NADPH Oxidases Revisited: From Function to Structure 303123751X, 9783031237515

Citation preview

Edgar Pick   Editor

NADPH Oxidases Revisited: From Function to Structure

NADPH Oxidases Revisited: From Function to Structure

Edgar Pick Editor

NADPH Oxidases Revisited: From Function to Structure

Editor Edgar Pick Department of Clinical Microbiology and Immunology, Sackler School of Medicine Tel Aviv University Tel Aviv, Israel

ISBN 978-3-031-23752-2 ISBN 978-3-031-23751-5 https://doi.org/10.1007/978-3-031-23752-2

(eBook)

# The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

This book is dedicated to Leora, my friend, companion, wife, confidant, and consoler at happy and difficult times. I know that it must be hard to live with someone who made the border between home and laboratory “intrinsically disordered” and whose working hours were “unstructured” (to borrow terms from the now fashionable structural nomenclature). I would not have had the courage to plan this book and initiate its editing without your support. You tried hard to persuade me that working on a book on NADPH oxidases was as satisfying as waiting eagerly for the result of a key experiment. In reverence to you and to the enthusiastic support of “Springer Nature,” I redefined “satisfying” as an oscillating parameter, fluctuating with my mood, reminiscences of decades at the laboratory bench, and the progress of the book. Your trust and encouragement also made me regard the term “Emeritus” not merely as a seal on the past but also as an unexpected prelude to a new beginning, culminating in the publication of this book. You made my transition from a generator of new data to a narrator of existing data as smooth and gentle as possible. To our daughters, Anat and Dana, a good slice of dedication is due, too. You competed valiantly with the “oxidase” for my time and attention but never for my love. I hope that you can forgive me in the hope that the collateral oxidative damage was minimal. Finally, dedication of this book could never be complete if it would not also be to the memory of my student, Yael Bromberg, whose radiant life with a passion for biochemistry was cut short prematurely. The moment, back in 1983, when Yael added cytosol to macrophage membranes not responding to arachidonic acid and got a burst of superoxide, “has made all the difference.” I wish that she could have read this closing sentence.

Preface

Our writings are as so many dishes, our readers guests, our books like beauty, that which one admires another rejects; so are we approved as men’s fancies are inclined. Pro captu lectoris habent sua fata libelli (Books have their own destinies) Robert Burton (1577–1640), The Anatomy of Melancholy (1621)

There is no better introductory sentence for this book than the above citation written in the inimitable graciousness of archaic English. I hope that “our readers” will be our “guests” and “approve” of the content and form of this book and of the choices that I made and forgive me for the omission of some undoubtedly important subjects. Before entering the process of planning this book on NADPH oxidases, I asked myself whether it was necessary and helpful. A colleague of mine, back from medical school days, who is a highly valued urologic surgeon, expressed doubts about the justification of publishing printed and even online versions of books in an era of unprecedentedly fast progress in science and medicine. The main reason for skepticism was his view of the contemporary reading public as glued to laptops, tablets, and mobile phones, thirsty for recent, fast, and brief information. A book comprising information on a rather large and heterogeneous biomedical subject and composed at a certain point in time cannot comply with such a demand. There was no choice but to “freeze” the information available at an arbitrary “now” and let the readers “melt the ice” and follow the newer developments independently on their own. A central point in this book is the accent on the history of how the discoveries were made which led to NADPH oxidases becoming such an important, actively investigated, and clinically relevant area of present biomedical research. The “NADPH oxidase” was born on July 16, 1908, the date on which Otto Warburg submitted the paper “Observations on the oxidative processes in the sea urchin egg” to the star biochemistry journal at the time, Hoppe-Seyler’s Zeitschrift fur Physiologische Chemie, describing the 6- to 7-fold increase in oxygen consumption on fertilization of the eggs. It took almost exactly 70 years to the publication of the finding by C.A. Foerder, S.J. Klebanoff, and B.M. Shapiro that oxygen consumption by the eggs was accounted for by the production of hydrogen peroxide, which served as substrate for a peroxidase-mediated crosslinking of tyrosyl residues in the egg membrane. The authors noted, with remarkable foresight, the similarity of the process to the response of leukocytes to phagocytosis. About three decades later, G.M. Wessel and coworkers identified the responsible enzyme as a dual oxidase (Udx1) regulated by calcium. A telling example for the strange ways in which discoveries are made and the attention they provoke (or do not provoke) is that the first observation of a process mediated by the NADPH oxidase in phagocytes was made in 1932 but was ignored for almost three decades. The senior author of the publication, R.W. Gerard, was not aware of the paramount importance of the discovery and did not consider it as one of his major achievements. The junior author, C.W. Baldridge, published several papers on oxygen consumption by other cells as a metabolic

vii

viii

parameter in various diseases but his name disappeared from the literature a short time after the publication of the phagocyte paper for an unknown reason.1 A quite similar case was the precocious and hardly noticed identification of the heart of the phagocyte NADPH oxidase, cytochrome b558, in horse leukocytes, by H. Hattori, and rabbit leukocytes, by Y. Shinagawa and coworkers, many years before its rediscovery in human phagocytes by A.W. Segal and O.T.G. Jones in 1978. The history of NADPH oxidase research can be divided into two periods. In the first, the focus was on the phagocyte and its role in host defense and led to the identification and characterization of the catalytic (Nox2) and regulatory (cytosolic) components of the enzyme complex, including the genetic basis of its function in health and disease (the molecular basis of chronic granulomatous disease (CGD)). The “biochemistry-centered” period of the seventies and eighties of the past century also led to deciphering the steps of the electron flow from NADPH to dioxygen through sequential redox centers and to defining the assembly of the cytosolic regulators with cytochrome b558 as the process responsible for NADPH oxidase activation. A most significant episode and an early consequence of the appearance of bioinformatics tools was the finding of the sequence homology between Nox2 and ferredoxin-NADP+ reductase (FNR), with emphasis on the FAD- and NADPH-binding domains. A few years later, the similarity between the bis-heme motif of ferric reductases (FR) and Nox2 was noticed, leading to the consolidation of the concept that Nox2 was evolutionarily a hybrid construct with its cytosolic part originating in the FNR superfamily and its transmembrane part, in the FR superfamily. The realization that a single molecule harboring all redox centers, carried electrons from NADPH via FAD and two hemes to dioxygen, represented a conceptual breakthrough and served as a common denominator for all members of the Nox family, to be discovered in the future. It was, by a rather remote comparison, evocative of the maturation of religions from heathen paganism to monotheism. The second period was initiated by the perception that Nox2 was but a member of a large group of proteins (the seven-members Nox family) with similar structures but a multitude of functions and wide representation within and outside the animal kingdom, from prokaryotes to eukaryotes. The presence of members of the Nox family in a variety of cells, tissues, organs, and organisms, together with the definition of their multiple functions led to the conclusion that the initial concept of Noxs being principally involved in host defense or causing auto-oxidative damage was much too restricted. The contemporary view is that Noxs are essential mediators of vital biological processes comprising signal transduction, gene expression, cell differentiation, oxygen sensing, protein cross-linking, muscle contraction, and thyroid hormone synthesis, to mention only some, possibly also serving as an explanation for the evolutionary conservation of molecules generating potentially toxic reactive oxygen species. This golden era also saw the identification of novel regulatory components and mechanisms acting on the newly discovered members of the Nox community. Noxs resemble the mythological god Janus, depicted as having two faces representing duality. Indeed, in parallel to their role in normal physiology, mutations in, and over- or under-expression or defective regulation of specific Noxs, are causative or enhancing factors in numerous pathological situations. The association of aberrant Nox function with disease made Noxs desirable druggable targets for the design of compounds to serve as therapeutic agents in clinical medicine. The birth of molecular biology and its offspring, molecular genetics, and the emergence of bioinformatics had a major impact on the methodologies used in NADPH oxidase research. Thus, classical biochemical methods were used to identify and purify cytochrome b558 and the cytosolic components associated with its activation. This was followed by the introduction of reverse genetics for the cloning of Nox2 and of cloning as an accessory approach, following prior isolation of the protein, as applied to the cytosolic components. The bioinformatics The author thanks Dr. William Nauseef for attempts to find biographical details related to the fate of Dr. C.W. Baldridge.

1

Preface

Preface

ix

revolution led to searches for translated nucleotides in genome databases of expressed sequence tags, becoming the almost exclusive approach to the discovery of new members of the Nox family and the homologues of the regulatory proteins. Two more advances had a marked impact on the elucidation of the structure and function of Noxs. One was the quantal leap from X-ray crystallography-based structural studies to cryogenic electron microscopy (cryo-EM) that allowed visualizing Nox structures at near-atomic resolution. The second development was the revolution in the ability to predict threedimensional structures of proteins from amino acid sequences, as illustrated by the programs AlphaFold and RoseTTAFold, which have already demonstrated amazing accuracy and confirmation by the experimental validation of the predicted structures. Models of every known Nox are now a routine component of protein sequence databases found in bioinformatics resource portals, next to X-ray- or NMR-derived actual structures, when such are available. To use Winston Churchill’s words, the predictive power of these programs is merely “the end of the beginning” and their rapid evolution is certain to lead to the visibility of cofactors, metal ions, and other ligands, the role of which in Nox function is paramount. Considering that protein– protein interactions are a key feature of NADPH oxidase assembly, a major challenge for future programs will be the ability to predict the structure of protein–protein interfaces, many of which are likely to involve unstructured (intrinsically disordered) regions, which may become ordered following interaction with another protein or ligand. Another issue that might hinder progress in solving and predicting Nox function from Nox structures is that Nox activation is the consequence of conformational dynamics occurring on a time scale. Prediction programs and cryoEM are catching “frozen” states of the protein in the course of switching from the resting to the activated conformation, offering superb static models but not the equivalent of a motion picture. Because of the fact that activation of three out of seven members of the Nox family is dependent on assembly with cytosolic components, future structural studies will have to be able to deal with multi-molecular complexes centered on the Nox component embedded in a milieu mimicking the membrane lipid bilayer (nanodiscs, lipodiscs, peptidiscs) interacting with cytosolic components. A major impetus for the extent and intensity of Nox investigations was the conviction that Noxs are significant participants in the pathogenesis of diseases. I hope to be excused for expressing the personal opinion that the Nox community might have rallied rather overenthusiastically and with insufficient rigor behind the fast growing list of Nox-disease connections. In addition to the canonical Nox-related diseases, comprising the various forms of CGD and hypothyroidism, the list of pathologic processes linked to the participation of one or more Noxs comprises cardiovascular, respiratory, hepatic, gastrointestinal, endocrine, neurological, skeletal, muscular, neoplastic, and many other disorders. Causal Nox-disease connections would, in most cases, fail to comply with even the most liberal interpretation of Koch’s postulates, in which the term “microorganism” is replaced by over- or under-expression a particular Nox. A direct consequence of the dominance of the Nox-disease connection is the enormous investment by both academia and industry in the design of drugs for the treatment of a plethora of diseases based on the evidence of the involvement of specific Noxes. As illustrated by the chapters in this book, their authors did not share the skepticism of the editor and discussed extensively, confidently and eloquently Nox-associated diseases and compounds exhibiting pan-Nox or Nox-specific inhibitory ability. The traditional hypothesis-driven drug design, the cherished hallmark of academic research, was replaced by ever larger high-throughput screening for druggability of Noxs by both biotech companies and academic researchers. The view was also expressed that the customary therapeutic approach of inhibition of Nox activity should be complemented by approaches enhancing the expression or activity of specific Noxs. Whether the widely shared enthusiasm for the Nox-disease connection and the resultant optimism for Nox targeting by drugs, as an effective therapeutic means, will fulfill the expectations cannot be answered at present and only the future can tell. The uncertainty of predictions is best illustrated by the statement in the

x

Babylonian Talmud that “From the day that the temple was destroyed, prophecy was taken from the prophets and given to fools and children.” A social phenomenon of interest is that involvement in basic and clinical Nox research has bonded investigators in a community very similar to what Ludwik Fleck called a “thought collective” in his unjustly little known book “Genesis and Development of a Scientific Fact.” A “thought collective” was defined as a community of persons mutually exchanging ideas or maintaining interaction, which also acted as a special “carrier” for the historical development of a field of thought. Fleck also introduced the term “thought style,” as domineering a thought collective. In elegantly styled phrasing, he stated that although the thought collective consisted of individuals, it was not a sum of them and the individual within the collective was hardly ever conscious of the dominant thought style exerting a compulsive force upon his thinking, with which it was difficult to be at variance. Past examples of such occurrences are the long-lasting predominance of the concept that the NADPH oxidase was a flavoprotein and nothing more (popular with the cytochrome deniers), followed by the idea that the enzyme was composed of two parts, a flavoprotein and a b cytochrome (the binary model), supplemented by miniproposals of additional components being involved, such as ubiquinone and iron-sulfur proteins. “Thought collectives” of the period were resistant to revolutionary proposals, such as the idea that the NADPH oxidase was both a flavoprotein and a b cytochrome. Chapters in this book, written by leading investigators, offer overviews of canonical Noxs (Part II) and their regulators (Part III), and of tools to identify Noxs, methods to measure their products, selective Nox inhibitors, and interaction of Noxs with non-Nox partners (Part IV). Further sections describe non-mammalian Noxs, of amoebas, yeasts, fungi, plants, nematodes, arthropods, and zebrafish (Part V). Two chapters deal with structural aspects of Duoxes and Nox “relatives” (Part VI), and two chapters, with basic, clinical and therapeutic aspects of CGD (Part VII). The closing chapter consists of an optimism-filled prediction for successful clinical translation of basic Nox research (Part VIII). With the intention to dilute somewhat the concentration of “heavy” science and supplement the rigor of biochemistry, molecular biology, and genetics with a compensatory dose of humanism, I invited prominent investigators, registered or unregistered members of the Nox guild, to narrate their personal journey to their discoveries, frequently the result of taking the road “less traveled by” (Robert Frost). Unfortunately, some of the founding fathers and mothers, who sowed the seeds of the solitary Nox2 tree that became the jungle-like Nox forest, are no longer with us. I, thus, felt that it was important that their contributions, ways to approach a scientific problem, and personal traits, including idiosyncrasies, should be narrated by the next generation, who had the unique opportunity to work with them as colleagues, postdocs, or students. I hope that both the personal recollections and the very intimate biographical sketches will convince the reader that even the most recent result is but a station in an open-ended selfcorrecting evolutionary process with occasional wrong turns. The contents of this book are proof for the quality, diversity, openness, and optimism of the present Nox “thought collective” and serves as an assurance for the bright future of Nox research. Christopher Columbus was awarded a “research grant” to find a novel way to the Indies and discovered America; let us hope that further work on Noxs will open equally unforeseen vistas of similar significance. Last but not least, a confession. My intention to edit this book had three roots. The first was the realization of the fact that the maturity of the Nox field justified an up-to-date compendium, to supplement the excellent “Methods and Protocols” volumes published recently. The second was the wish to include in the book personal recollections of investigators who had a crucial impact on the field as well as testimonies of scientists who had the privilege of working closely with leading figures of the past. The third root was, I admit, using the opportunity of editing this book as a compensation for having to abandon many years of “hands on the bench” work on the wonderful subject of Noxs.

Preface

Preface

xi

My thanks go to the authors of the chapters in this book for the excellence of their contributions and for the tolerance and goodwill exhibited during the lengthy process leading to the publication of what is truly “their” book. I thank Springer Nature for their effort to turn a dream into reality. In particular, I would like to thank Mr. Bibhuti Bhusan Sharma, for his key role in helping me navigate the tempestuous waves of book editing, and Drs Alexis Rivas, Amrei Strehl, Sabine Schwarz, and Paul Roos, for making sure that I do not get lost in the publishing labyrinth. Tel Aviv, Israel

Edgar Pick

Contents

Part I 1

History, Recollections, and Homages

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Edgar Pick

3

2

The Phagocyte Oxidase: The Early Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . John T. Curnutte and Alfred I. Tauber

65

3

Reflections on My Life in Noxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. David Lambeth

81

4

The Discovery and Characterisation of Nox2, a Personal Journey . . . . . . . . . Anthony W. Segal

91

5

Reminiscences on Positional Cloning of X-CGD Gene (Aka CYBB, gp91phox, Nox2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Stuart H. Orkin

6

On Katsuko Kakinuma: Spectroscopic Studies of Redox Centers in NADPH Oxidase – “Identifying and Observing the Key Players that Pass an Electron to Oxygen” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Hirotada G. Fujii and Lucia S. Yoshida

7

Pierre Vignais, from One Respiratory Chain to Another. . . . . . . . . . . . . . . . . 119 Marie-Claire Dagher

8

Gary M. Bokoch, the Rac-n-Rho Man: His Fascination with Rho-GTPases . . . 123 Becky A. Diebold

9

History and Discovery of the Noxes: From Nox1 to the DUOXes . . . . . . . . . . 133 Albert van der Vliet

Part II

Canonical NADPH Oxidases

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Thomas L. Leto and Miklós Geiszt

11

Physiological Functions and Pathological Significance of NADPH Oxidase 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Yoko Nakano and Botond Bánfi

12

Nox4: From Discovery to Pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Louise Hecker, Kosuke Kato, and Kathy K. Griendling

xiii

xiv

Contents

13

Nox5: Molecular Regulation and Pathophysiology . . . . . . . . . . . . . . . . . . . . . 215 Livia L. Camargo, Francisco Rios, Augusto Montezano, and Rhian M. Touyz

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2 . . . . . . . . . . . . . . . . . . . . . . . . 229 Françoise Miot and Xavier De Deken

Part III

NADPH Oxidase Regulators

15

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 Pham My-Chan Dang and Jamel El-Benna

16

The NADPH Oxidase Activator p67phox and Its Related Proteins . . . . . . . . . . 263 Hideki Sumimoto, Akira Kohda, Junya Hayase, and Sachiko Kamakura

17

p40phox: Composition, Function and Consequences of Its Absence . . . . . . . . . 275 Taco W. Kuijpers and Dirk Roos

18

Rho Family GTPases and their Modulators . . . . . . . . . . . . . . . . . . . . . . . . . . 287 Yuan Lin and Yi Zheng

Part IV

Tools, Inhibitors, and Neighbors

19

Tools to Identify Noxes and their Regulators . . . . . . . . . . . . . . . . . . . . . . . . . 313 Katrin Schröder

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323 Jacek Zielonka and Matea Juric

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives . . . . . . . 343 Christopher M. Dustin, Eugenia Cifuentes-Pagano, and Patrick J. Pagano

22

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes . . . . . . 379 Tiphany Coralie de Bessa and Francisco R. M. Laurindo

Part V

Non-Mammalian NADPH Oxidases

23

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399 Laurence Aubry and Bernard Lardy

24

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413 Michael Breitenbach, Mark Rinnerthaler, Jiri Hasek, Paul J. Cullen, Campbell W. Gourlay, Manuela Weber, and Hannelore Breitenbach-Koller

25

NADPH Oxidases in Fungi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429 Daigo Takemoto and Barry Scott

26

Plant NADPH Oxidases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445 Gad Miller and Ron Mittler

27

Nematode Noxes: The DUOXes of Caenorhabditis elegans . . . . . . . . . . . . . . . 467 Danielle A. Garsin

Contents

xv

28

NADPH Oxidases in Arthropods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477 Ana Caroline P. Gandara and Pedro L. Oliveira

29

NADPH Oxidases in Zebrafish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 S. M. Sabbir Alam and Daniel M. Suter

Part VI

Structure

30

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex . . . . . 507 Jing-Xiang Wu, Ji Sun, and Lei Chen

31

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP) Enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . 521 Wout Oosterheert, Sara Marchese, and Andrea Mattevi

Part VII

Pathology

32

Chronic Granulomatous Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537 Marie José Stasia and Dirk Roos

33

Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557 Giorgia Santilli and Adrian J. Thrasher

Part VIII 34

Future

Quo Vadis NADPH Oxidases: Perspectives on Clinical Translation . . . . . . . . 575 Ulla G. Knaus, Ajay M. Shah, and Victor J. Thannickal

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587

Part I History, Recollections, and Homages

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique Edgar Pick

Abstract

This chapter narrates the history of the almost a centurylong work on the phagocyte NADPH oxidase, from the first report on the “respiration” of dog leukocytes phagocytizing bacteria (1932) to the contemporary efforts to predict and solve the structure of the enzyme. The main stations on this stellar road are discussed within their historical context, not omitting the occasional wrong turns and dead ends. The first pivotal findings were that the enzyme used NADPH as reducing power and that the primary product was superoxide, a radical differing from molecular oxygen by the mere addition of a single electron. Solving the intricacies of this “modest” chemical reaction rested on basic discoveries made at the laboratory bench, combined with work on the genetic basis and clinical aspects of chronic granulomatous disease. The most significant discoveries were: the flavoprotein nature of the enzyme; identification and purification of cytochrome b558; cloning of Nox2 and p22phox; the bis-heme motif; presence of all redox centers in Nox2; design of the cell-free system; discovery and identification of the cytosolic components; cloning of p47phox and p67phox; involvement of Rac and RhoGDI, and solving the mechanism of the assembly of the functionally active NADPH oxidase complex. Keywords

NADPH oxidase · Superoxide · Cytochrome b558 · Nox2 · p22phox · p47phox · p67phox · Rac · RhoGDI · CGD

E. Pick (✉) Department of Clinical Microbiology and Immunology, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel e-mail: [email protected]

1

Prehistory A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die and that the growing generation is familiar with the truth from the outset Max Planck, Scientific Autobiography and Other Papers (1949)

1.1

The “Respiration” of Phagocytes: The First Paradigm

All attempts to narrate the genealogy of NADPH oxidases (abbreviated to Nox, for the singular form)1 leads to work focused on trials to identify the source of the energy required for the mechanical act of phagocytosis. This fact appears strange today due to our awareness of the lack of connection of the phagocyte Nox (Nox2)2 to the process of particle engulfment. An important finding was that leukocytes of patients suffering from an inborn defect in fighting certain infections, linked to phagocytes, had unaltered ability to take up particles [1]. The report describing normal phagocytosis In this chapter, I used the general term “NADPH oxidase” and the abbreviated form “Nox”, approved by the Human Genome Organization (HUGO) Human Gene Nomenclature Committee (HGNC) (https:// www.genenames.org) in 2000. This applies to the Nox homologs (also termed Nox isoforms) Nox1, Nox2, Nox3, Nox4, and Nox5, the “numbered” forms referring to the catalytic subunits, only. The approved abbreviated term for “dual oxidase” is “Duox” and the catalytic subunits are known as Duox1 and Duox2. I used “Noxs” as the plural of “Nox”, for lack of a better term. 2 This chapter discusses principally the history of research focused on the phagocyte NADPH oxidase (phagocyte Nox). This is also known as cytochrome b558, flavocytochrome b558, or cytochrome b-245 and is a heterodimer consisting of two subunits, known since 1991, as gp91phox and p22phox. In accordance to the rules of HGNC, the catalytic subunit gp91phox is also known as Nox2 (somewhat unjustly, considering the fact that it was the first to be discovered and characterized). The cytosolic components, p47phox, p67phox, p40phox, and Rac (1 or 2) are not integral parts of the phagocyte NADPH oxidase but assemble with cytochrome b558, to form the phagocyte NADPH oxidase complex. 1

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_1

3

4

E. Pick

by leukocytes of patients with Chronic Granulomatous Disease (CGD) was published in 1966, 34 years after a one and one-third pages-long gem of a paper, with the prophetic title “The Extra Respiration of Phagocytosis” [2]. By researching for writing this chapter, I discovered that the senior author, Ralph Waldo Gerard (1900–1974), was a well-known neurophysiologist at the University of Chicago, who was apparently unaware of the groundbreaking quality of the paper he coauthored with C. W. Baldridge. I was wondering whether he was ever interested in the oxidative burst; in his biography, there is no mention of such interest. Using the legendary Warburg manometer, the authors showed that dog leukocytes phagocytizing Micrococcus Luteus in the presence of serum (opsonization?) showed a four-fold increase in oxygen consumption within 15 min; a remarkable finding was the total absence of augmented oxygen uptake when leukocytes were phagocytosing India ink (suspension of carbon particles). Although the authors interpreted the enhanced respiration as being required for particle uptake, the lack of response to carbon particles suggested a different mechanism. A little known contemporary report described different results, expressed in the lack of enhanced oxygen uptake by leukocytes obtained from the peritoneal cavity, following the phagocytosis of starch, carbon particles and staphylococci [3].

1.2

A Dominant Biochemist Enters the Field: Manfred L. Karnovsky and a Fluctuating Paradigm

The interest of Manfred Karnovsky in leukocyte metabolism was aroused by studies on possible metabolic changes in phagocytosed tubercle bacilli. As a control, he investigated the metabolism of the host cell (the phagocyte), which turned into a life-long preoccupation that had a major influence on the future of the field. Described in his first publication on this subject, Karnovsky observed an increase in oxygen consumption by elicited peritoneal guinea pig leukocytes and monocytes consequent to phagocytosis of Mycobacterium tuberculosis bacilli [4]. The enhanced oxygen consumption was also elicited by phagocytosis of nonbacterial particles but the initial tendency to look upon the event as an energetic requirement of the mechanical act of particle uptake was abandoned in light of the finding that phagocytosis proceeded normally under anaerobic conditions [5]. It is of interest for the history of ideas that Karnovsky noted that the increment in oxygen consumption was accompanied by an increase in the hexose monophosphate shunt (HMPS) but stated that these phenomena “appear to be concomitants of particle intake and, perhaps, not of prime importance in the process”. He also declared that the resistance of oxygen uptake to cyanide indicated that the process was “not cytochrome

linked” (meaning “not mitochondria-linked”). Who could have imagined that it was all about a very different cytochrome? A good example for the convergence of ideas, just ahead of a paradigm shift, was the almost simultaneous work originating in Zanvil Cohn’s group at the Rockefeller Institute [6]. They confirmed that phagocytosis was fueled by glycolysis and that cyanide inhibition of oxygen uptake had no effect on phagocytosis; however about 30% of oxygen uptake was not inhibited, a phenomenon defined, as in the work of Karnovsky, as “not cytochrome linked” and was suggested to be related to the process of post-phagocytosis killing of bacteria. In retrospect, both the Harvard and Rockefeller groups discovered the cyanide resistance of the yet hypothetical “non-cytochrome” cytochrome responsible for the killing of bacteria, to be identified about two decades later. Five years after the publication of the findings described in Refs. [5, 6], an attempt was made to establish the enzymatic basis of the respiratory stimulation but the paradigm shift was partial [7]. Two findings, to stay with us, were that the enhanced oxygen uptake and the stimulation of the HMPS were interlinked and that both processes were resistant to cyanide. In a mixture of findings that are true to this very day and of some that were wrong, Karnovsky suggested that the enzyme responsible was an NADH oxidase located in the cytosol, which contained flavin adenine dinucleotide (FAD), acted as a diaphorase reducing nitro blue tetrazolium (NBT), and which “presumably” generated hydrogen peroxide (H2O2). The paper contains an interesting polemic with Judah Quastel, and a brief rather negative reference to a seminal paper by Filippo Rossi, the only two investigators who, at the time, were focused on the same subject. The “evolution” of Karnovsky’s views of what we call today Nox is interesting to follow for both the readiness to correct and the difficulty to abandon former beliefs. A key element that promoted the understanding of the function and structure of Nox was the clinical input derived from the early reports dealing with the biochemical basis of CGD (reviewed in [8]). In an article published in 1973, Karnovsky offered direct evidence for the production of H2O2 (though quantitatively, very low) but maintained his former claims that the enzyme was a NADH oxidase, present predominantly in the cytosol, and generating H2O2 as the primary product [9]. The description by Rossi in 1964 of a granule-bound cyanideinsensitive enzyme with a marked preference for NADPH, was mentioned, as well as the difficulty of explaining how H2O2 generated in the cytosol could reach the phagocytosed bacteria supposed to be its target. An interesting demonstration of how slowly did our knowledge of Nox progress in the past is a publication labeling the enzyme as “An elusive Pimpernel” [10]. The production of superoxide (O2•-) was acknowledged, but several assumptions, listed as facts, were erroneous, such as NADH being the reducing substrate, the

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

5

enzyme being located in the cytosol, that there were multiple oxidases involved, and that activity assessed in subcellular fractions did not represent that of the intact cell. The rather graceful way in which scientific controversies were dealt with at the time is shown by the statement by the authors that “The quest for truth in this field (attempts to elucidate the enzymologic basis of the respiratory phenomenon) has become a rather contorted history over the years, characterized by some controversy, but, perhaps fortunately little acrimony”. It took another 7 years for Karnovsky to admit that the enzyme he studied so intensively for three decades was in fact an NADPH oxidase, which generated O2•- and was localized not in the cytosol but, as expected, in the plasma membrane and phagosomal membrane resulting by the invagination of the former [11]. This review paper also gave credit to the fundamental discoveries of Bernard Babior, among which was the finding that the enzyme was latent in resting phagocytes but became activated upon stimulation, a clear indication for the existence of a regulatory mechanism, the elucidation of which keeps investigators busy to the present day.

1.3

Karnovsky’s Alter Ego: Judah Hirsch Quastel

The Canadian biochemist Judah Quastel (Fig. 1.1) was inspired by the work of Karnovsky and was engaged in similar studies but was convinced that the enhanced oxygen consumption was linked to the “bactericidal processes accompanying phagocytosis” (exact citation from Ref. [12]). Oxygen consumption by elicited guinea pig peritoneal leukocytes was stimulated by uptake of bacteria but not by that of inorganic particles and was associated with an increase in the HMPS pathway. There was a rather indirect demonstration of phagocytosis-associated H2O2 formation but the mechanism proposed was oxidation of NADPH or NADH (NADPH being more effective) catalyzed by flavin mononucleotide (FMN). As far as a paradigm shift was concerned, this article comprised three essential elements associated with the contemporary view of Nox structure and function: (1) The participation of NADPH (as opposed to the incorrect proposal by Karnovsky of NADH); (2) The involvement of a flavin (although, the wrong one), and (3) The generation of H2O2 (although, the proposed pathway was incorrect). An article published 2 years later provided a synthesis of the Quastel model by emphasizing the superiority of NADPH over NADH, the role of the HMPS in the reduction of NADP+ to NADPH, the superiority of FMN over FAD, the enhancement of the reaction by Mn2+, and the generation of H2O2 as a product of oxidation of NADPH in the presence of oxygen [13]. Quastel went down many wrong alleys but I urge the present day reader to stick with

Fig. 1.1 Judah Hirsch Quastel (1899–1987)

NADPH, replace H2O2 by O2•- and FMN by FAD, and get a clumsy but rather modern Nox.

2

History

2.1

A Gentleman from Verona: Filippo Rossi

Experience is by industry achieved, And perfected by the swift course of time William Shakespeare, Two Gentlemen of Verona

The Shakespearean “roots” of Filippo Rossi (Fig. 1.2) are evident by having done his groundbreaking work in Padova (The Taming of the Shrew) and spending most of his scientific life in Verona (Two Gentlemen of Verona). The body of work, related to Nox, coming from the group led by Filippo Rossi is of a size which makes it almost impossible to review and one can argue about what parts belong to the Bible and which should be considered Apocrypha. It is rather surprising that the description of what looked like an “anomaly” in the metabolic response of leukocytes to phagocytosis (in the sense used by Kuhn [14]) was hardly noticed, in spite of the fact that it clearly belonged to the Bible. In two modestly written papers, published in two unpretentious journals (both appearing today under different

6

E. Pick

activity in the granules of phagocytosing cells suggested the existence of an electron transport system (diaphorase?), and (4) The enhanced Nox activity could be the result of changes in phosphatide turnover, based on a report by Karnovsky on phagocytosis-induced changes in the phospholipid composition of leukocytes [19].

2.2

Fig. 1.2 Filippo Rossi (1926–2022)

names and well before the time that the “impact factor” dictated journal choice), Rossi and Zatti laid the foundations of what became the start of the modern age of Nox research [15, 16]. The study showed that phagocytosis of Staphilococcus aureus or Micrococcus luteus (the latter bacteria having been used in the first description of an enhanced oxygen uptake by phagocytosing leukocytes [2]) by guinea pig peritoneal leukocytes was associated with an increase in oxygen consumption, which was cyanide resistant and accompanied by a stimulated HMPS. The phagocytosisinduced oxygen uptake was unrelated to the production of lactic acid. However, the paramount finding was that oxygen uptake was closely associated with oxidation of NADPH by an enzyme localized exclusively in the leukocyte granules. The granules from phagocytosing cells exhibited an increase in NADPH oxidase activity described as “dramatic”, exceeding the much lesser increase in NADH oxidase activity. As always, new discoveries, as original as they may be, do not occur in isolation from former findings. Thus, Rossi and Zatti followed the work of Quastel [12, 13] and added Mn2+ to the system, to find, indeed, enhanced NADPH and NADH oxidation. They also found that addition of the electron carrier menadione to intact cells or to granules derived from resting cells reproduced the metabolic changes elicited by phagocytosis (see comment on the Mn2+ and menadione artefacts, further down in the text). Luckily, they did not dwell on these results and rushed ahead, reaching the following conclusions: (1) Stimulation of Nox did not involve disruption of the Nox-containing granules, indicating that the Nox-containing granules represented a category distinct from lysosomes, in accordance with findings made two decades later that the bulk of the core of the enzyme, cytochrome b558, was stored in what became known as “specific granules” [17, 18]; (2) Stimulation of NADPH oxidation in the granules, caused by phagocytosis, was the primary event that promoted the oxidation of glucose by the HMPS; (3) The enhanced Nox

A Minor Diversion

The present day reader will be intrigued by the seeming dependence of the phagocytosis-elicited metabolic changes on Mn2+, found by Quastel [13] and Rossi [15]. Fortunately, the claim was revisited a decade later, when it was found that Mn2+-dependent NADPH oxidation by leukocyte granules was a nonenzymatic chain reaction involving O2•-, with the granules serving as the source of O2•- [20, 21]. It was only in 1998 that the menadione-stimulated Nox from resting cells was found to be distinct from and unconnected to the enzyme responsible for O2•- production by stimulated phagocytes [22].

2.3

Two Stars are Born: The Person and the Radical

Therefore the problem is not so much, to see what nobody has yet seen, but rather to think concerning that which everybody sees, what nobody has yet thought Arthur Schopenhauer, Parerga und Paralipomena (1851) The superoxide radical is a minor, but not trivial, product of oxygen reduction Irwin Fridovich [23]

It is interesting that one of the major developments in the history of Nox came not as an expected consequence of earlier findings (what Kuhn defined as “normal science” [14]) but by a combination of three components: noticing that something was missing, an external input, and the right person to link the first two components. Thus, by 1973 it was clear that phagocytizing leukocytes consume oxygen, oxidize NADPH, and regenerate NADPH by activating the HMPS. The “purpose” of this metabolic event was unrelated to the mechanical aspects of particle engulfment and what was missing was a “product”. A purpose (a bactericidal effect on phagocytosed microorganisms) and a product (H2O2), were proposed by Quastel [12] but the correct biochemical sequence culminating in the generation of H2O2 remained unknown. It was Bernard Babior who found that the Nox of leukocytes generates O2•- and that this “minor but not trivial product of oxygen reduction” was the enzyme’s raison d’ être. Interestingly, one root of the hypothesis-driven search for O2•- as a product of phagocytosing leukocytes was the

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

work of Irwin Fridovich, showing that biologic systems can produce O2•- [23] and that cells, including aerobic microbes, contain the enzyme superoxide dismutase (SOD), destroying O2•- [24, 25]. A second root was the finding of Seymour Klebanoff that hypochlorous acid, generated by H2O2, leukocyte myelopereroxidase and halides, was bactericidal, defining the killing mechanism of H2O2 [26]. A third root was the seemingly paradoxical finding that patients with congenital deficiency in myeloperoxidase were efficiently combating infections, suggesting that H2O2 might not be the sole killing molecule [27]. It was the genius of Babior to hypothesize that leukocytes must produce another oxygenderived bactericidal molecule and connecting this to the, at the time, little known O2•- was “to think what nobody has yet thought” (see Chap. 2 by J.T. Curnutte and A.I. Tauber, for an intimate description of the events around this discovery). The paper, coauthored with John Curnutte and Ruby Kipnes, published in 1973, described the production of O2•- by human granulocytes phagocytizing latex particles [28]. O2•was detected by its property to reduce ferricytochrome c; the identity of the reducing species proven by its sensitivity to SOD, and O2•- generation was shown to be stimulated by particle uptake. In its four pages, the distinct observations of the events of the post-phagocytic metabolic changes, made over four decades, were concentrated into a simple chemical reaction: 2O2 þ NADPH → 2O2 • - þ NADPþ þ Hþ Finally, oxygen uptake, Nox activity, the secondary stimulation of the HMPS, and killing of bacteria made biological sense. The paper was cited 3731 times untill May 2023. O2•production by stimulated leukocytes became “normal science” and was confirmed by numerous groups (reviewed in [29]). Opposition to the finding was almost inexistent and the fact that the only such attempt came from the laboratory of another “founding father” of Nox research [30] and provoked a response from Babior [31] made it a curiosity in the triumphal acceptance of O2•- as the primary reactive oxygen species (ROS). It also became clear that the origin of H2O2 was by dismutation of O2•-, which was spontaneous and not mediated by SOD [32].

2.4

The New “Yellow Enzyme”

The critical discovery that Nox was a FAD-containing flavoprotein had its parental roots in the “prophetic” statement by Quastel that “a system catalysed by flavin is involved in the oxidation of the reduced di- and tri-phosphopyridine nucleotides by leukocytes” (though, he meant FMN) [12]). The discovery was made in the rush by many groups to

7

isolate and purify an active O2•- producing enzyme from stimulated cells. The basic principle of these experiments was to expose leukocytes to a range of particles to be phagocytosed, followed by the disruption of the cells and collection of the granular fraction obtained by centrifugation. The sedimented granules were expected to contain the activated Nox and were assayed as such, or following solubilization by a variety of detergents, for NADPH-dependent O2•- production [33]. Support for the legitimacy of using granules from stimulated phagocytes as bona fide representatives of activated Nox came also from the finding that the particulate fraction from leukocytes of patients with the X-linked form of CGD failed to produce O2•- [34]. In yet another landmark paper, surprising by the modesty of its statement, Babior showed that Nox-containing particles lost their activity when assayed in the presence of a low concentration of a detergent and that the loss could be prevented by the presence of FAD [35]. Babior, most probably, added FAD because, to cite literally from the paper, “a role for the flavin as a prosthetic group which transports reducing equivalents from NADPH to oxygen is suggested by the fact that virtually every flavoprotein described to date has been a redox enzyme”. Once the hypothesis existed, Babior turned to the “bench” and showed that: (1) FAD acted on the enzyme derived from stimulated phagocytes; (2) The effect was specific for isoalloxazine (the redox center of FAD) since adenosine monophosphate (AMP) and adenosine diphosphate (ADP) did not work (but this was not all the story because FMN and riboflavin were also inactive), and (3) FAD was unable to repair the lack of activity of enzyme from cells of CGD patients. The finding that the effect of FAD was evident only on detergent-treated particles suggested that the detergent released FAD from the enzyme, which in turn suggested that FAD was a prosthetic group of the enzyme which should be regarded as a flavoprotein. The nature of the FAD—apoprotein bond (covalent or non-covalent) was brought up, though remained unsolved. Three follow up papers, two of which resulting from collaboration with Theodore Gabig, added quantitative information on the ability of FAD and inability of FMN to support activity and on the optimal timing of FAD supplementation [36–38]. The fact that the last paper in the series was published in 1981 [38] raises questions about the effectiveness of crossfertilization of scientific concepts, shown by the absence of any mention of cytochrome b558 being part of Nox in spite of the fact it was identified and its existence published 3 years earlier. In the wider science-historical context, the major discovery that Nox was a member of the large family of flavoproteins and the prediction by Babior that the enzymebound FAD undergoes reduction and oxidation at a rate equal to the turnover of the enzyme, became the “boon and the

8

E. Pick

bane” of future Nox research, for two reasons. The first was that it led to excessive accent being placed on the flavoprotein, considered to be the only redox center of the enzyme and, in a way, delayed the acceptance of cytochrome b558 as a key component of Nox. The second was that it fostered the “binary” model of Nox, as consisting of a NADPH-binding FAD flavoprotein and a cytochrome, containing the hemes. It took 15 years until the “monotheistic” model of Nox, in which all redox centers were present in a single molecule, became accepted.

O2 -

1e-

NADPH

O2

FAD·Enz

?

2e-

FADH2·Enz 1e-

1eCyt bred

2.5

The Flavoprotein Becomes the Prima Donna: The Era of the Binary NADPH Oxidase

Further studies on the flavoprotein nature of Nox, which followed Babior’s basic discovey, overlapped an intense effort by multiple groups to isolate, purify, and characterize the enzyme. Such studies were performed by exposing intact leukocytes to phagocytic stimuli, followed by cell disruption and the separation of a particulate fraction, based on extensive evidence showing that this contains all of the active enzyme. There was one dissident attempt to isolate the enzyme from a granule fraction of unstimulated cells and activate it by dialysis but the fact that the activity did not show preference for NADPH and was also present in material derived from cells of a CGD patient indicated that it was not related to Nox [39]. With plans to purify the enzyme in mind, most investigators attempted to solubilize the enzyme in the presence or absence of a protective agent, usually a phospholipid. An important observation made in the course of these efforts was that the FAD analogue, 5-carba-5-deaza-FAD, capable only of two-electron transfer, could not replace FAD and acted as an inhibitor of Nox activity [40]. This paper by the group of Alfred Tauber is a highly instructive illustration of the predominant model of Nox in the early eighties of the last century, characterized by the accent placed on the flavoprotein and minor interest in the participation of cytochrome b558. The preparation was derived from human leukocytes phagocytosing opsonized zymosan or exposed to the increasingly popular soluble stimulant, phorbol myristate acetate (PMA). The subcellular particles were solubilized by detergent and a post-centrifugation supernatant used as the activated Nox. This was found to contain a cytochrome with the spectral characteristics of cytochrome b558 but it could not be reduced by NADPH under anaerobic conditions, a finding that the authors used as an argument against its participation in O2•- production. They presented a close to correct scheme of the electrons flow in the enzyme (Fig. 1.3), while stating that the study did not allow determining whether FAD or heme was the site of oxygen reduction.

O2 -

Cyt box NADP+

O2

Fig. 1.3 Schematic representation of electron transfer during O2•generation proposed by Tauber and coworkers at the time when it was not known whether the site of oxygen reduction was flavin or heme and the evidence for the involvement of a cytochrome was considered “at best circumstantial” (reproduced from Ref. [40])

The findings were felt to be suggestive of a binary model of Nox, consisting of a flavoprotein and a cytochrome b, although a present day look at the data finds these compatible with a single molecule, too. The binary model gradually became dominant, principally based on reports of separation of the flavoprotein from the cytochrome, with the accent placed on the characterization of the flavoprotein. Gabig described the complete separation of a flavoprotein from cytochrome by bile salt fractionation of a granular fraction from PMA-stimulated leukocytes and proposed that the flavoprotein represented Nox, based on its reduction by NADPH anaerobically and re-oxidation by oxygen, qualities lacking in the cytochrome fraction [41–43]. It is significant that none of the two fractions was capable of O2•production but the flavoprotein exhibited NADPH-dependent reduction of artificial electron acceptors under anaerobic conditions, an activity fitting the definition of a diaphorase [43]. These findings were a step forward compared to numerous failed attempts to demonstrate NADPH-dependent reduction of artificial electron acceptors by the particulate fraction of stimulated phagocytes under aerobic conditions and variable degrees of success under anaerobic conditions [44, 45]. The binary Nox proposal was supported by the description of two categories of CGD patients; one group lacking both flavoprotein and cytochrome and the second, lacking only the flavoprotein [41, 42]. The Discussion sections of these papers make for interesting reading today. The authors were wondering about the mechanism by which a single mutation could affect two proteins and were at a loss to explain the defect in the cytochrome-positive patient (most likely, a patient with a mutation in a cytosolic component and the apparent lack of flavoprotein being the result of a methodological error or dissociation of FAD from the cytochrome). Additional prestige to the centrality of a flavoprotein was brought by the careful biophysical studies coming from the laboratory of the grande dame of Japanese Nox research,

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

Katsuko Kakinuma (see Chap. 6 by H.G. Fujii and L.S. Yoshida). Using electron paramagnetic resonance (EPR), she found in both resting and stimulated leukocytes a signal typical of flavin semiquinone radicals [46]. The NADPH dependence of the EPR signal in stimulated cells was studied under anaerobic conditions and revealed that the FAD free radical was in the proximity of a transition metal and that the distance between the two components might decrease as a result of stimulation. This hypothesis contained the “prophetic” suggestion that the metal was likely to belong to cytochrome b558. In spite of this observation, Kakinuma did not espouse the binary Nox model and did not abandon the idea of direct production of O2•- by the flavoprotein, a belief strengthened by the claim to have isolated a 67 kDa protein, in which no heme was detected [47]. The refusal by such a prominent investigator to consider cytochrome b558 as a component of Nox, even as a part of binary enzyme, demonstrates the pressure of what Ludwik Fleck called the “thought collective”, expressed in a dominant “thought style” [48]. A publication, the straight interpretation of which supported the concept of the binary Nox, contained the grains of a paradigm change. Niels Borregaard, an authority on the fractionation and characterization of leukocyte subcellular particles, and Tauber, reported the close association of a b cytochrome and a flavoprotein in the specific granules of neutrophils and their correlated absence in the specific granules in cells of two patients with the X-linked form of CGD [49]. Preceding future studies, they also described the translocation to the plasma membrane of the two seemingly associated proteins upon stimulation of the intact cells with PMA, coinciding with the membrane localization of Nox activity. The publication also contained an element lacking in most other reports, namely the description of flavoproteins unrelated to Nox, present in the cytosol and plasma membranes, before stimulation. Re-reading this paper in 2022, one tends to hold one’s breath, in the expectation that the authors will abandon the binary enzyme model but this did not happen. Sticking to “normal science”, in Kuhn’s terminology, they supported the concept popular at the time but suggested a model very similar to the one accepted today, provided that one considered FAD and heme as belonging to a single molecule (Fig. 1.4). The model expressed the correct stoichiometry of one molecule of flavoprotein associated with two molecules of cytochrome and proposed the two-electron reduction of FAD by NADPH, followed by two one-electron transfer steps to O2, via the cytochrome, leading to the formation of two molecules of O2•-. A review by Parkinson and Gabig supported the binary Nox model but the signs of the approaching revolution were in the air [50]. The authors made two predictions, soon to be fulfilled; these were the design of a cell-free system of Nox activation and the isolation of cytochrome b558, holding the promise of solving the puzzle.

NADPH

9

FAD

FADH·

NADP+

FADH2

b-cytochrome (red)

O2

b-cytochrome (ox.)

O2-

b-cytochrome (red)

O2

b-cytochrome (ox.)

O2-

Fig. 1.4 Hypothetical model for Nox, representing an evolved version of that appearing in Fig. 1.3. The model is based on the important finding by Borregaard and Tauber of the colocalization of an FAD containing flavoprotein with cytochrome b558 in the specific granules of neutrophils and the absence of both in the specific granules of cytochrome b558-negative CGD patients. In the spirit of the binary concept, the association of two molecules of cytochrome with one molecule of flavoprotein is shown, reflecting the ratio found in the granules (reproduced from Ref. [49])

The funeral of the binary Nox model was heralded by a publication with the modest and uncommitting title “The association of FAD with the cytochrome b-245 of human neutrophils”, by Segal and Jones and coworkers [51]. This originated in the laboratories of those responsible for one of the most fundamental discoveries in Nox history, the identification of cytochrome b558, described by the authors, with British understatement, as “a likely component of the microbicidal oxidase system” [52]. Notwithstanding the fact that the new cytochrome was described in 1978, attention was focused on the flavoprotein and, in this respect, the decision to compare the amount of FAD to that of cytochrome b558 in the plasma membrane and specific granules of leukocytes was original and anteceded later work [41, 49]. The FAD: cytochrome ratio was found to be 1:1 and was maintained in the course of partial purification of the cytochrome. Membranes of neutrophils from patients with X-linked CGD, lacking cytochrome, had about half the amount of FAD, indicating that some FAD was not associated with cytochrome, whereas membranes of patients with the autosomal recessive form of CGD, not lacking cytochrome, had unaltered amounts of FAD. In the spirit of the binary Nox model, the authors placed the FAD protein in a “pathway between NADPH and a cytochrome b” but made the prophetic statement that “it is possible that FAD is normally attached to the cytochrome”. This prediction was voiced a decade before it became evident that both NADPH and FAD binding domains were in the cytosolic dehydrogenase region (DHR) of Nox2. The correct FAD:heme ratio of 0.5:1 in purified cytochrome b558 from neutrophils was reported in 1995 [53], confirming the early measurement in specific granules [49].

10

E. Pick

2.6

Off the Path, Briefly

In the early eighties of the last century, it was proposed that ubiquinone functioned as an additional redox station between FAD and cytochrome b558. The hypothesis originated in the observation that neutrophils contained ubiquinone [54, 55] and that ubiquinone-5 stimulated oxygen consumption by intact neutrophils and by disrupted cells, in the presence of NADPH or NADH [54]. The claim by Crawford and Schneider [54] that ubiquinone-5 stimulated oxygen consumption, associated with O2•- and H2O2 generation, was not supported by the actual data (low amounts and lack of controls). Additional backing for the hypothesis was derived from enrichment in ubiquinone content of phagolysosomes obtained from phagocytosing leukocytes, which paralleled the production of O2•- [56]. Based on complex assessment of redox poise in neutrophils, Gabig and Lefker suggested an electron flow sequence NADPH → FAD → ubiquinone → cytochrome, the activation of Nox involving “unblocking” the flow from ubiquinone to cytochrome [57]. The “ubiquinone hypothesis” was abandoned following reports that could not confirm the presence of ubiquinone in intact neutrophils and cytoplasts devoid of intracellular granules and in phagolysosomes enriched in Nox activity [58, 59]. The small amounts detected were associated with mitochondria. The rise and fall of the ubiquinone hypothesis should serve as a warning against not including appropriate controls and over-interpreting of results. Here and there, an iron-sulfur protein was proposed to serve as yet another redox station; however, a search for it in human neutrophil membranes by EPR failed to detect any such compound [60].

2.7

The Campaign to Purify the Pre-activated NADPH Oxidase

This implies that the components of the activated NADPH oxidase complex are loosely associated and that isolation of the enzyme is likely to continue to be an exercise in tireless perseverance and saintly patience John F. Parkinson and Theodore G. Gabig [50]

The predominant procedure used in attempts to isolate and purify Nox was to first activate Nox in intact cells by particles to be phagocytosed (frequently, opsonized) or by soluble stimulants (mostly, PMA), followed by separating a particulate fraction, solubilization of the particles, subjecting the soluble material to fractionation, and assessing Nox activity. These efforts overlapped the studies meant to prove that Nox was a flavoprotein and the identification of cytochrome b558 as a component of Nox. The “Nox purification” period covered the decade 1978–1988; the last years coincided with designing methods to purify cytochrome b558.

The first successful solubilization of the enzyme of phagocytosing human neutrophils was published in 1978 [36] and was followed by the description of the basic characteristics of the soluble enzyme [37, 38] but there was no attempt to purify the enzyme except for establishing that it passed through a membrane filter retaining molecules larger than 300 kDa [36]. Tauber and Goetzl solubilized phagocytosing or PMA-stimulated human leukocytes, using the ionic detergent sodium deoxycholate, and performed the first size-based fractionation by gel filtration on Sephacryl S-200 [61]. Two peaks were recovered, corresponding to the exclusion volume and to a molecular mass of 150 kDa; material of both peaks possessed lege artis NADPHdependent, azide-resistant Nox activity. Nox partitioned into large structures whenever gel filtration was performed with buffer in the absence of detergent, due to the association of cytochrome b558 with membrane phospholipids forming liposomes [62]. It is remarkable that the paper did not refer to a possible relationship of Nox activity to cytochrome b558 although the description of the cytochrome as a component of Nox was published 1 year earlier [52]. The “Verona School” of Rossi continued its tradition of excellence and was deeply involved in trials to isolate an active Nox. Using the particulate fraction of PMA-stimulated guinea pig leukocytes solubilized by sodium deoxycholate, an active Nox was recovered as a high molecular weight pellet, clearly separated from material with diaphorase activity, which remained in the supernatant following centrifugation at 100,000 × g [63]. The Nox but not the diaphorase activity was associated with the presence of cytochrome b558, as shown by a typical reduced minus oxidized spectrum, a midpoint potential of -245 mV, and a FAD:cytochrome ratio of close to 0.5:1 [63]. The authors were concerned by the slow rate of cytochrome reduction by NADPH under anaerobic conditions, an issue that kept the Nox community busy for quite some time. By subjecting solubilized PMA-stimulated pig neutrophil subcellular particles to size exclusion chromatography, Nox activity was recovered in a high molecular mass complex containing cytochrome b558 and phospholipids, distinct from a smaller molecular mass fraction that contained a NADPH/ NADH-dependent diaphorase [64]. A common feature of reports from this period was an unusually low FAD:cytochrome ratio, due to loss of non-covalently bound FAD during lengthy procedures in the presence of detergent. Nox derived from the particulate fraction of PMA-stimulated guinea pig leukocytes or macrophages was associated with a high molecular mass complex containing cytochrome b558 and phospholipids, exhibiting an unlikely sized 32 kDa band on SDS-polyacrylamide gel electrophoresis (SDS-PAGE) [65, 66]. If proof was needed for the difficulty with which a new paradigm conquers the field, the work of another pioneer of

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

Nox research, Andrew Cross, then, active in Bristol with Jones, on Nox derived from PMA-stimulated pig neutrophils is a perfect example [67]. Membranes solubilized by Lubrol PX and sodium deoxycholate exhibited stimulationdependent Nox activity and contained a flavoprotein and cytochrome b558. NADPH partially and slowly reduced both FAD and cytochrome under anaerobic conditions. The thiol-alkylating agent. p-chloromercuribenzoate (PCMB) inhibited activity and cytochrome reduction by NADPH, suggestive of the presence of an essential cysteine in the DHR of Nox2. The authors adhered to the binary Nox interpretation of these findings and proposed the existence of a FAD-protein distinct from cytochrome b558, participating in the electron transport path scheme popular at the time: NADPH → FAD protein → cytochrome b558 → O2 → O2 • The group of Minakami solubilized Nox from pig leukocytes phagocytosing oil droplets, exhibiting the characteristic properties (cyanide resistance, NADPH preference, enhancement by FAD) and the presence of some cytochrome b558 [68]. In an original approach, detection of the “Nox protein” rested on the certainty that it must contain the NADPH binding site and 2′3′-dialdehyde NADPH was used as an affinity label [69]. A molecule of 66 kDa was the dominant protein labeled and an excess of NADPH or pretreatment with PCMB prevented labeling. FAD and heme could not be detected due to technical reasons and, although the authors paid lip service to the binary Nox model, it is likely that this was one of the great missed occasions to identify Nox2 as the catalytic component of Nox (pig Nox2 has a theoretical molecular mass of 55.83 kDa). Based on the binary model, Koyama and coworkers first purified an NADPH-dependent cytochrome c reductase from solubilized porcine leukocyte membranes [70] and combined the reductase with cytochrome b558, purified from the same membranes. The authors concluded, based on NADPH oxidation and O2•- production, that the results indicated Nox activation but an analysis of the reported turnover and other quantitative data did not uphold this claim [71].

2.8

The Cytochrome Deniers

One Step Forward, Two Steps Back – The Crisis in Our Party Book by Vladimir Ilyich Ulyanov (Lenin) (1904)

The finding that cytochrome b558 was the paramount component of Nox was published in 1978 [52]. In the following decade, investigators were busy purifying and identifying the activated Nox from stimulated phagocytes and although all were aware of the existence of cytochrome b558, its centrality

11

was doubted or denied. The reasons for this anomaly were: (a) Wrong identification of the component responsible for Nox activity expressed in the claim that it did not contain heme; (b) The “historic” dominance of the idea that Nox was a flavoprotein and no more; (c) Difficulty in accepting the fact that both FAD and heme were present in the same molecule and the reporting of widely divergent FAD:heme ratios; (d) The slow and incomplete reduction of cytochrome b558 by NADPH under anaerobic conditions, which appeared to be incompatible with rates of O2•- generation; (e) The fact that purified cytochrome b558 from resting cells could not be shown to express Nox activity before the advent of the “cell free system”, and (f) The fact that cytochrome b558, derived from stimulated phagocytes, lost its expected Nox activity in the course of purification. The group of Pierre Vignais solubilized the particulate fraction of PMA-stimulated bovine leukocytes and purified Nox by a sequence of anion exchange, size exclusion chromatography and isoelectric focusing, all of which were performed in the presence of the detergent used for solubilization [72] (see Chap. 7 by M-C. Dagher). The purified enzyme had a molecular mass of 65 kDa by SDS-PAGE but only traces of FAD and cytochrome b558 were found to be associated with the pure protein. The material produced O2•-, oxidized NADPH, was resistant to cyanide, and phospholipid enhanced its activity. The authors stated in the Discussion section that “cytochrome b has been claimed to be a component of the O2-—generating oxidase”, referring to Ref. [52], which makes for strange reading today. They concluded that the absence of cytochrome in the purified material indicated that it was not directly involved in the generation of O2•- but might exert some control on the course of the respiratory burst. The absence of FAD was explained by the technical difficulty to release the tightly bound flavin from the protein (an argument negated by our present day knowledge of the non-covalent attachment of FAD to Nox2). In a corollary paper, further arguments were harnessed against cytochrome b558 being a component of the oxidase [73]. Among these were the frequently brought up difficulty to reduce the cytochrome by NADPH and, strangely, the lack of effect of CO on oxygen uptake (an issue that proved to be a hard nut to crack). It was even more unexpected that one of the founding fathers of Nox research and the person who first described O2•- generation by phagocytes did join the anti-cytochrome “club”. The group of Babior purified Nox from solubilized membranes of zymosan-stimulated human leukocytes by affinity chromatography on red agarose (red stands for a triazine-based azo dye, known as reactive red 120, which was bound to agarose). The fraction with Nox activity corresponded to a protein with a molecular mass of 65 kDa by SDS-PAGE, which contained FAD [74]. The active fraction lacked spectral evidence for cytochrome b558, a fact

12

attributed by the authors to the low amount of protein, insufficient for detection of the cytochrome. Using a modified solubilization procedure, an active Nox was purified which contained FAD but only traces of cytochrome b558 and appeared as a single band on native gel electrophoresis [75]. SDS-PAGE revealed the presence of three proteins of molecular masses of 67, 48, and 32 kDa and the authors proposed that the active oxidase was a complex of 150 kDa consisting of the three subunits. The absence of a detectable cytochrome was subject to two alternative interpretations: that the enzyme was a flavoprotein, capable of reducing oxygen, or that a downstream component was lost and the flavoprotein was altered in the course of purification, acquiring an oxygen reducing ability. Intrigued by the marked difference between the results of Babior’s group and their own [63–66], the Verona group did what was rarely done and carefully repeated the experimental procedure, as performed by Babior. In a brief report, they concluded that Nox purified on red agarose did contain cytochrome b558 and was (paradoxically) poor in FAD [76]. To crown this disturbing period in the history of Nox research, Green and Pratt solubilized Nox from the particulate fraction of PMA-stimulated human leukocytes and separated a cytochrome b558-containing fraction from a flavoprotein, 51 kDa in size, by gel filtration [77]. None of the two fractions displayed Nox activity, but some was detected in the exclusion volume, in which minor amounts of the cytochrome and flavoprotein co-eluted. This work represented an unexpected return to the binary Nox concept, ignoring the identification of cytochrome b558 in 1978 and the ensuing research activity, and was an example of “two steps back”, appearing in the title of Lenin’s political pamphlet. A remarkable fact about the follow-up publication of Babior [75] was that it ignored the discovery of the cytosolic participation in Nox activation, which became evident in 1984/1985. In an admittedly retrospective judgement, it is surprising that whenever an active Nox was isolated, it referred to a single molecular species. By present day concepts, an active phagocyte Nox means a complex of cytochrome b558 assembled with the cytosolic components and one would expect that, at least, some of the complexes would withstand the purification procedure and the associated cytosolic components would be detected. Consequently, one wonders what the identities of the 67 kDa, 48 kDa, and 32 kDa bands, described by Babior, were (p47phox?, p67phox?). To the credit of Green and Pratt is that their publication was the first to mention the possibility that the lack of activity of the two purified components was caused by the absence of a third component, which they called “cytosolic factor”, thought to harbor the NADPH binding site [77].

E. Pick

2.9

Before “Curtain Down”

An interesting “left over” from the numerous efforts to purify Nox was the marked inhibitory effect of PCMB [57, 61, 63, 67, 69]. It might be easy to discard these data on the basis that most work with PCMB was done when the amino acid sequence of Nox2 (the certain target of PCMB) was unknown and the specificity of PCMB for cysteines is uncertain (possible binding to methionine and histidine). However, there are strong indications that PCMB binds to a specific cysteine in the NADPH-binding site of Nox2, a property that explains its inhibitory action on Nox activity. That PCMB acted on cytochrome b558 was first shown by its ability to elevate the midpoint potential from -245 mV to -175 mV in solubilized guinea pig leukocyte Nox [63]. O2•- production by solubilized pig leukocyte Nox was also inhibited by PCMB, targeting on cytochrome b558 being supported by the parallel inhibition of its reduction by NADPH under anaerobic conditions [67]. In a quite extraordinary prediction, Gabig and Lefker explained the mechanism of PCMB inhibition by it binding to a single sulfhydryl (cysteine) in the proximity of the NADPH binding site [57]. Finally, Minakami’s group offered direct experimental evidence for such a mechanism by showing that PCMB prevented the labeling of a 66 kDa purified Nox with the NADPH analog, 2′3′-dialdehyde NADPH [69]. Based on present day knowledge, it is likely that PCMB indeed acts by binding to cysteine 537 in the nicotinamide binding sub-site of the NADPH binding domain of Nox2. A Cys537Arg mutation in Nox2 caused X91+ CGD, due to the failure of NADPH binding [78, 79]. Recent work revealed that the Nox inhibitors VAS2870 and VAS3947, acted by covalent alkylation of Cys668 in the DHR of bacterial Nox5, located in the nicotinamide binding site and representing a cysteine conserved among all Noxs (corresponding to Cys537 in Nox2) [80].

3

“Modern Times”: The Transition to Molecular NADPH Oxidase Research

(inspired by the title of the film by Charlie Chaplin)

3.1

Cytochrome b558 Phylogeny: The Pink Oxidase

The first description of cytochrome b558 appeared in an article by Hattori, the original purpose of which was to study what was called “Nadi oxidase” in the granular fraction of horse leukocytes [81]. The Nadi reaction, by which the Nadi oxidase was defined, consists of the formation of indophenol blue from dimethyl-p-phenylenediamine and α-naphtol, and

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

was originally described by no lesser person than Paul Ehrlich, while still a medical student [82]. Hattori found that what he defined as “stable” Nadi oxidase reaction appeared exclusively in the leukocyte granular fraction, which was distinct from mitochondria. This fraction contained a molecule with a unique absorption peak at 558 nm, when in the reduced state, defining it as a b cytochrome. Quite amazing was the observation that the granular fraction exhibited a pinkish color. The author concluded that the function of the newly discovered cytochrome “may be connected with a certain important role in the metabolism of neutrophil leukocytes”. Five years later, Shinagawa and coworkers made the unexpected connection between the work of Karnovsky on the enhanced oxygen uptake of phagocytosing leukocytes [5] and the description by Hattori of a cytochrome b with an absorption peak at 558 nm, when in the reduced form. A granular fraction was prepared from rabbit leukocytes, free of mitochondria, and the dithioniteinduced reduced minus oxidized spectrum was recorded. This was a typical cytochrome b558 spectrum, with absorption peaks at 558 nm, 525 nm, and 426 nm, and a trough at 410 nm [83]. The cytochrome was reduced slowly by NADPH and NADH (?) under anaerobic conditions and reduction was cyanide-resistant. Both reports were ignored for almost two decades.

3.2

The Cytochrome b Revolution

The modern era of Nox research had its origins in the close cooperation of a clinician-scientist, Anthony W. Segal, and a biochemist with a background in mitochondria and heme, Owen T. G. Jones (see Chap. 4 by A.W. Segal). Chronologically, the first publication mentioning a newly identified component of Nox described the presence of a b type cytochrome in homogenates of human leukocytes with a reduced minus oxidized spectrum exhibiting peaks at 559 nm and 429 nm (corresponding to the 558 nm and 426 nm peaks found by Hattori and Shinagawa in horse and rabbit leukocytes) [84]. The key discovery was, however, the absence of both 559 nm and 429 nm peaks in leukocytes of four CGD patients; the observation was made at a time when the modes of inheritance were not yet clearly established. The article contained a number of historically valuable curiosities, such as the reference in the introductory section to the author’s own work on the failure of NADH (!)-dependent NBT reduction by leukocytes of CGD patients and, although it was proposed that the cytochrome carries electrons to oxygen and may pump protons into the phagocytic vacuole, there was no suggestion of the enzyme generating O2•-. The major publication that initiated the cytochrome b revolution appeared in Nature on November 30, 1978 [52]. Human leukocytes phagocytosing opsonized zymosan

13

were submitted to spectroscopy. Activation of Nox by phagocytosis resulted in the appearance of a broad absorption peak at 416 nm, of unknown origin, and a minor peak at 560 nm. Exposure of leukocytes to PMA also resulted in spectroscopic changes. The ability to detect the 560 (558?) nm peak in whole cells and to “catch” the cytochrome in the reduced form was a technical tour de force. The authors next showed that whole cell homogenates and purified phagocytic vacuoles exhibited a reduced minus oxidized spectrum with sharp peaks at 429 nm and 560 nm. The cytochrome was missing in the cells of six CGD patients with the X-linked form of the disease.

3.3

Challenging the Cytochrome Revolution

The finding by Segal and Jones of the absence of cytochrome b in the majority of CGD patients led other investigators to look for the status of cytochrome b in CGD patients in their care. In a brief case report, Borregaard (known for significant work on the subcellular location of Nox components [17, 49]) and coworkers described two CGD patients, brother and sister, with the autosomal recessive mode of inheritance, expressing normal cytochrome b content in their leukocytes [85]. Encouraged by the finding that leukocytes of a patient with the X-linked form of disease also contained cytochrome b (probably, an X91+ form of CGD), the investigators reached the sweeping conclusion that “this finding invalidates the argument that cytochrome b is part of an oxygen-consuming electron transport system”. In an exceptionally good-tempered response, Segal and Jones expressed their satisfaction at the confirmation of the detection of a cytochrome b in leukocytes by Borregaard and reported that retesting of the leukocytes of CGD patients found to lack cytochrome b, as described before [52], yielded identical results [86]. The definitive response to the CGD—cytochrome b dilemma appeared in a seminal paper, showing that, functionally, CGD could be divided in two categories, characterized by either the absence of cytochrome b or by the inability of cytochrome b to be reduced following activation of Nox [87]. Dithionite-generated reduced minus oxidized spectra of intact leukocytes showed the presence of cytochrome b in normal subjects, its absence in two male CGD patients (X-linked) and its presence in two female CGD patients. However, typical cytochrome b spectra could also be seen following stimulation of leukocytes of normal subjects by PMA under anaerobic conditions but were absent in leukocytes of both male (X-linked) and female CGD patients. The failure of male CGD patients’ cells to exhibit a cytochrome b spectrum was due to the absence of cytochrome or the presence of an “abnormal” cytochrome. The failure of female CGD patients’ cells to exhibit a cytochrome b spectrum was proposed to be due to either the absence of a

14

component proximal to cytochrome b (a left over from the binary Nox model) or to “a defect in the mechanisms which activate or coordinate the system”. This second remarkable prediction was found to be true following the design of the cell-free system, which led to the discovery of the cytosolic components and the finding that most cases of the autosomal recessive mode of CGD were due to the absence of the cytosolic components p47phox or p67phox [88]. The use of the terms “activate” and “coordinate” represented a remarkable prediction since a nomenclature established many years later coined an analogue of p67phox as Nox “activator” (NOXA1) and an analogue of p47phox, as Nox “organizer” (NOXO1) (reviewed in [89]). The acceptance of cytochrome b558 was followed by the determination of its midpoint potential as –245 mV, a value sufficiently low to permit direct reduction of oxygen to O2•[60]. It took almost 15 years until it was realized that this represented the mean value of the midpoint potentials of two hemes, of -225 mV and -265 mV [90]. Finally, the issue of the ability of cytochrome b558 to be reduced by NADPH was revisited. The slow and incomplete reduction of cytochrome, originating from PMA-stimulated leukocytes under anaerobic conditions, meant to “freeze” it in the reduced state, was incompatible with the rate of O2•production [67, 90, 91]. This was interpreted as suggesting that not all electrons from NADPH to oxygen pass through cytochrome and was used as an argument in a future controversy about whether cytochrome b558 harbors all the redox centers (see Sect. 5.1). In a surprising turnabout, the brilliant experimental skills of Cross provided the answer to the dilemma. Human leukocytes were stimulated by PMA and the particulate fraction solubilized. NADPH was added in the presence of oxygen and steady state spectra were recorded. It was found that the rate of NADPH-dependent cytochrome reduction by extracts of stimulated cells was equal to the rate of NADPH-dependent O2•- production. The reduction was resistant to cyanide, was specific for NADPH, and inhibited by PCMB, demonstrating that oxygen was required for rapid electron flow within the cytochrome [92]. A decade later, Vasilij Koshkin made use of a system of cytochrome b558 activation by reconstitution with anionic phospholipids, which he developed (see Sect. 6.8), and suggested that the reason for the low heme reduction under anaerobic conditions was the nonidentity of the two hemes [90], manifested in only one being easily accessible to reduction by FAD. Under aerobic conditions, there was a marked facilitation of electron transfer, dependent in some way on interaction with oxygen, which was reflected in efficient heme reduction [93]. An unexpected detail in this paper was the prediction that the two hemes were at different locations in the cytochrome molecule, manifesting a difference in their reducibility under anaerobic conditions, made before evidence for this became available.

E. Pick

3.4

“The Gold Rush”: The Race to Purify the Cytochrome

(inspired by the title of the film by Charlie Chaplin) A natural outcome of the almost certainty that cytochrome b558 was a central and, most likely, the only catalytically active component of Nox was the effort by the “who is who” of Nox research to purify the cytochrome. Segal’s group were the first to publish the purification of cytochrome b558 from leukocytes of chronic myeloid leukemia patients [94]. Critical steps were removal of a flavoprotein by the use of specific detergents and affinity chromatography on heparin-agarose. In addition to its priority value, this work was an example of how a major methodological advance can tolerate certain errors. Cytochrome b558 was described as a single protein (wrong), seen as a broad band of 68–78 kDa on SDS-PAGE, suggesting that is was a glycoprotein (right), and having low affinity for CO (controversial). Nox activity was absent in the extract used for purification, derived from PMA-stimulated leukocytes, this being explained by the removal of the flavoprotein, thought to act as the electron donor to the cytochrome, in accordance to the still prevalent binary Nox concept. About 3 years later, Segal published a paper describing a modified purification procedure that led to the finding that the protein of 76–92 kDa, described before, was associated throughout the purification procedure with a second protein, 23 kDa in size [95]. The two proteins were present in equimolar amounts. Both subunits were absent in leukocytes of patients with the X-linked form of CGD. Still a captive of the binary Nox concept, Segal considered the possibility that one of the subunits was the cytochrome and the other, the flavoprotein, with preference for the large subunit representing the heme-binding apoprotein. An interesting point raised in this paper was the expression of uncertainty about the relation of the larger subunit to the protein encoded by the gene for Nox2 on the X chromosome, cloned a short time before (see Sect. 4.1), as apparent in dissimilarity in amino acid composition. It should, however, be noted that Segal performed the amino acid analysis on a partially purified preparation of whole cytochrome that might have contained both subunits [96]. On the other side of the Atlantic, another group was involved in an independent effort of purifying cytochrome b558. The affinity of the leading investigator, Algirdas Jesaitis, yet another major contributor to Nox research, for cytochromes had its origin in work done a decade earlier on corn coleoptiles, leading to the characterization of a cyanideinsensitive b cytochrome (personal communication [97]). By starting from human leukocyte membranes solubilized by octyl glucoside (a nonionic detergent that became very popular in cytochrome b558 studies) and using a series of affinity chromatography steps, a protein was isolated revealing two

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

polypeptides banding at 91 kDa and 22 kDa by SDS-PAGE [98]. The 91 kDa protein was seen as identical to the 68 kDa to 78 kDa protein described by Segal; the 22 kDa component was seen as identical to the 23 kDa protein described by Segal. The larger protein was a glycoprotein (the protein core was 50 kDa in size) and the two components formed a heterodimer [99]. The authors also described the absence of the 22 kDa component in cells of patients with the X-linked form of CGD. It is of historical interest that the 22 kDa protein was considered as the likely candidate for carrying the heme, based on the claim that the protein encoded by the gene for Nox2 on the X chromosome did not show a heme binding motif (see Sect. 4.1). The prominence of the work of Segal and Jesaitis, its impressive confirmation by the cloning of the genes, and the clinical significance reflected in understanding the pathogenesis of CGD, overshadowed the considerable amount of work done by other groups. The group of David Lambeth purified a cytochrome from bovine leukocytes with a typical spectrum and close to expected midpoint potential but yielding three bands of 11, 12, and 14 kDa on SDS-PAGE [100]. The group of Dirk Roos purified a cytochrome from human leukocytes with a typical spectrum and a molecular mass of 127 kDa [101], most likely identical to the large subunit described by Segal [94] and Jesaitis [98]. Unexpectedly, the authors argued against a role of the cytochrome in Nox activity on the basis that CO was bound to the cytochrome (with very slow kinetics) and did not inhibit Nox activity in intact leukocytes (two findings to be interpreted differently in the future). The group of Rossi purified a cytochrome from pig leukocytes yielding a protein of 31 kDa although a shadow of a protein in the 100 kDa range was also visible [102]. Finally, Vignais and Morel isolated a cytochrome from bovine leukocytes, revealing three bands of 64, 56, and 20 kDa on SDS-PAGE [103], quite different from the results of Lambeth [100].

3.5

Cytochrome at Work: O2•- Production by Purified Cytochrome b558

It is surprising that the numerous attempts to purify cytochrome b558 were not associated with testing the purified material for Nox activity. The logic inherited from early work on isolating an active Nox would have been to purify cytochrome from cells stimulated with an activator of the respiratory burst. Not attempting to do this was due to the still dominant binary Nox model, in which cytochrome required an upstream electron donor for activity, and to the belief that the cytochrome had a low chance of remaining in an activated state throughout the purification process. To the best of my knowledge, the first demonstration of a catalytically active cytochrome b558 was the ability of a

15

highly purified cytochrome from resting guinea pig macrophages to produce O2•- in an amphiphile-activated cell-free system (see Sect. 6.2) in the presence of purified cytosolic components [104]. The ability of purified cytochrome b558 to produce O2•- in vitro was confirmed with cytochrome derived from membranes of resting human leukocytes in the presence of cytosol [53] or recombinant cytosolic proteins [105, 106]. It was, therefore, surprising that the group of Kakinuma was unable to activate cytochrome b558, purified from pig leukocytes in a cell-free system containing total cytosol and myristic acid, as the activating amphiphile [107]. Possibly the most eloquent example of cytochrome b558 engaged in “do it by yourself” was the description by Koshkin of the production of O2•- by purified cytochrome b558, under specific conditions of relipidation, in the absence of cytosolic components [108, 109]. The closing sentence of the first paper [108] serves as a most adequate summary for this subsection: “. . .the present report offers direct experimental proof for the proposal that the complete electron transporting machinery of the O2•- generating NADPH oxidase is contained in the cytochrome b558 dimer”.

3.6

No “Rapping” in Cytochrome b558

The rise and fall of Rap1A, as a component proposed to be associated with cytochrome b558, serves as an example for a pseudo-paradigm shift. The involvement of a GTP-dependent regulatory protein (G protein) in Nox activation was first proposed by Seifert [110] and by Gabig [111] on the basis of enhancement of amphiphile-stimulated O2•production in the cell-free system by nonhydrolysable GTP analogs and its inhibition by nonhydrolysable GDP analogues. The resistance of the process to pertussis and cholera toxins indicated that the G protein involved was not a member of the heterotrimeric family. Mark Quinn and coworkers were the first to describe the association of the Ras-like small GTPase, Rap1A, with cytochrome b558 and suggested that it may represent the sought after Nox activity regulatory G protein [112]. A subsequent study by the same group described regulation of the cytochrome—Rap1A complex formation by phosphorylation of Rap1A [113]. Doubts about the functional significance of the claimed association of Rap1A with cytochrome b558 became dominant when it was found that patients with the X-linked mode of CGD, lacking cytochrome b558, had normal expression of Rap1A [114]. Two studies by the group of Gabig offered apparent support for the involvement of Rap1A in Nox activation but the methodological complexity of the work made their interpretation difficult [115, 116]. In an in vivo study, it was found that leukocytes of Rap1A-/- mice had reduced stimulated O2•- production but also changes in adherence and

16

E. Pick

chemotaxis, suggesting that the site of action of Rap1A was upstream from Nox [117]. The definitive downfall of the Rap1A hypothesis was the result of the finding that in highly purified cytochrome b558 preparations, Rap1A was no longer detectable but the material fully supported Nox activity in a cell-free system [105, 106, 118]. The Rap1A saga was finally sealed by the finding that the G protein having a key role in oxidase function was Rac1 or 2 (see Sect. 8).

3.7

Filling in the Blanks: Heme Coordination: Six or Five?

Even before the purification and characterization of cytochrome b558 was completed, the structure of the hemes was an issue that was actively investigated. There was general agreement, going far back in the prehistory of Nox research, about the cyanide resistance of Nox but the binding and effect of CO were still controversial. Readers who are familiar with Hitchcock’s masterpiece, the film “Shadow of a Doubt”, will recall the attempt of the criminal uncle to poison his niece by locking her in a closed space gradually filling with carbon monoxide (CO). Poisoning by CO is caused by the ability of CO to compete with oxygen for binding to the heme of hemoglobin, characterized by a pentacoordinated iron, with the sixth coordination site binding oxygen. Pentacoordinated heme iron is also found in other enzymes involved in interaction with oxygen, such as cytochrome c oxidase and cytochrome P450, allowing the formation of iron-oxygen complexes (reviewed in [119]). The elucidation of the structure of cytochrome b558 heme and its refractoriness to CO and cyanide represented a paradigm shift, which was mainly the product of research by Japanese investigators. In a fundamental study, Iizuka and coworkers assessed the absorption spectra of cytochrome b558 at low temperature in the presence and absence of CO and found these to be undistinguishable, leading to the conclusion that the cytochrome does not form a complex with CO [120]. They also confirmed previous findings that CO did not inhibit O2•- production by stimulated leukocytes, as expected from its inability to compete with oxygen. A low spin hexacoordinated heme iron, with bis-imidazole axial ligation being characteristic of cytochrome b558, was first proposed by Hurst and coworkers [121], and supported by an independent study, which also predicted axial coordination of the iron to two histidines [122]. A curiosity, in accordance with the state of knowledge at the time, was the proposal that one of the histidines was located in Nox2 and the other, in p22phox [122]. Using EPR on purified material, further confirmation of the low spin, hexacoordinated, bishistidine-linked character of cytochrome b558 was provided by the groups of Kakinuma and of Segal, in accordance with

its resistance to cyanide and its lack of CO binding [123, 124]. The final touch to the paradigm shift was establishing a strict correlation between the Nox activity of purified cytochrome b558 in a cell-free system and the lowspin hexacoordinated state of the heme [125]. Artificial induction of a high-spin state of the heme led to a reduction of cell-free Nox activity concomitant with the loss of low-spin heme. An attempt to assign a role to the high-spin state was based on the finding that anionic amphiphiles used as Nox activators in the cell-free system (arachidonic acid (AA) and SDS) were found to cause a transition of the cytochrome b558 heme iron from the low-spin hexacoordinated state to a high-spin pentacoordinated state [126]. The authors postulated that Nox activators worked by causing the pentacoordinated form of the reduced heme iron to bind oxygen and generate O2•-. A subsequent study by Fujii and coworkers could not confirm the existence of a transient activation-related pentacoordinated state by showing that although AA induced a high-spin heme iron, myristic acid did not in spite of the fact that both amphiphiles were capable of Nox activation [127].

3.8

Platonic Love: Cytochrome b558 Heme Iron and Oxygen Keep Distance

The hexacoordination of the heme iron in cytochrome b558 posed a challenge to unveiling the mechanism of single electron transfer from the reduced iron to oxygen. Isogai and colleagues delivered the death blow to the pentacoordination of heme iron in cytochrome b558 hypothesis. In an inventive methodological approach, the authors reverted to a variation of the binary Nox model by activating purified cytochrome b558 with NADPH cytochrome P-450 reductase [128]. Using a range of sophisticated biophysical methods, including EPR, the authors presented evidence for a low-spin oxidized heme, with g values similar to those reported by other investigators [123, 124]. The authors also confirmed the lack of effect of CO, cyanide and azide on the absorption spectrum and Nox activity of cytochrome b558 and noted the extremely rapid reoxidation of reduced heme iron, with no evidence for the presence of intermediate species. What was unique in this work was the proposal that an electron was directly transferred from the reduced heme iron to oxygen, without ligation of oxygen to the iron via a sixth coordination site [128]. Further work, involving measurement of oxidation-reduction kinetics of cytochrome b558 by rapid scanning spectroscopy, strongly supported the absence of a heme iron—oxygen intermediate. Instead, the one-electron reduction of oxygen occurred by an “outer sphere” mechanism at or near the heme edge [129].

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

17

This proposal represented a major paradigm shift and it took more than two decades for proof for its veracity, based on structural studies, to become available. This was the result of solving the crystal structure of Nox5 from Cylindrospermum stagnale (csNox5) [130]. Two hemes were detected; one, close to the cytosolic side of the membrane and the second, toward the outer side. Both were found to be hexacoordinated. The structure revealed a small cavity located above the outer heme which was occupied by a water molecule, to be replaced by the oxygen substrate. The cavity was lined by two basic residues, involved in oxygen binding and catalysis. In a surprising prediction, Isogai suggested, back in 1995, that the weak interaction of oxygen with heme in Nox2 was stabilized by hydrogen bonding, with a basic amino acid near the heme pocket [129]. The significance of a basic residue “in the right place” was also shown by the serendipitous discovery of an X-linked CGD patient, with normal cytochrome b558 content but no O2•- production, due to an Arg54Ser mutation, which replaced a positively charged residue with an uncharged one [90]. The structural data of csNox5 offered direct proof for dioxygen binding not occurring through direct coordination to the iron of the heme but rather by noncovalent interaction with the prosthetic group and surrounding hydrophilic side chains [130]. With a high degree of certainty, we can assume that this model applies to Nox2 and to all the other Noxs.

acid composition, no resemblance to other cytochromes, and no evidence for a heme binding region. In an article published 3 years after the original paper, the correct open reading frame of 571 amino acids was described, corresponding to a 65 kDa protein [134]. In a most candid statement, Orkin admitted that “On the basis of the computer searches it could only be speculated initially that the X-CGD gene encoded a membrane glycoprotein that might be involved in the assembly of the cytochrome b but was unlikely to be the cytochrome itself”. Segal and coworkers proved that the gene was indeed coding for the large subunit of cytochrome b558 by separating the large and small subunits and sequencing 43 amino acids from the N terminus of the large subunit. These corresponded to nucleotides 19–147 in the X-CGD gene, originally thought to be noncoding; adding 101 amino acids to the translation product led to the expected molecular mass of 65 kDa [135]. Mary Dinauer and coworkers provided additional proof by using immunoblotting with antibodies to a peptide corresponding to 20 residues from the complementary X-CGD gene DNA sequence, and to recombinant proteins, corresponding to the entire or part of the X-CGD gene DNA sequence. The antibodies detected a protein of 90 kDa in leukocytes, absent in cells of X-linked CGD patients, and also recognized the large glycosylated subunit of purified cytochrome b558 [136].

4

Cytochrome b558: A Roundup

4.2

4.1

“Clone or Perish”: Reverse Genetics in Action

The cloning of the gene for the small subunit of cytochrome b558 involved isolation of cDNA clones by immunoscreening an expression vector library, prepared from differentiated promyelocytic leukemia HL60 cells, with a polyclonal antiserum to the small subunit [137]. The deduced amino acid sequence showed no similarity to known cytochromes, with the exception of some similarity to a histidine-bearing region of cytochrome c oxidase and one histidine aligned with the iron coordinating histidine of myoglobin. At the time, the authors were unable to decide whether the hemes were present on the large or small subunits or shared by the two. It is of interest that a high proline content was detected in the deduced sequence, a finding the relevance of which was revealed only when the role of p22phox in Nox assembly was discovered (see Sect. 11.1). This publication was also the first to describe the fact that RNA encoding p22phox was present in all cells but the protein was only expressed in phagocytes that also expressed the large subunit. Cloning of the p22phox gene, mapping the chromosomal location and characterizing the exon-intron structure led to solving the genetic basis of the autosomal cytochrome negative form of CGD, resulting from defects in the gene encoding p22phox [138]. As far as I was able to find out, this was the first

The customary approach to the molecular analysis of inherited disorders was the identification of a specific protein related to the disease followed by that of the corresponding gene. In spite of the fact that the protein missing or defective in the X-linked form of CGD was known [52, 84, 86] and purified [94, 98], the identification of the gene was based on a different approach, known as “reverse genetics” (reviewed in [131]). The first step in this approach was to map the position of the gene related to the X-linked form of CGD and, next, to focus on a segment in which mutations were strictly correlated with the disease [132]. In a major breakthrough, Stuart Orkin and coworkers cloned the gene for the X-linked mode of CGD by applying successfully the “reverse genetics” methodology [133] (see Chap. 5 by S.H. Orkin). The complete X-CGD cDNA was found to be 4.27 kb in length and contained a single open reading frame encoding 468 amino acids, with an initiation codon at base 322 [133]. The conclusions make for fascinating reading today, with the authors arguing against the predicted protein being related to cytochrome b558, based on different amino

“Forget-Me-Not”: The p22phox Gene

18

E. Pick

publication in which a novel nomenclature for the two subunits of cytochrome b558, previously called “large, β, or heavy chain” and “small, α, or light chain”, was introduced. The large subunit was coined gp91phox (based on its glycoprotein nature and the molecular mass of 91 kDa of the human protein), predominantly known as Nox2, at present, and the small subunit, p22phox (based on the molecular mass of 22 kDa of the human protein) [138]. The authors made the important observation of the absence of both subunits in the autosomal CGD due to a defective p22phox, demonstrating that expression of the mature form of gp91phox required association with p22phox. This complemented the earlier reciprocal observation that both gp91phox and p22phox were absent in cells of patients with the X-linked mode of CGD, due to defects in the gene coding for gp91phox, showing that the stability of p22phox required association with gp91phox [139]. Distinct studies focused on the role of heterodimer formation in the synthesis, processing and maturation of cytochrome b558 [140, 141]; discussing these is beyond the scope of this chapter. Three years after cloning the gene for p22phox and noticing the richness of the deduced protein in prolines, the same authors described a CGD patient with a Pro → Gln mutation at p22phox residue 156 [142]. Cytochrome b558 was present but not functional and the authors predicted (correctly) that the mutated Pro residue was in a part of p22phox located in the cytosol. It took 3 more years for an explanation to be found, described in two papers that represented the overture to the molecular era of cytochrome b558–cytosolic components interaction [143, 144] (see Sect. 11.1). Finally, there is a large amount of, in part, controversial information on the relation of p22phox gene polymorphism to cardiovascular and cerebrovascular diseases (reviewed in [145]).

4.3

Hide-and-Seek: Where Are the Hemes?

It appears strange that the number and location of hemes in the cytochrome b558 heterodimer took so long to be established and with so many wrong turns. As mentioned in Sects. 4.1 and 4.2, the predictions made for the location of the heme(s), based on the deduced sequences of gp91phox and p22phox, were predominantly in favor of p22phox or shared by the two subunits [135, 137]. Experimental support for the wrong hypothesis (p22phox), using a non-customary methodology, came first from the Segal laboratory [146]. In an interesting illustration of how the path to the truth passes through half-truths, Quinn and coworkers started from the proposal and the evidence that cytochrome b558 was a multiheme protein [120–122, 137] and suggested that the hemes were located within the membrane lipid bilayer (true), one residing in Nox2 (true) and one shared by p22phox and Nox2

(wrong) [147]. In 1995 Cross and coworkers reported in a seminal paper that cytochrome b558 contained two hemes with distinct midpoint potentials [90]. Surprisingly, they adopted the Quinn model of one heme being shared between p22phox and Nox2 or between two p22phox subunits, and one heme, residing in Nox2. This paper also contained two significant predictions. First, it suggested that Arg54, located at the beginning of a transmembrane α helix close to the membrane surface, might form hydrogen bonds with a heme propionate group, as hypothesized by Isogai [129], and proven to be the case for basic residues in the solved structure of csNox5 [130]. Second, it proposed that the third transmembranal α helix contained three potential hemeliganding histidines, the identity of two of which was confirmed by future work. Fujii, known for his work on the hexacoordination of cytochrome b558 heme iron, proposed that the heme was associated with Nox2 and correctly predicted its ligation to His101 and His209 [148]. This paper was published before the two-hemes model became common knowledge and was based on comparison with cytochrome P450 and on awareness of the report by Roos and coworkers on point mutations at His101 and His209 in X-linked CGD [149]. Definitive proof for the location of hemes in Nox2 was obtained by using stable transfection of cDNA of Nox2 and p22phox into African green monkey kidney cells (COS7) [140], a methodology of major impact, pioneered by Dinauer (reviewed in [150]). Spectral analysis and redox titration also showed that the hemes were present in Nox2 and not in p22phox [151]. This significant publication contained the first portrayal of the six transmembrane α helices and ligated hemes, inspired by the analogy of the bis-heme motif of yeast ferric reductase with that of Nox2 (see Sect. 4.4), It also delineated the correct path of electrons: FAD → inner heme → outer heme → outer sphere, where oxygen was encountered. Finally, in impressively designed experiments, involving stable transfection of cell lines with cDNA of Nox2 subjected to site-directed mutagenesis focused on individual histidines, Dinauer and coworkers found that Nox2 maturation, cytochrome b558 dimer formation and O2•- production did not occur in cells transfected with cDNA with mutated His101, His115, His209, His222 [152].

4.4

Bis-Heme Motif Archeology: Ferric Reductases

The ferric reductase of Sacharomyces cerevisiae, coded by the FRE1 gene, was found to have significant sequence homology and spectral and midpoint potential similarities with Nox2 [153, 154]. Four critical histidine residues were identified in the FRE1 protein and were predicted to coordinate a bis-heme structure, as proven by being required for ferric reductase activity and by a heme spectrum [155]. The

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

suggested coordination of the hemes between pairs of histidines located in α helices, the position of the hemes (one above the other) within the lipid bilayer and perpendicular to the plane of the membrane, favorable for transmembrane electron transport, represented a major paradigm shift, echoed by its relevance to Nox2. Two pairs of analogous residues were indeed recognized by spacing and context in Nox2 (His101 and His115; His209 and His222), representing the first identification of the hemes involved in electron transport from FAD to oxygen, later confirmed by the Dinauer group [151, 152]. The evolutionary origin of the bis-heme motif, forming a key part of the ferric reductase domain of the ferric iron reductase superfamily, and its relation to Noxs has been reviewed [156]. The shift from a metal reductase to an O2•--generating protein took place in early eukaryotes. Much was also learned from the similarities of Noxs to yet another member of the ferric reductase superfamily, the “sixtransmembrane epithelial antigen of the prostate enzymes” (STEAPs), which go back in phylogeny to mitochondria and photosynthetic cytochromes [157] (see Chap. 31 by W. Oosterheert, S. Marchese, and A. Mattevi).

5

From the Golden Calf to NADPH Oxidase Monotheism

The binary oxidase concept was shattered definitely by in vitro and in silico evidence for all redox centers being present in Nox2. The detailed information on the location and function of the hemes answered all the questions related to the most distal step in the electron flow, the reduction of dioxygen to O2•-, but left the issue of the location of the NADPH and FAD binding sites, unsolved.

5.1

Functional Evidence for the Presence of All Redox Stations on Cytochrome b558

Knoller and Pick first reported that highly purified cytochrome b558 fully replaced phagocyte membrane in a cellfree Nox activation assay, suggesting but not proving that all redox centers were present in the cytochrome [104]. The significance of this finding was augmented by it being reproduced in a system in which all cytosolic components were purified recombinant proteins [106], shown before not to be flavoproteins and, thus, not likely to contain the NADPH and FAD binding sites [158]. However, due to the, at the time, prevalent idea that NADPH binding sites might be present in p67phox (see Sect. 5.2), the definitive proof was provided by the demonstration by Koshkin that O2•- generation by purified cytochrome b558 could be achieved in the absence of cytosolic activators via a direct effect of anionic

19

phospholipids [108, 109] (see Sect. 6.8). Cytosolindependent Nox2 activation in vitro was used in the elucidation of the mechanism responsible for defective O2•- production in some cases of X91+ CGD [159].

5.2

Identifying the Binding Site for NADPH: A Bumpy Road

Umei was the first to successfully label a 66 kDa component in a solubilized Nox preparation of pig leukocytes, derived from phagocytic vesicles, with NADPH 2′,3′ dialdehyde [69]. The labeled protein was assumed to be membraneassociated but no proof was provided for its identity with Nox2 (non-glycosylated pig Nox2 has a molecular mass of 56 kDa). A follow-up study performed on the particulate fraction of resting and PMA-stimulated human leukocytes led to labeling of a 66 kDa protein [160]. Its identity with Nox2 was negated by the labeling of a protein of the same size in material derived from cells of X-linked CGD patients. Segal and coworkers used 2-azido-NADP+ as an affinity label and detected labeling of a 93 kDa protein, suggested to be glycosylated Nox [124]. Doussière and coworkers used a radiolabeled NADPH analogue and found it binding to a 65 kDa protein in the plasma membrane of bovine neutrophils [161]. A subsequent study by the same group led to the detection of a protein of 80–100 kDa, which was identified as Nox2 by parallel immunoblotting [162]. Using a different affinity label and human leukocytes, Lederer identified an NADPH binding protein of 76 kDa, found to coincide with Nox2 by immunoblotting, and absent in material from cells of a patient with X-linked CGD [163]. The differences between reports and the relation of labeling to prior exposure of cells to Nox activators [162, 163] were, probably, related to the particular methodologies used. Typical of the changing climate of how new paradigms are born, it was not the painstaking in vitro work described above, but the in silico data to emerge later, that served as the decisive argument in accepting the idea that Nox2 was the bearer of all redox centers.

5.3

“The Wrong Man”

The title of this subsection is derived from Hitchcock’s most somber film about an innocent man accused of a crime he did not commit. The search for the location of the binding site for NADPH on the wrong component was less dramatic but slowed down progress in the right direction. The participation of Babior in choosing the “wrong man” is a good example for the truth that even the greatest scientist might be wrong, occasionally.

20

The origin of the paradigm that the NADPH binding site might be present in one of the cytosolic components was in experiments performed by D. Sha’ag in the course of his Ph. D. thesis work in the Pick laboratory. He found that the cytosolic component(s) participating in cell-free Nox activation could be isolated by affinity chromatography on agarose on which the NADPH analog 2′,5′-ADP was immobilized [164]. The component(s) bound to 2′,5′-ADP was eluted by NADPH, but not by NADH, and also by ATP and GTP. The same component(s) were also bound to immobilized ATP and GTP and could be eluted by free nucleotide triphosphates [165]. The two reports were of key importance in the effort to identify and characterize the cytosolic components (see Sect. 7.1). In a series of reports, Babior and collaborators described the property of NADPH 2′,3′ dialdehyde, in the presence of cyanoborohydride, to abolish the ability of cytosol of resting human neutrophils to support cell-free Nox activation and suggested that the affected protein was p67phox [166– 168]. Paradoxically, affinity labeling of membrane and cytosol with NADPH 2′,3′ dialdehyde and 3H- cyanoborohydride revealed a 67 kDa protein in the membrane but not in the cytosol [166]. The most likely explanation for this was that lack of inactivation does not necessarily mean lack of binding, a fact supported by the known ability of NADPH 2′,3′ dialdehyde to also serve as a substrate and by the methodological acrobatics required to demonstrate inactivation [69, 166]. In a final attempt to document the cytosolic origin of the NADPH binding protein, human neutrophils were stimulated by PMA and the membrane fraction incubated with buffer containing 0.3 M KCl, to release proteins expected to be translocated to the membrane in the process of Nox activation. A comparison of eluate from PMA-stimulated cells with that of resting cells, following affinity labeling with [2′-32P]NADPH, revealed the presence of a 32 kDa protein, restricted to the eluate from stimulated cells [169]. Typical for the confusion dominant at the time, the relation of this protein to the cytosolic protein identified by the same investigators by Nox inactivation, was not commented upon. A good illustration for how resistant dominant investigators are to paradigm shifts is the fact that the Babior group returned, 9 years later, with a toned-down version of p67phox possessing an NADPH binding site, localized in the N-terminal part (residues 1–199) [170]. Admitting the strong evidence for the presence of the binding site on Nox2, Babior suggested a complementary role of p67phox in electron transfer, in the eventuality of the failure of the canonical path. The final shot of the cytosolic era was the description of a cytosolic protein of 52 kDa, labeled with a photodependent NADPH analog, detected in membranes of guinea pig neutrophils following cell-free Nox activation, dependent on the translocation of the protein from the cytosol [171].

E. Pick

5.4

Briefly, Back to FAD

From the “almost right” prophecy of Quastel [12], and the finding by Babior that FAD was a sine qua non cofactor of Nox, the concept of the flavoprotein nature of the enzyme was gradually established (see Sects. 2.4 and 2.5). However, the binary model of Nox refused to die and, up to the early nineties of the last century, the “ancient régime” view of a flavoprotein upstream from cytochrome b558 coexisted with the ultimately victorious “cytochrome only” paradigm. However, the fact that highly purified cytochrome b558, used in experiments serving as proof for it comprising all the redox centers, was found to be FAD-free, generated some confusion [53, 104, 106, 108, 109]. Several studies provided the answer to this “pseudo-paradox” by showing that FAD bound non-covalently to Nox2 and dissociated from Nox2 in the course of cytochrome b558 purification. Nox activity of the FAD-depleted cytochrome could be regenerated by re-binding of FAD, if present in large excess in the assay buffer. Segal and coworkers showed that the FAD content of leukocyte membranes did not increase following Nox activation and was low in membranes of cells from X-linked but not autosomal recessive p47phox-deficient CGD patients, negating the existence of a cytosolic flavoprotein [124]. Rotrosen and coworkers found that purification of cytochrome b558 was associated with loss of non-covalently bound FAD and increased dependence on added free FAD for Nox activity [105]. They successfully reflavinated cytochrome b558, depleted of FAD, and isolated by gel filtration a cytochrome active in the cell-free system in the absence of added FAD. An unnoticed (and uncited) detail in these experiments was that reflavination was enhanced in the presence of cytosolic components, a finding noticed 12 years later and suggested to serve as a possible mechanism by which cytosolic components regulate electron flow in Nox2 [172]. The successful reflavination of FAD-depleted cytochrome b558 was reproduced by Nisimoto and Lambeth [53] and Hideki Sumimoto and coworkers confirmed the decreasing Nox activities associated with sequential steps in the process of cytochrome b558 purification, paralleled by increased dependence on added free FAD [173]. A most convincing proof for the flavoprotein nature of Nox2 was provided by Doussière and coworkers, who used a radioactive photoactivable FAD derivative for labeling deflavinated bovine leukocyte membranes [174]. The FAD analog was found to bind to a protein in the 80–120 kDa range, identified by immunoblotting as Nox2; enzymatic deglycosylation shifted the mass of the detected protein to 50–60 kDa.

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

5.5

In Silico Apotheosis: Copy-Pasting from Ferredoxin-NADP+ Reductase

With due respect to the significance of experimental evidence in defining Nox2 as the home of both NADPH and FAD redox centers, the precise definition of the residues involved in binding the NADPH substrate and the FAD cofactor was the result of applying an early mode of bioinformatics. This originated in noticing the homology of sequence motifs in Nox2 with sequences in a superfamily of enzymes that catalyze FAD-enabled electron transfers from NADPH to one-electron carriers. This was made possible by Andrew Karplus solving the structure of spinach ferredoxin-NADP+ reductase (FNR), an enzyme which catalyzes the reversible electron transfer between NADP(H) and the iron-sulfur protein ferredoxin [175]. The protein has an FAD binding domain in the N-terminal half and an NADP+ binding domain in the C-terminal half. Members of the superfamily expressing homology comprised cytochrome P-450 reductase, cytochrome b5 reductase, nitric oxide synthase, nitrate reductase, and sulfite reductase. Three groups of investigators, Segal and coworkers [124]; Rotrosen and coworkers [105], and Sumimoto and coworkers [173] aligned sequence regions of Nox2 with sequence segments of members of the FNR superfamily and published the detected homology regions. The finding that Nox2 was a member of the FNR superfamily represented a major paradigm shift and there is reason to believe that the three groups of investigators reached this conclusion independently, in a very narrow time interval. Rotrosen et al. [105] acknowledged awareness of data of Segal et al. presented at a phagocyte workshop in 1992, at the page proof stage; Sumimoto et al. [173] acknowledged awareness of work by Rotrosen et al. and cited the paper [105]. The following sequence homologies of gp91phox with spinach FNR were detected (the position of residues in the Nox2 sequence is indicated): the pyrophosphate(405MLVGAGIGVTPF416) [105, 124, 173]; ribose(442YWLCRD447) [105, 173]; adenine- (504GLKQK508) [105], and nicotinamide-binding (535FLCGP539) [105, 124, 173] segments of the non-continuous NADPH binding domain, and the isoalloxazine- (335LEWHPFTLTSA345) [124, 173], and ribitol-binding (350FFSIHIRIVGD360) [173] segments of the non-continuous FAD binding domain. Recent progress in solving the structure of members of the Nox family [130] allowed a more sophisticated comparison between corn root FNR and Nox proteins, based on overlay modeling, and substantiated the conclusions reached by alignment [176]. The correct assignment of the binding domains was confirmed by the loss of Nox activity in leukocytes of patients with the X91+ form of CGD, caused by mutations in the binding sites for NADPH or FAD. Examples are a Pro415His mutation in the NADPH

21

pyrophosphate binding site [177], a Cys537Arg mutation in the NADPH nicotinamide binding site [78, 79], and a His338Tyr mutation, in the FAD isoalloxazine binding site [178]. Aligning sequences proved to be more persuasive than site labeling and Nox activity inhibition experiments and the “monotheistic” concept of all redox centers being present in one molecule eliminated binary and even more “pagan” Nox models. A fact to remember is that fundamental information about both the bis-heme redox station, located in the membrane, and the NADPH and FAD redox stations, located in the cytosolic DHR of Nox2, was acquired by discovering their similarities to and evolutionary origin in the ferric iron reductase and FNR superfamilies, respectively. The rather rigid definition of NADPH- and FAD-binding sites, appearing above, is certain to be revised as the result of the marked advances in structural studies of Noxs by cryoelectron microscopy (cryo-EM) and other recent approaches. The specific role of individual residues within the “sites” will be better known, more residues participating in binding will be identified, and a vision of the protein and ligands “wiggling and jiggling by thermal motion” should be closer to reality than the present static rendition (the author thanks Andrew Karplus for suggesting this imagery).

5.6

The Enigmatic Insertion Sequence

Soon after the alignment studies were published, a more in depth analysis revealed the presence of an insertion sequence in Nox2 (residues 484–504), absent in other members of the FNR superfamily [179]. In the resting state, the insertion might prevent access of NADPH to its binding site and its displacement could be related to Nox2 activation [179]. A thorough reexamination of the function of the insertion sequence revealed that, contrary to expectations, its deletion did not result in constitutive Nox2 activity but that it might serve as a target for the translocation of cytosolic components, the binding of which could cause it swerving off the NADPH binding site [180]. It is likely that the insertion sequence contains a binding site for p47phox [181, 182] and that its involvement in binding of p67phox is secondary to and dependent on that of p47phox [183]. Alignment of Nox2 with csNox5 pointed to the insertion sequence of Nox2 corresponding to a region in Nox5 responsible for binding the Ca++-dependent regulator EF-band [130, 184]. In the Nox5 structure, the EF-band binding segment is unstructured [130], a characteristic also evident in the insertion sequence in Nox2, as shown by the fact that it was not visible in the crystal structure of the 385–570 segment of Nox2 (PDB 3AIF) [180]. A more complete understanding of the modus operandi of the insertion sequence is not yet available and, thus, represents a paradigm shift “in progress”.

22

E. Pick

Yet another issue originating in the comparison of structures of members of the FNR superfamily to that of Nox2 was the function of the C-terminal aromatic residue, which is Phe570 in Nox2. In FNRs, this residue occupies a site critical for the interaction of NADPH with FAD [176, 185] (see Sect. 10.2).

5.7

A Retrograde Paradigm Shift Attempt: “Nobody’s Perfect”

The sovereignty of Nox2 as the only catalytic element of the phagocyte Nox appeared to be established. It was, however contested by no lesser person than Babior, who claimed that the hemes in Nox2 do not function as the terminal electron donors to dioxygen [186]. This claim was based on the frequently discussed discrepancy between the rate of O2•production and that of reduction of Nox2 hemes by NADPH (calculations were based on experiments in the cell-free system [187]). The iconoclastic arguments of Babior were answered by a courageous Letter to the Editor by Cross (who worked in the same institution) [188], contesting the calculations and emphasizing the need to perform kinetic experiments in the presence of oxygen, as shown before [92]. The unsatisfactory response by Babior [189] is worth reading and serves as a good example for the veracity of the statement that “Nobody’s perfect”, to cite from Billy Wilder’s unique comedy “Some like it hot”.

6

The Unexpected Emergence of the Cytosol

6.1

A Disputed Birth Certificate

For a membrane imbedded enzyme, possessing all the cofactors required for catalysis, the possible involvement of additional non-catalytic components required for the induction of activity was not foreseen (to the best of my ability to search the literature). Interest focused on transduction pathways from the membrane receptors to Nox, with overwhelming accent on protein kinase C (PKC). It was suggested that the cytosolic components were discovered in the process of searching for the molecular basis of the cytochrome b558-positive autosomal recessive form of CGD [190]. This statement referred to the previously unexplained mechanism of the failure of phagocytes of patients with autosomal recessive CGD, expressing cytochrome b558, to produce O2•- [87], and to the finding that hybridization of monocytes of a patient with X-linked cytochrome b558-negative CGD with monocytes of a cytochrome b558-positive CGD patient, yielded fused cells with normal Nox activity [191]. It was also claimed that the absent or fading activity of Nox derived by extraction from stimulated cells instigated

attempts to develop a method of activating Nox derived from resting cells. Another potential incentive for exploring possible non-cytochrome participants in Nox activation was the finding by Segal and coworkers that leukocytes of patients with the autosomal form of CGD, stimulated with PMA, failed to phosphorylate a 44 kDa protein [192]. The phosphorylated protein was recovered in the membrane and granule fraction and was absent in the cytosol, and its size was corrected to 47 kDa [193], two results the significance of which became obvious only when the cytosolic components were identified and their translocation to the membrane was found to be causally linked to Nox activation. Typical for the period (1985–1986), following the discovery of the cytosolic components but preceding the identification of p47phox and p67phox, some authors did not abandon the binary oxidase model and reasoned about the possibility that the 47 kDa protein might function as an electron carrying molecule upstream of cytochrome b558. With due respect to these plausible scenarios, the author of this chapter took the liberty of telling a different narrative, based on his personal involvement in the early history of the discovery that Nox function was controlled by cytosolic components unrelated to the catalytic core, cytochrome b558.

6.2

The Cell-Free System

The existence of cytosolic modulators was discovered independently in two laboratories, in close temporal proximity, in the course of attempts to elicit Nox activation in membranes of resting guinea pig macrophages [194] and horse leukocytes [195]. The path to the discovery of the cell-free system in the Pick laboratory started in trials to find a common mechanism for the production of O2•- by guinea pig macrophages exposed to a variety of unrelated stimuli [196]. An important factor was the awareness to sporadic reports describing inhibition of Nox activation in leukocytes by agents interfering with liberation or oxidation of AA and to the finding that many Nox activators caused liberation of AA or its oxidized metabolites (reviewed in [197]). Most valuable contributions to the birth of the cell-free system were reports originating in the laboratories of two major figures in Nox history, Kakinuma and Karnovsky, who found that some long chain unsaturated and medium length saturated fatty acids induced O2•- production by intact leukocytes [198–200]. With the above results serving as background, a systematic study in guinea pig macrophages showed that: (1) O2•production in response to eight out of ten stimulants was associated with liberation of AA; (2) Inhibition of phospholipase A2 (PLA2) blocked O2•- generation; (3) AA and other long chain unsaturated fatty acids elicited O2•- production;

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

(4) Oxidation products of AA by lipoxygenase were not more active than AA; (5) Only the ionized form of the fatty acids was active, and (6) The other product of PLA2 action, lysophosphoglyceride, was inactive [201]. This led to the hypothesis that stimulants of Nox activation in phagocytes activate a PLA2 and that AA produced by the action of the enzyme on macrophage phospholipids activates Nox. The likelihood of such a process occurring was enhanced by the account of the richness of macrophages in AA-containing phospholipids (25%) [202] and was extrapolated to other phagocytes, although AA-containing phospholipids in polymorphonuclear leukocytes amounted to only 14%, the dominant fatty acid being oleic acid (31%) [203]. The decision by Bromberg and Pick to attempt activation of Nox in homogenates of unstimulated guinea pig macrophages by AA was prompted by the ability of AA to activate Nox in whole cells, supplemented by the resistance of AA-elicited activation to PLA2 inhibitors, suggesting a direct effect on the enzyme [201]. The ability to activate Nox in a broken cells preparation derived from resting macrophages by AA was first described by Pick at the Pasteur Institute—Weizmann Institute Symposium in Paris, commemorating a century from the discovery of phagocytosis by Metchnikoff, in June 1983. The fundamental characteristics of what came to be known as the “cell-free system” were described in a paper published in October 1984 [194]. These were the following: (1) Addition of AA to a nuclei-free homogenate of resting macrophages led to NADPH-dependent O2•- production; (2) O2•- generation required AA to be in the ionized form and was maximal at concentrations of AA close to the critical micellar concentration (CMC); (3) Long-chain unsaturated fatty acids with one, two, or three double bonds were also active but long-chain saturated acids and prostaglandin E2 (a cyclooxygenasederived metabolite of AA), were not; (4) Stimulants known to activate Nox in intact cells, such as PMA or the chemotactic peptide formyl-Met-Leu-Phe (fMLP), did not elicit O2•production, indicating that signal transduction functioning in whole cells was not operating in the cell-free system; (5) FAD but not FMN enhanced AA-induced O2•- production; (6) O2•- production was supported by NADPH (Km = 49 μM) but not by NADP, NADH and NAD, and (7) No O2•- generation was elicited by AA in separated particulate and cytosolic fractions of homogenates but activity was recovered by combining the two fractions. A common error in reviews dealing with Nox is to look upon the cell-free system as merely a methodological advance. That this is not so is evident in the Abstract of the paper, stating that “AA-elicited O2•- formation is dependent on the cooperation between a subcellular component

23

sedimentable at 48,000 × g (probably containing the O2•forming enzyme) and a cytosolic factor”. In the Discussion section, various mechanisms by which fatty acids could act were considered, with emphasis on the “striking dependence of the activation on a cytosolic component” [194]. The identity of the cytosolic component was a central issue, the preferred candidate being a PKC, inspired by the report of AA activating a PKC in leukocytes [204].

6.3

Meanwhile, in Ghent

The very fact that a significant scientific novelty so often emerges simultaneously from several laboratories is an index both to the strongly traditional nature of normal science and to the completeness with which that traditional pursuit prepares the way for its own change Thomas S. Kuhn, The Structure of Scientific Revolutions, 1970 [14]

Heyneman and Vercauteren, described the activation of Nox in horse leukocyte homogenates by the sodium salt of oleic acid (OA) [195]. Their paper was published in December 1984, a narrow 2 months after the publication of the Bromberg and Pick paper [194], but the work in both laboratories was done independently and the investigators were unaware of a similar system being developed elsewhere (a perfect illustration of the statement by Kuhn). It is also of interest that the observation leading to the design of the cellfree system was essentially identical, namely that Nox was activated in intact phagocytes by unsaturated fatty acids. Heyneman first reported that the sodium salt of OA elicited Nox activation by horse leukocytes, expressed in enhanced oxygen consumption and the generation of O2•- and H2O2 [205, 206]. The OA-activated Nox was localized in the particulate fraction and was absent in material derived from OA-treated leukocytes of a CGD patient [207]. Heyneman and Vercauteren next found that the sodium salts of OA and linoleic acid elicited oxygen uptake and O2•- production when added to a homogenate of horse leukocytes [195]. The characteristics of the system were identical to those of Bromberg and Pick, most notably the absolute requirement for the participation of the cytosol. An important supplement was the finding that a homogenate of leukocytes of a CGD patient did not respond to OA, and a reminder of the infancy of Nox research was the observation that a homogenate derived from leukocytes treated with the thiol alkylating agent PCMB was refractory to activation by OA (see Sect. 2.9).

24

6.4

E. Pick

A Controversial Background Nevertheless Leads to Doing the Right Experiments

The roots of the cell-free system were in the belief that activation of Nox in phagocytes involved the participation of a PLA2 and the consequent liberation of AA. Paradoxically, the role of PLA2 in Nox activation in vivo, remains a controversial subject and the author of this chapter must admit that an attempt to make sense of the large amount of older literature dealing with the subject failed to lead to a clear picture (see Discussion section in [197]). Reports that Nox activation in intact cells by opsonized zymosan [208], AA itself [209], and by PMA (known to be mediated by PKC) [210], was blocked by PLA2 inhibitors, were only some of numerous examples contributing to the confusion. The uncertainty was further augmented by the suggestion that PKC might activate a PLA2 and, indirectly, cause AA release [211]. Additional reasons for the controversial status of the subject were the heterogeneity of phagocyte PLA2s [212], and the unresolved discrepancy between a cytosolic PLA2 acting on an already assembled Nox [213], and data showing that AA derived from cytosolic PLA2 activity was responsible for oxidase assembly [214]. Significantly, a comprehensive review of the molecular pathways linking membrane receptors to Nox activation failed to mention the role of PLA2 and fatty acids [215]. Fortunately, the role of PLA2 was called into question following the success of the cell-free systems and, once the new paradigm was established, the history of its parenthood became irrelevant.

6.5

The Cell-Free Becomes “Legit”

The confidence in the cell-free system being an as close as possible representation of Nox activation in vivo was enhanced by two descriptions of its application to human leukocytes. Linda McPhail and coworkers listed as reasons for engaging in the design of a cell-free system their earlier findings that agents activating Nox in vivo acted by various transduction mechanisms [216], the ability of fatty acids to elicit O2•- production by intact leukocytes [200], and the description of the AA-activated cell-free system in guinea pig macrophages [194]. McPhail and coworkers described a cell-free system in human leukocytes at the meeting of the American Federation for Clinical Research in May 1984 (Abstract [217]) and the results were published in May 1985 [218]. Its properties were similar to those described in guinea pig macrophages [194] and horse leukocytes [195], sharing its most prominent feature, the requirement for a cytosolic component. The report also contained the novel observation that treating whole cells with subliminal

concentrations of Nox stimulants enabled membranes to respond to AA in the absence of cytosol, representing the first indicator of stimulus-induced translocation of the to be discovered cytosolic components. Curnutte designed a cell-free system and presented it at the meeting of the American Society of Hematology in December 1984 (Abstract [219]). In a publication appearing back to back with that of McPhail and coworkers, Curnutte described an essentially identical AA-activated cell-free system in human leukocytes [220]. The background for his work was the awareness of the ability of AA to activate Nox in intact cells with a very brief lag time, implying a direct action on the enzyme. A crucial finding made by Curnutte was that the membrane fraction from leukocytes of a patient with the X-linked form of CGD was unable to support AA-elicited O2•- generation when mixed with either its own cytosol or that of a normal subject, whereas the cytosol of the patient’s cells cooperated efficiently with membranes of normal leukocytes [220]. The close parallelism between the failure of membranes of X-linked CGD patients to produce O2•-, when derived from intact stimulated cells and the inability to respond to AA in the presence of normal cytosol served as a dominant argument for the cell-free system representing a fair model for Nox activation in vivo. This key initial observation was followed by a more extended study on cells of patients with X-linked and autosomal recessive cytochrome b558-negative CGD, showing failure of membranes but not the cytosol to support AA-elicited O2•- generation in a cell-free system when combined with the reciprocal component originating from cells of normal subjects [221]. Finally, the cytosolic contribution to Nox activation acquired significant backing from the description of the inability of cytosol of cells of patients with autosomal recessive cytochrome b558-positive CGD to cooperate with normal membrane in a cell-free system [222, 223]. It is of historical interest that in the paper published in 1989 [223], Curnutte predicted the multitude of cytosolic components and suggested that one of these might be the 48 kDa protein, the phosphorylation of which was described by Segal and coworkers [192, 193]. All questions related to the identity of the cytosolic components were answered by the introduction of the semi-recombinant cell-free system in which full amphiphile dependent activation was achieved in mixtures of purified relipidated cytochrome b558 and recombinant p47phox, p67phox, and Rac1GTP [106]. p40phox was absent from semi-recombinant cellfree systems, with a rare exception [118] and I am not aware of published data on the possible influence of its inclusion (however, see Sect. 11.6.3). Skepticism toward the cell-free system being a fair representation of events in the intact cell was minimal. One publication stated that activation of Nox in disrupted leukocyte preparations by high concentrations of AA reflected nonphysiological changes in the lipophilic environment of

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

the enzyme, mimicking the action of detergents [224]. Questions about fatty acids having a direct effect on Nox were also raised on the basis of a report that fatty acids acted in intact cells by activation of phospholipase C [225].

6.6

Cell-Free Activators: Beyond Fatty Acids

The cell-free system was the gateway to the discovery of the cytosolic components and represented the closest in vitro model of Nox complex assembly, amenable to rigorous quantification. However, elucidating the mechanism of cellfree activation also provided unexpected clues to the molecular details of the assembly. Key sources in the search for a mechanism of action were publications setting out a requirement for both an anionic and a hydrophobic character of fatty acids able to elicit ROS production by intact phagocytes [199], as well as a demand for an ionized form of fatty acids for whole cell [226] and cell-free [194] Nox activation. A rather confusing contribution was a plethora of publications reporting activation of Nox in intact cells by a variety of surface-active agents, comprising various action mechanisms. These included the nonionic detergents saponin [227] and digitonin [228–230], and the ionic detergents deoxycholate [228] and SDS [231, 232]. The proposed mechanisms of action had in common a first step of binding to the plasma membrane, followed by the delivery of a messenger to activate Nox. ROS production by guinea pig leukocytes in response to detergents was found to be limited to anionic detergents with a C12 alkyl chain, at concentrations below the CMC, and dependent on a sulfate or carboxylate anion head [231]. Bromberg and Pick described the ability of SDS to activate Nox in a cell-free system derived from guinea pig macrophages under experimental conditions identical to those of the AA-activated cell-free system [233]. Unlike the rather nonspecific effect of surface-active agents on intact phagocytes, the cell-free system was activated specifically by anionic amphiphiles, such as SDS, sodium dodecyl sulfonate, and lithium dodecyl sulfate (LiDS), but was unresponsive to cationic, amphoteric, and nonionic detergents (reviewed in [197]). Activation by anionic amphiphiles suggested that long chain unsaturated fatty acids, consisting of a hydrophilic carboxylic group (head portion) to which a hydrophobic long chain of alkyl groups, with a variable number of double bonds (tail portion), was attached, acted by an identical mechanism. Many questions related to the molecular mechanism of cell-free activation by amphiphiles were not resolved. Thus, long-chain unsaturated fatty acids were found to be the most active [194, 234], but certain medium-chain saturated fatty acids (12–16 carbons) were also reported to be capable of

25

activation [235–237]. Questions were also raised by the finding that the trans form of AA failed to activate Nox and prevented activation by the cis form [238], whereas the trans form of the shorter and less unsaturated oleic and linoleic acids were active [234, 235]. Physicochemical properties were found to influence markedly the ability of amphiphiles to be active in the cell-free system; thus, fatty acids with the lowest Krafft point were the most active [239] (the Krafft point is the temperature above which the solubility of a detergent rises sharply). To this category belonged the fatty acids which were the most effective in the cell-free system because their Krafft points were below the temperature at which cell-free assays were usually run (24 °C). As noted before for AA, the concentrations of unsaturated fatty acids active in the cell-free system were close to their CMC, when the assays were performed at 24 °C. In sharp contrast, the activating concentrations of SDS and LiDS were 70–100 times lower than their CMC values, a finding for which no explanation is yet available (reviewed in [197]).

6.7

An Unnoticed Paradigm Shift: Assembly and Catalysis

The cell-free system was the first to indicate clearly that a process of assembly of a complex from multiple parts was at the heart of Nox activation. Canonical cell-free assays were started by the addition of the amphiphilic activator, followed by NADPH, to initiate O2•- generation. The lack of awareness of the fact that the order of addition was important obscured the distinct roles of the activator and the substrate (an illustrative example was the inadvertent reversal of order in the cell-free system described by Curnutte, in which the addition of the substrate (NADPH) preceded that of the activator (AA) [220]). In a variety of specially designed “two-steps” assays, the incubation of membranes and cytosol with the amphiphilic activator (assembly step) preceded the addition of NADPH (catalytic step) [240–242] (reviewed in [243]). Such separation allowed focusing on the assembly process by varying the concentration of membrane and cytosolic components, the concentration of activator, and the duration and temperature of incubation. An illustrative example was a study in which full assembly was achieved in 5 min at 25 °C but only after 30 min, at close to 0 °C [240]. Although the effect of temperature was possibly mediated by factors other than the kinetics of assembly, such as the nature of the membrane lipid bilayer [244, 245], the importance of time and temperature supported a view of Nox assembly as a multi-component chemical reaction, with conventional Vmax and Km parameters.

26

6.8

E. Pick

The Poor Man’s Cell-Free: Renouncing the Cytosol

A methodological advance in cell-free studies was replacing phagocyte membranes by solubilized membranes, yielding liposomes, following lowering the concentration of the solubilizing detergent to values below the CMC by dialysis or dilution [62]. Native membrane lipids were occasionally replaced by specific purified or synthetic exogenous phospholipids [246] that were also used in the preparation of purified cytochrome b558 liposomes [104]. Abandoning the trodden path, Koshkin developed a cellfree system consisting of purified cytochrome b558 relipidated with a mixture of low purity phosphatidylcholine (PC) and phosphatidic acid (PA) or PA alone, which generated O2•- in the absence of cytosol [108, 109]. Grading the ability of various phospholipids to support cytosol-independent Nox activation revealed that the anionic character of the phospholipid determined its efficiency, as demonstrated by the potency of PA alone and by the finding that the activity of PC preparations was inversely related to their purity, indicating that contaminating anionic phospholipids were responsible for activity [108]. The anionic amphiphile LiDS, enhanced cytosol-independent activation but at concentrations much lower than those customary in the canonical cell-free assay. The cytosol-independent effect of LiDS suggested that cytochrome b558 was a potential target of anionic amphiphiles. In a series of extensive studies, Jesaitis and coworkers used intermolecular resonance energy transfer from a fluorescent donor, with affinity for cytochrome b558, to the heme acceptors to investigate whether the amphiphiles exerted an allosteric effect on the cytochrome. AA, SDS, and LiDS indeed induced a pronounced conformational change in cytochrome b558, at concentrations known to be effective as Nox2 activators in the cytosol-dependent cell-free system [247]. Similar conformational changes were caused by the anionic phospholipids PA and phosphatidylserine (PS) but not by the neutral PC or phosphatidylethanolamine (PE) [248]. In yet another example of a paradigm shift arising independently from the work of unrelated investigators, Paolo Bellavite and coworkers described the induction of O2•-generation by solubilized plasma membranes of pig neutrophils exposed to PA, in the absence of cytosol [249]. The impact of the Bellavite report was somewhat diminished by the fact that the source of O2•- was a whole membrane fraction which, at the time, did not eliminate the possible presence of a non-cytochrome component.

7

The Cytosolic Aristocracy: p47phox and p67phox

Following the discovery of cytosolic participation in cell-free Nox activation, the postulate that activation was the result of productive assembly of cytochrome b558 with cytosolic components matured into the paramount concept and model (reviewed in [250–252]). This paradigm and the finding that most cases of autosomal recessive forms of CGD were due to a cytosolic defect, served as the impetus for attempts by many investigators to isolate, identify, and characterize the cytosolic component or components. The “guesses” as to the identity of cytosolic component(s) comprised PKC, resting on the extensive evidence for the involvement of PKC in Nox activation by the popular phorbol ester, PMA [253]. The data of Segal and Heyworth [192, 193] on the failure to phosphorylate a 44 (47) kDa protein in leukocytes of patients with the autosomal recessive form of CGD, associated with a fault in the cytosol, seemed compatible with a protein kinase (though, as it turned out later, the cytosol contained the target of the kinase). A more careful reading of the literature would have led to the abandonment of the PKC hypothesis since Seifert and Schultz showed that the PKC inhibitor H7 did not affect O2•- production in the cell-free system [234]. Another candidate was a protein bearing the binding site for NADPH; this hypothesis led to the finding that the cytosol contained at least two proteins required for Nox activation [164, 165]. Finally, the early literature on the enhancing effect of guanine nucleotides on Nox activity [110, 111, 240, 242] served as the impetus for the idea that the cytosol contributed an essential G protein. This prediction turned out to be fulfilled but, paradoxically, served as the methodological justification for experiments that led to the identification of p47phox and p67phox and not of a G protein [254].

7.1

The Unlikely Road to Success: The Wrong Affinity Gels Yield the Right Proteins and a Notorious Rabbit Emerges

The identification of p47phox and p67phox was characterized by a paradox (binding of the proteins to the “wrong” affinity chromatography gel) and by sheer luck (a polyclonal antiserum that recognized both proteins). Robert Clark and coworkers applied leukocyte cytosol to GTP-agarose, in the belief that the cytosolic component was a G protein, or to 2′,5′-ADP-agarose, and hoped to elute the component with GTP or ATP [254]. The choice of 2′,5′-ADP-agarose and ATP was based on the work of Sha’ag and Pick on recovery

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

of cytosol-related Nox activating material bound to 2′,5′-ADP-agarose, by ATP [164]. A single peak was eluted from both affinity gels by GTP or ATP and supported cellfree Nox activation by leukocyte membranes. A critical step was raising a polyclonal rabbit antiserum against active peak material (the notorious B-1 antiserum), which recognized two proteins on SDS-PAGE, corresponding to molecular masses of 47 and 67 kDa, in the single peak eluted from both affinity gels and was present in the cytosol of myeloid cells [254]. The presence of the two proteins correlated with the ability of cytosol to support cell-free activation and they were both absent in non-myeloid cells. Later work demonstrated that undifferentiated myeloid leukemic cells lacked both p47phox and p67phox but these appeared in the course of differentiation to neutrophil or monocyte phenotype induced by retinoic acid and interferon γ, respectively [255]. The expression of the cytosolic components correlated with the ability of the differentiated cells to produce O2•- in response to PMA and the capacity of the cytosol to support cell-free Nox activation. On reading the paper about three decades later, one wonders about the absence of the “third cytosolic component” (Rac) and the mechanism of Nox activation in its absence, a query also applying to the activity of a fraction eluted from 2′,5′-ADP-agarose, to which macrophage cytosol was applied [164]. The likely explanation is that Rac was present in the membrane as a loosely attached protein, as described by Vignais and coworkers [256]. The absolute validation of the function of the two components was supplied by the absence of one or the other of the two proteins in the cytosol of leukocytes of patients with autosomal recessive CGD and their presence in patients with X-linked cytochrome b558-negative CGD [254]. A larger study, involving 36 patients with autosomal recessive CGD with defective cytosolic activity in the cell-free system, confirmed the absence of p47phox (33% of all CGD patients) or p67phox (5% of all CGD patients), assessed by immunoblotting with the B-1 antiserum [257]. In a back-to-back publication, in the same issue of Science, Harry Malech and coworkers used a conventional approach to fractionate leukocyte cytosol by anion exchange chromatography [258]. Whereas this led to the detection of a single fraction with minimal ability to support cell-free activation, an ingenious procedure to detect cryptic activity by supplementing fractions with sub-activating amounts of cytosol, led to the identification of three fractions, which, when combined, could replace cytosol. Use of the B-1 antiserum led to the identification of two fractions as p47phox and p67phox, and a third component of unknown nature. The p47phox and p67phox fractions were capable of restoring the cell-free activity supported by cytosols of distinct groups of autosomal recessive CGD patients, proving that these were lacking either p47phox or p67phox; the third fraction was

27

incapable of restoring the cell-free activity supported by cytosol of any CGD patients [258]. A noteworthy finding was that both cytosolic components translocated to the plasma membrane of leukocytes stimulated by ligands eliciting O2•- generation and in a cellfree system consisting of membranes and cytosol exposed to an activating concentration of AA, the translocation being closely correlated with Nox activation [259]. These findings supported a role of cytosolic components as activators of the electron flow in Nox2 but the mechanism by which this was achieved and the particular contribution of each component were not understood. Following the isolation of p47phox and p67phox and establishing their relation to two forms of autosomal recessive CGD, the presence of the two proteins in several species was demonstrated. Roos and coworkers used ion exchange chromatography to isolate p47phox from human neutrophil cytosol, based on the principle of combining fractions inactive by themselves [260]. Follow-up work led to the isolation of p67phox, distinct from p47phox, and of an unidentified protein that enhanced the activity of p47phox combined with p67phox, in the presence of GTP [261]. Segal and coworkers isolated and characterized a 47 kDa phosphoprotein from the cytosol of PMA-stimulated human neutrophils [262]. Ishimura and coworkers purified p47phox from porcine leukocyte cytosol, which cooperated with a 63 kDa protein, their activating ability being enhanced by a partially purified third component [263]. Vignais and coworkers isolated and characterized p47phox from bovine leukocytes, validating the ability to be phosphorylated and to translocate to the membrane upon cell activation [264]. There were two reports on the isolation of p67phox, from porcine and bovine leukocytes. Ishimura and coworkers [265] were among the first to report a native molecular mass of 180 kDa, a property that turned out to be revelatory of the structure of p67phox. Dagher and Vignais [266] also noted a native molecular mass larger than expected, its appearance in the course of differentiation of myeloid leukemic cells, and the translocation to the membrane of PMA-stimulated leukocytes.

7.2

Cloning of p47phox and p67phox Following Isolation: “The Last of the Mohicans” Approach

Most work on Noxs starts these days with cloning of the proteins. p47phox and p67phox were, however, cloned by two groups following their isolation and partial characterization. Malech and coworkers obtained p47phox cDNA clones by using the notorious B-1 antiserum to screen an expression library derived from differentiated HL-60 myeloid precursor cells [267]. Positive clones were identified by the ability of

28

the expressed protein to interfere with the detection of p47phox by immunoblotting with antibody B-1. The amino acid sequence deduced from the cDNA contained a very basic sequence and a serine-rich region, representing an optimal target for PKC-mediated phosphorylation, two facts of capital importance for the function of the protein (see Sect. 11.1). Recombinant p47phox produced in Escherichia coli (E. coli) was able to restore deficient cell-free activity of cytosols lacking p47phox from autosomal recessive CGD patients. The authors noted a segment of sequence similarity with ras GTPase activating protein (ras GAP)). It was of interest that ras GAP shared amino acids with phospholipase C-148 (PLC-148) and non-receptor tyrosine kinases, two proteins containing src homology region 3 (SH3) domains [268, 269]. Clark and coworkers also cloned the cDNA of p47phox [270]. Expressed fusion proteins of a size close to 47 kDa were recognized by B-1 antiserum and reconstituted O2•production by p47phox-deficient cytosol of an autosomal recessive CGD patient. The deduced amino acid sequence and the presence of serines surrounded by basic residues, serving as potential phosphorylation sites, essentially confirmed the data of Malech and coworkers [267]. Clark also detected a significant homology to proteins encoded by the src oncogene, as well as to PLC-148 and α-fodrin, proteins that contained a SH3 domain (reviewed in [271]). It is of interest that the sequence similarity data led to the identification of the C-terminal SH3 domain by the Malech group [267] and the N-terminal SH3 domain by the Clark group [270]. Soon after the cloning of p47phox, the cloners joined forces and generated cDNA clones encoding for p67phox by screening an expression library derived from differentiated HL-60 cells with the B-1 antiserum [272]. Recombinant protein, produced using cDNA encoding p67phox, restored the lacking ability of cytosol from p67phox-deficient CGD patients to support cell-free activation but failed to restore that of cytosol from p47phox-deficient patients. The nucleotide sequence predicted a protein of 526 amino acids and the C-terminal region showed sequence similarity with the SH3 domain present in proteins possessing the src oncogene. A second region of similarity with SH3 was detected in the midst of the p67phox sequence (residues 243–298). The identification of two SH3 domains in both p47phox and p67phox was a significant paradigm shift that led to the discovery of the key role of SH3 domains in protein–protein interactions participating in Nox assembly (reviewed in [250–252]). The authors also noted the presence of proline-rich regions “together with SH3 motifs” in both p47phox and p67phox, an observation the significance of which became apparent only several years later [143, 144] (see Sect. 11).

E. Pick

8

Evolution Runs Out of Genes and Has to Borrow from Neighbors

The absolute requirement for either Rac1 or Rac2 in the assembly and consequent activation of Nox was established at the completion of a long-winded and tortuous road, the origins of which were multiple and unconnected. The first indication for the involvement of a G protein originated in the work of Seifert and Schultz, who found that nonhydrolysable GTP analogs enhanced and nonhydrolysable GDP analogs inhibited amphiphile-dependent cell-free Nox activation [110, 234]. Gabig and coworkers independently described similar results and made the important observation that the effect of GTP was resistant to pertussis and cholera toxins, indicating that trimeric G proteins were not involved [111]. Further supportive evidence for the dependence of Nox activation on a G protein accumulated but failed to reveal its identity [240, 242, 273]. Finding “the right” G protein was delayed by going up a blind alley by concentrating on Rap1A, based on the claim that it co-purified with cytochrome b558 [112] (see Sect. 3.6). The birth of the cell-free system led to a worldwide effort to fractionate the cytosol, in the course of which what emerged to be a complex of Rac with Rho GDP dissociation inhibitor (RhoGDI) was isolated by several groups. Malech and coworkers isolated a fraction of neutrophil cytosol (NCF-3), the supplementation of which with fractions known to contain p47phox and p67phox, was required for cell-free Nox activation but there was no clue to the identity of the protein with the exception of the finding that addition of GTP to the mixture of the 3 fractions doubled activity [258]. Paradoxically, a targeted trial to find the G protein by affinity chromatography of neutrophil cytosol on GTP-agarose led to the isolation of p47phox and p67phox [254]. In a reversed paradoxical situation, Sha’ag and Pick found that a fraction of macrophage cytosol (σ1) which did not bind to GTP-agarose turned out later to contain Rac1-RhoGDI [165]. A hint that σ1 might contain the sought-after G protein was the unexpected finding that it was present in thymus, lymph nodes, brain and a myeloma cell line and cooperated with a macrophage cytosol fraction (σ2), found later to contain p47phox and p67phox [274]. Preliminary results indicated a molecular mass of 30 to 52 kDa. This was the first indication that one of the cytosolic components required for Nox activation might be a protein not limited to myeloid cells. A cytosolic component of 50 kDa, potentially identical with the Rac-RhoGDI complex, was isolated from pig neutrophil cytosol [275]. Finally, as mentioned above, Roos and coworkers isolated a component of human neutrophil cytosol (SOC I) which cooperated with fractions containing p47phox and p67phox; significantly, the cooperation was dependent on the

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

presence of GTP but the authors refrained from associating it with a G protein [261].

8.1

The Exhilaration of Identifying a “New Protein” and the Frustration of Finding That It Was Already Known

The identification of the “third cytosolic component” was achieved by a combination of old-fashioned protein purification and the use of the cell-free system. Pick and Arie Abo purified the protein to near homogeneity from the σ1 fraction of the cytosol of guinea pig macrophages [276]. It was originally thought to be a single protein of 46 kDa but was found to be a heterodimer, composed of two proteins of 22 kDa and 24 kDa, and it was suggested that one of the two proteins was a small GTPase. In a collaborative study with Segal and Alan Hall, the purified dimer was subjected to tryptic digestion and amino acid sequencing and regions of homology were found with regions in the small Rho GTPase Rac1 and in the regulatory protein RhoGDI [277]. Thus, σ1 was a complex of Rac1 and RhoGDI. A key finding was that recombinant Rac1 in the GTP-bound state was able to replace σ1 in a cell-free system, proving that the component in the Rac1-RhoGDI complex required for Nox activation was Rac1 [277]. Indeed, the 22 kDa and 24 kDa proteins of the complex could be separated by reverse phase chromatography or by anion exchange chromatography in the presence of sodium cholate [278]. This allowed amino acid sequencing and immunoblotting of the two isolated proteins and confirmed their identity with Rac1 and RhoGDI. It also allowed testing of individual fractions in the cell-free assay, demonstrating that activation was supported only by the 22 kDa protein, which bound [35S] GTPγS, and a fully functional complex could be reconstituted by removal of the detergent [278]. The expectation of discovering a novel cytosolic component present only in myeloid cells was not fulfilled; both proteins were already known and performed various functions in many types of cells. Thus, Rac1 and Rac2 were discovered as members of the Rho GTPases family by the isolation of cDNA clones in a differentiated HL-60 cells library [279]. RhoGDI was first isolated by Yoshimi Takai and coworkers from the rabbit gut as a protein which inhibited the dissociation of GDP from Rho GTPases [280] and was found to belong to the RhoGDI-1 (or RhoGDIα) subfamily (reviewed in [281]). The ubiquity of both Rac1 and RhoGDI explained the precocious finding that σ1 was also present in non-myeloid cells and confirmed the prediction made in the Discussion section of the paper that it might be the bearer of Nox-unrelated functions in other cells [274]. Almost simultaneously with the identification of Rac1, Gary Bokoch and Ulla Knaus identified as Rac2 a protein

29

purified from human neutrophil cytosol, which augmented subliminal cytosol dependent cell-free Nox activity [282]. A more complete purification sequence and characterization of Rac2 comprised immunoblotting and amino acid sequencing [283]. Unlike Rac1, which was expressed in many cells and was the dominant isoform required for Nox activation in guinea pig macrophages [277] and human monocytes [284], Rac2 was specific for the hematopoietic lineage and was the dominant Nox-related form in human [283] and murine [285] neutrophils. However, the dependence of the involvement of a particular Rac isoform on the nature of the receptor trigger leading to Nox activation suggested that mechanisms upstream of Rac proper might be involved. The physiological form of Rac proteins involves a posttranslational modification by which a 20-carbon lipophilic geranylgeranyl isoprene is attached to the cysteine in the C-terminal 189CAAL192 sequence by the enzyme geranylgeranyltransferase type I (GGTase-I) (a process known as prenylation) (reviewed in [286]). In a series of elegant studies, it was shown that Rac translocated to the membrane by virtue of a combined mechanism consisting of the chargedependent binding of the positively charged C-terminus to the negatively charged inner aspect of the membrane and the insertion of the hydrophobic prenyl tail into the lipid bilayer [287–291]. Insufficient charge-dependent binding, such as seen with Rac2, which contains only 3 basic residues compared to 6 basic residues in Rac1, was fully compensated by prenylation [288]. The importance of prenylation-dependent hydrophobic interaction with membranes was illustrated by the findings that prenylated Rac1 was bound to protein-free PC vesicles and that such vesicles prevented cell-free Nox activation supported by prenylated Rac1 [289]. On the other hand, the relevance of the positive charge at the Rac C-terminus was indicated by the poor binding of a prenylated [p67phox-Rac1] chimera in which the Rac moiety was replaced by the uncharged C-terminal residues 189–192, versus strong binding by a chimera containing C-terminal residues 178–192, encompassing the 6 basic residues [290]. Further proof for the cooperation of charge and hydrophobicity was provided by the enhanced binding of prenylated Rac1 to protein-free PC vesicles, enriched in anionic phospholipids [291]. The dual nature of membrane attachment is also used as a signal transduction means, in response to stimulation of surface receptors acting via the generation of specific phosphoinositide messengers (reviewed in [292, 293]). Prenylated Rac1 and 2 were first shown to support cellfree Nox activation by Takai and coworkers [294, 295]. Prenylated Rac was found to be a superior activator compared to the nonprenylated form and interacted with “small G protein guanine dissociation stimulator” (smg GDS) [296]), which stimulated GTP uptake and enhanced Nox activation, and with RhoGDI, which acted as an

30

inhibitor. The ability to support Nox activation was dependent on Rac being in the GTP-bound form, a state that could be achieved by a GDP to GTP exchange procedure at low Mg2+ concentration [289], by using the Q61L mutant, which was constitutively in the GTP-bound form [297], or by the action of a guanine nucleotide exchange factor (GEF) [298].

8.1.1 Getting Help from GEFs The enhancement of Nox activity in vitro by exogenous GTP and nonhydrolysable GTP analogs added to the assay medium served frequently as proof for the involvement of a G protein. However, nucleotide exchange to GTP only occurs at very low Mg2+ concentrations causing the dissociation of GDP [283], a situation not existing in cell-free assays. Heyworth and coworkers authored the definitive study on Nox activation by prenylated Rac showing that it could bind exogenous GTP by virtue of the presence of a hypothetical “stimulatory guanine nucleotide exchange protein” (probably inspired by Takai’s smg GDS) in the membrane, whereas nonprenylated Rac was unresponsive to a GEF-like protein and required preloading with GTP [299]. Both posttranslationally modified and unmodified Rac were equally potent Nox activators, when in the GTP-bound form. The proposal that prenylated Rac binds GTP from the medium with the assistance of a “GDP/GTP-exchange stimulating protein” was also made by Vignais [256] and Bokoch proposed the existence of a membrane-associated GEF, which was responsible for GDP to GTP exchange on Rac in the process of dissociation from RhoGDI [300]. Pick and coworkers described the ability of the GEFs Trio and Tiam 1 to support cell-free Nox activation by mixtures of membranes, p67phox, prenylated Rac-GDP, and GTP, in the absence of p47phox and amphiphile [298, 301]. In another example for the complexity of GEF action, its role in Nox activation was illustrated by a cell-free system meant to imitate the whole cell situation. Macrophage membrane liposomes responded by O2•- production upon exposure to Rac(GDP)-RhoGDI complex, p67phox, a GEF (Trio or Tiam 1), and GTP. Enrichment of membranes with the anionic phospholipid phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P3), reproducing membrane phospholipid remodeling by phagocyte receptor stimulation [292], resulted in enhanced Nox activation [302]. Discussing the detailed mechanism of action of GEFs acting on Rho proteins is beyond the scope of this chapter (reviewed in [303]) (see Chap. 18 by Y. Lin and Y. Zheng). The participation of Rac-focused GEFs in Nox activation in vivo and its connection to phosphoinositides-mediated signal transduction are also not covered (reviewed in [304] and, briefly, in [305]). 8.1.2 Untypical Rac CGDs Only relatively few cases of CGD affecting Rac2 were described. Neutrophils of a patient with an Asp57Asn

E. Pick

mutation in one allele, expressing both the mutant and normal alleles, produced lower amounts of O2•- in response to some but not all stimuli, and the cytosol failed to cooperate with normal membranes in a cell-free assay [306]. Asp57 is essential for binding GTP and is conserved in mammalian Rho GTPases. The level of Rac2 was low (41%) and the mutant Rac2 had an impaired capacity to bind GTP but unaltered binding of GDP. The presence of the mutant Rac2 interfered with the ability of normal cytosol or wild type Rac1-GTP to support O2•- production in a cell-free system. The authors suggested that the dominant negative behavior of Rac2 Asp57Asn in the intact cell was the consequence of sequestration of a Rac-specific GEF. The mutant Rac2 was, indeed, shown to bind GEF in vitro but did not respond by the expected nucleotide exchange [307]. Expression of Rac2 Asp57Asn in NIH/3 T3 cells resulted in characteristic phenotypic changes seen in cells with dominant negative mutants of Rac1. Three other patients had a Rac2 Glu62Lys mutation which caused lack of response to the GTPase activating protein (GAP), leading to impaired GTP hydrolysis resulting in a prolonged GTP-bound state and consequent over-activation of downstream Rac effectors and O2•- overproduction [308]. Glu62 in Rac1 was identified as one of four residues making direct contact with a bacterial GAP domain, suggesting that mutating the corresponding residue in Rac2 could inhibit binding of GAP and lead to extended Rac activation [309]. The opposite effect of GAP (enhanced GTP hydrolysis) was demonstrated by the inhibition by a bacterial GAP domain of Nox2 activation in a cell-free system, supported by Rac1-GTP but not by Rac1 exchanged to nonhydrolysable GTP (Y. Litvak, Z. Selinger, S. Molshanski-Mor, E. Pick, unpublished results).

8.2

Divorce, Rac-RhoGDI Style

Both Rac1 and Rac2 are present in the cytosol of phagocytes exclusively in a complex with RhoGDI [310]. Among the three major forms of RhoGDI that exist in mammals, Rac bound with highest affinity to the ubiquitously expressed RhoGDI1, also known as RhoGDIα (basic facts about RhoGDI are reviewed in [281, 311, 312]). RhoGDI interacted with both the GDP-bound and GTP-bound forms of Rac but the affinity for the GTP-bound form was ten-fold lower [313]. The structural particulars of the Rac1–RhoGDI complex were solved by X-ray crystallography and revealed two regions of contact. The first was in the N-terminal region in RhoGDI, which contacted the switch I and II regions of Rac and interfered with nucleotide exchange; the second was in the C-terminal region, forming a “pocket” responsible for the hydrophobic binding of the prenylated tail of Rac, although

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

the C-terminal polybasic region of Rac also contributed to binding [314, 315]. Dissociation of the Rac-RhoGDI complex is an obligatory step for participation of Rac in Nox activation by allowing translocation to the membrane and interaction with an effector (p67phox). This process raised two questions to which no fully satisfying answers were available. The first was the mechanism(s) of dissociation; the second was the nature of the signal(s) from phagocyte receptor stimulation to dissociation. Takai and coworkers showed for the first time that the combination of AA and nonhydrolysable GTP induced translocation of Rac to the membrane in a cell-free system derived from differentiated HL-60 cells and suggested dissociation of the Rac-RhoGDI complex as the likely mechanism [316]. In a similar cell-free system, consisting of neutrophil membranes and cytosol, Abo and coworkers showed that SDS caused translocation of Rac2 (but not RhoGDI) to the plasma membrane in the absence of added GTP [317]. This process also occurred in intact neutrophils stimulated with fMLP or PMA [317], as shown before by Quinn and coworkers [318]. Bokoch and coworkers took a direct approach by actually assessing the ability of some fatty acids and phospholipids to disrupt Rac-RhoGDI complexes; thus, AA and PA dissociated complexes at concentrations similar to those activating Nox in the cell-free system [310]. In retrospect, the actual data revealed only very partial dissociation of the complexes and the majority of results were based on an inhibitory effect of fatty acids and phospholipids (among them, phosphoinositides) on the ability of RhoGDI to prevent GDP dissociation from and GTP hydrolysis by Rac [310]. The suggested mechanism of dissociation was competition with the isoprenyl tail of Rac for binding to the C-terminal hydrophobic grove of RhoGDI. A possibly related mechanism was phosphoinositides partially “opening” but not fully dissociating RhoA-RhoGDI complexes, allowing targeting of the complex to the membrane, followed by GDP to GTP exchange on RhoA, full dissociation from RhoGDI, and attachment of RhoA to the membrane [319].

8.2.1

A Paradigm Shift: GEF Is Responsible for the Divorce Philips and coworkers were first to describe Rac-RhoGDI complex dissociation, resting solely on GDP to GTP exchange on Rac [320]. Bokoch and coworkers found that dissociation of Rac-RhoGDI complexes leading to translocation of Rac to the membrane depended on the presence of a membrane-associated GEF and nonhydrolysable GTP [300, 321]. In this model, some of the Rac-RhoGDI complex was bound to the membrane by affinity of Rac for GEF or for membrane phospholipids [310] and Rac was subjected to GDP to GTP exchange by GEF, resulting in dissociation of the complex and binding of Rac-GTP to the membrane.

31

However, the precise course of events was not fully understood; structures of complexed Rho proteins revealed that only one partner could be accommodated by the Rho protein, making binding of both RhoGDI and a GEF unlikely but not impossible [322] (reviewed in [323]). Pick and coworkers put forward two models of Rac-RhoGDI dissociation; the first centered on minimal requirements in an in vitro situation [324] whereas the second attempted to reproduce in vitro, events likely to occur in vivo [302]. A Rac1 (or 2)-RhoGDI complex was generated from recombinant Rac produced in E. coli, prenylated enzymatically, and recombinant RhoGDI. The complex dissociated by exposure to liposomes containing any of the four anionic phospholipids, phosphatidylinositol (PI), phosphatidylglycerol (PG), PS, or PA, and free Rac attached to the liposomes. Of significance was the finding that a Rac (GDP)-RhoGDI complex was able to elicit modest cell-free Nox activation by membranes enriched in anionic phospholipids in the presence of p67phox, which was enhanced by the participation of p47phox. This model suffered from the absence of a process leading to the exchange of GDP to GTP and from the fact that the enrichment in monovalent anionic phospholipids did not emulate a situation likely to occur in intact phagocytes. In a model mirroring in vivo events, Rac1(GDP)-RhoGDI complexes were exposed to liposomes, imitating the proportion of monovalent anionic phospholipids in the membrane of resting phagocytes but lacking membrane proteins, supplemented with PtdIns(3,4,5)P3, at a concentration likely present in the membrane of stimulated phagocytes [292, 304, 325]. In the presence of a dbl. family GEF and GTP, the complexes dissociated [302]. Extending this finding to phagocyte membranes, native membranes responded by Nox activation upon exposure to Rac1(GDP)-RhoGDI complexes, p67phox, GTP and a dbl. family GEF; enrichment of the membranes with PtdIns(3,4,5)P3 led to a significantly enhanced response [302]. An essential issue in this model was the double affinity of dbl. family GEFs for prenylated membrane-bound Rac, via the dbl. homology (DH) domain, and for PtdIns(3,4,5)P3 in the membrane, via the pleckstrin homology (PH) domain (see [322]). This ideal co-localization, helped by direct binding of PtdIns(3,4,5)P3 to Rac1 [326], assured efficient GDP to GTP exchange at membrane microdomains where membrane receptor-derived signals caused elevated PtdIns(3,4,5)P3 levels by stimulating a phosphoinositide 3-kinase (PI(3)K) [304].

8.2.2 An Unresolved Dissonance There is experimental evidence for cell-free Nox activation supported by a Rac1(GDP)-RhoGDI complex, in the absence of added GTP, backed by rigorous verification of the identity of Rac-bound nucleotide [327]. This study should be scrutinized together with an earlier one reporting Nox

32

E. Pick

activation supported by a Rac1-RhoGDI complex, in the absence of exogenous GTP, lacking, however, identification of the nucleotides bound to Rac [278]. Although the predominant evidence was that free Rac1 of recombinant origin [277, 299] or derived by dissociation of the Rac-RhoGDI complex [278] supported Nox activation only when in the GTP-bound form, Rac1-GDP expressed significant activity in the presence of high concentrations of p47phox and p67phox [328]. The dissonant data on the activity of Rac-GDP had little influence on the consensus opinion that Rac-GTP was the physiological activator of Nox and the contradictory results were likely the consequence of special in vitro conditions.

A Protein For all Seasons: p67phox Emerges as the Direct Nox2 Activator

9

As time requireth, a man of marvellous mirth and pastimes, and sometime of as sad gravity, as who say: a man for all seasons Robert Whittington, of Thomas More, Vulgaria (1521)

It was firmly established that all redox centers carrying electrons from NADPH to O2, were located in Nox2 (see Sect. 5). However, in the resting phagocyte, in vivo, and in the isolated membrane or purified cytochrome b558, derived from resting cells, in vitro, there was no electron flow in spite of the presence of substrate, cofactors, and O2. Thus, a central focus of research became the mechanism initiating the flow, representing the biochemical correlate of Nox2 activation. Since Nox2 activation was linked to translocation of cytosolic components to the plasma membrane (or its equivalent in the phagocytic vacuole), it was postulated that the “purpose” of translocation was for one or more of the components to establish direct contact with Nox2. The prominent candidate that emerged for such a role was p67phox (see Chap. 16 by H. Sumimoto).

9.1

Binding of Cytosolic Components to Nox2: More Questions than Answers

Binding of cytosolic components to Nox2 was mostly assessed by indirect methods and the field is ripe for a thorough review by novel approaches centered on binding proper. Focus was on binding of p47phox, based on the finding that Nox2 assembly was initiated by such a step [183] (see Chap. 15 by PM-C. Dang and J. El-Benna). The first binding site for p47phox proposed was the sequence 559–565, located at the C-terminus of Nox2, based on inhibition of Nox2 assembly by a peptide corresponding to these residues [329]. Definitive proof for binding was provided by the finding that in a mixture of neutrophil cytosol and Nox2

peptide 551–570, activated by SDS, the peptide was bound selectively to p47phox, as shown by chemical cross-linking [330]. This was followed by the identification of sequences 86–94, in the cytosolic loop B, and 451–458, in the DHR, as potential binding sites for p47phox, pivoted on screening p47phox with random-sequence peptide phage display libraries and on the ability of peptides comprising these residues to inhibit Nox2 activation in a cell-free system [331]. The three binding sites for p47phox, mentioned above, and an additional binding site (residues 494–498) were listed in Ref. [332]. The original description of the 494–498 sequence as a binding site for p47phox could not be traced but was compatible with reports of residues belonging to the insertion sequence (see Sect. 5.6), such as 491–504 and in particular D500 [181], and residues 484–500 [182] or 484–504 [180], participating in interaction with p47phox. However, in an in cellulo study, the deletion of residues 488–497, within the insertion sequence, had no effect on translocation of p47phox to the membrane and a peptide corresponding to residues 488–497 was ineffective in inhibiting Nox activation in a cell-free system [333]. The significance of residues 86–94 was brought into the limelight by the finding by Patrick Pagano and coworkers that a chimeric peptide, consisting of residues 86–94 linked to a 9-residues peptide derived from the HIV-coat protein (to make it cell permeable), decreased O2•- production ex vivo by mouse aortic rings treated with angiotensin II and lowered angiotensin II-induced blood pressure in mice, in vivo [334]. An alternative mechanism for the inhibitory action of peptide 86–94 could be its interference with the interaction of basic residues in loop B (R91 and R92) with the N-terminal half of the NADPH-binding domain, supposed to “pull” the FAD-binding domain toward the lower heme and promote electron transfer from FAD to heme [335]. Direct interaction between Rac and Nox2 was first proposed by Heyworth and coworkers [321] and claimed to be confirmed experimentally by Bokoch and coworkers on the basis of a minor increase in fluorescence upon binding of a fluorescent GTP derivative-labeled Rac2 to cytochrome b558 [336]. Next, a pull-down assay consisting of recombinant glutathione S-transferase (GST)-Nox2 DHR and Rac (1 or 2), led to the identification of residues 419–430 in Nox2 as a binding site for Rac [337]. This finding was supported by the inhibitory effect of a peptide, corresponding to this region, on Nox2 activation in a cell-free system and of a cell permeable form of the peptide in intact neutrophils, and by the loss of Nox activity and abrogation of Rac binding by mutating essential residues in the putative Rac binding sequence in Nox2 [336]. Binding of Rac to cytochrome b558 was found to be competed for by the Rho GTPase Cdc42 [338] in spite of the fact that Cdc42 was incapable of supporting Nox activation [339]. These results raised some questions because of the unexpected equal binding abilities of GTP- and GDP-bound

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

Rac and the mediation of binding by the Rac insert domain, a region the role of which in Nox activation was controversial [290, 328, 340, 341]. The conspicuous absence of any description of a binding site(s) for p67phox on Nox2 contrasted with the evidence for binding of p67phox to cytochrome b558. This was proven by overlay and affinity precipitation experiments [342] and was followed by the observation that binding to cytochrome b558 was enhanced by C-terminal truncation of p67phox [343]. The latter finding was the first indication that binding of p67phox to Nox2 might be dependent on the relief of autoinhibition, representing in retrospect a paradigm shift, the significance of which was not understood at the time of the discovery (see Sect. 11.7).

9.2

p67phox and p47phox: The Indispensable Activator and the Dispensable Organizer

The lack of information on a binding locus (or loci) for p67phox on Nox2 was the consequence of lack of interest for this issue from the part of the Nox research community. This was surprising in light of the evidence having arisen independently in the groups of Lambeth and Pick that Nox2 could be activated in a cell-free system by p67phox in association with Rac-GTP, in the absence of p47phox [344, 345]. An overlooked secondary finding was that a C-terminally truncated form of p67phox (at residue 246) exhibited an augmented ability to support p47phox-independent Nox2 activation [344]. p47phox-independent Nox activation was clearly an in vitro artefact, as shown by the requirement for concentrations of p67phox and Rac much higher than in the presence of all three cytosolic components and by the in vivo reality of the impairment of O2•- production by leukocytes of CGD patients with p47phox deficiency [346]. However, the facts that no other combination of two cytosolic components (p47phox + p67phox or p47phox + Rac) was capable of supporting Nox2 activation and that p67phox deficiency caused a more severe form of CGD than p47phox deficiency [347, 348] were strong arguments for attributing to p67phox a paramount role in Nox2 activation. Further support for the centrality of p67phox came from the proposal by Lambeth that membrane-bound Rac oriented p67phox for optimal interaction with cytochrome b558 [288]. The idea was experimentally applied by Pick and coworkers, who were aware of the key role of prenylation in tethering Rac to the membrane [286, 287], by designing a cell-free system, consisting of full-length p67phox or p67phox truncated at residue 212, and prenylated Rac1, lacking p47phox and not requiring an anionic amphiphile [289]. Turnover rates of O2•- generation similar to those seen in the canonical amphiphile and p47phox-dependent cell-free system

33

were obtained at concentrations of p67phox and prenylated Rac 10 to 20-fold lower than with the combination of p67phox and non-prenylated Rac. Fusion proteins (chimeras) between truncated p67phox (1–210 or 1–212) and Rac were found to be capable of activating Nox2 in a cell-free system in the presence of an anionic amphiphile, in the absence of p47phox, though higher concentrations of the chimeras were required than in the presence of p47phox [340, 349]. As found with the combination of p67phox and prenylated Rac, a [p67phox(1–212)-Rac1 (1–192)] chimera, prenylated on the Rac moiety, was a capable Nox2 activator in the absence of p47phox and anionic amphiphile [290, 350, 351]. The centrality of p67phox in Nox2 activation was reflected in the use of the name NOXA2 for p67phox, (suggesting its function as a Nox2 activator), and NOXA1, for the homolog acting on Nox1 and Nox3 [89, 352]. The appellation NOXO2 for p47phox, and NOXO1 for the homolog acting on Nox1 and Nox3, was also meaningful, being derived from the term “organizer” [89, 352]. This nomenclature implied that p67phox was the direct effector of the (still hypothetical and yet unproven) conformational change in Nox2, whereas p47phox acted as a co-carrier of p67phox to the membrane. This role of p47phox was based on its activation-dependent binding to p22phox and to PtdIns 3,4-bisphosphate (PtdIns(3,4)P2) in the lipid bilayer (see Sects. 11.1 and 11.2) as well as to Nox2 proper (see Sect. 9.1), and on the constitutive binding to p67phox, via the C-termini of the two proteins (reviewed in [250–252]).

9.3

The Unexpected Regulator of p67phox

It might appear strange to start this section with the statement that the finding that Rac functioned as a regulator of p67phox was unexpected and, based on previous knowledge about Rac function, unpredictable. This statement rests on the fact that around the time that the search for a protein interacting with Rac in Nox assembly was going on, there was a wealth of information on well-defined Rac effectors. None of these was p67phox. The best known was a family of serine/threonine kinases, known as p21-activated kinases (PAKs) (reviewed in [353]). PAK was first isolated from rat brain and described as a 65 kDa protein interacting with the GTP-bound forms of Rac1 and Cdc42 by the intermediary of a conserved motif at its N-terminus, known as p21-binding domain (PBD) or GTPase binding domain (GBD) [354]. Another member of the family was isolated from a mouse DNA library (mPAK3), with affinity for the GTP-bound forms of Rac1 and Cdc42 but not for RhoA and the GDP-bound forms of Rac1 and Cdc42 [355]. The PBD was highly conserved among members of the PAK family and within it there was a short motif, known as “Cdc42/Rac interactive binding” (CRIB), consisting of 14–16 residues which conferred binding to Rac

34

and Cdc42 [356]. A second protein that interacted with Rac, though with lower affinity than with Cdc42, was the WiskottAldrich syndrome protein (WASP), which also possessed a sequence motif similar to CRIB [357]. Interaction of Rac and Cdc42 with WASP protein, leading to activation, involved the relief of autoinhibitory contacts between the CRIB motif and the C-terminal region of the protein [358]. This was in accordance with the proposal that binding of Rac and Cdc42 to effectors possessing PBDs (GBDs) induced a structural rearrangement in CRIB, or that such effectors existed is several conformations, one of which was stabilized by binding to Rac or Cdc42 [359]. The fact that p67phox served as the effector component of Rac (1 or 2) in the assembly of the phagocyte Nox complex was established by Hall, Segal and coworkers in a study representing a major paradigm shift in Nox history [360]. By using a pull-down assay of Rac, exchanged to [α-32P]GTP, with GST-p67phox, binding of Rac1 to p67phox but not to 47phox was found. Binding was dependent on Rac being in the GTP-bound form and mutations in residues 35, 38, and 40, positioned in the switch I region, led to lack of interaction with p67phox, which was associated with a failure to support Nox2 activation. Yet another significant observation was that Rac was bound to p67phox at a site within the N-terminal 199 or 246 residues, a finding well correlated with the ability of p67phox truncated at residue 246 to support Nox2 activation in a cell-free system [361]. It is of interest that the authors of this crucial paper did not regard Rac as a carrier for p67phox but as acting following the translocation of p67phox to the membrane by a Rac-independent pathway. In a follow-up paper, Hall and coworkers found that, in addition to residues in the switch I region (residues 22–45), there was an additional C-terminal site (residues 143–175) in Rac involved in interaction with p67phox, the primary site being the switch I region, expected to undergo a conformational change upon GDP to GTP exchange [362]. A puzzling finding was that PAK bound to Rac at the same two sites in spite of the fact that there was no sequence similarity between p67phox and the PAK family and no PBD domain and CRIB motif were present in p67phox [360]. Competition between p67phox and PAK for Rac was demonstrated by the ability of recombinant PBD of mouse PAK (residues 65–137) [355] to inhibit Nox2 activation in a cell-free system; interestingly, a peptide corresponding to the CRIB motif (residues 68–85) was inactive [291, 363]. In an elegant study, exploiting the inability of Cdc42 to activate Nox2, Thomas Leto and coworkers found that this was due to different residues at positions 27 and 30; A27 and G30, found in Rac, conferred Nox activating ability, whereas K27 and S30, present in Cdc42, did not [364]. The relevance of these residues for the interaction of Rac1 with p67phox was also demonstrated by the finding that mutating A27 to K or G30 to S in the Rac1 moiety of a [p67phox(1–212)-Rac1

E. Pick

(1–192)] chimera caused loss of Nox2 activating property [350]. Thus, interaction between p67phox and Rac was required even in the presence of physical fusion between the two moieties in the chimeric construct. Further evidence for the involvement of switch I residues in the interaction of Rac with p67phox was produced by Lambeth and coworkers, based on the effect of mutations in Rac (including residue D38, also described by Diekmann et al. [360]) on cell-free Nox2 activation [365] and on Rac to p67phox binding [366]. A further mutational study, focused on residues T35 and D38 [367], and a report on the loss of Nox2 activating capacity by Rac1 glucosylated at residue T35 [368] rounded off the already massive evidence. The claim by Lim and coworkers that both Rac1 and Cdc42 bound to p67phox, based on dot blot assays, appears incompatible with the functional assays and with the absence of a CRIB motif in p67phox and was not confirmed [369]. A further claim by Lim that the binding site for Rac on p67phox was located at residues 170–199 was not supported by later work [370]. Following the preliminary delineation of the binding site for Rac on p67phox within residues 1–199 [360], a more complete definition of the site appeared in a seminal paper by Sumimoto and coworkers [371]. The binding region consisted of four tetratricopeptide (TPR) motifs (a 34 amino acids sequence), located within the 153 N-terminal residues of p67phox, but the isolated three or four TPRs did not bind Rac. Residue R102 in the third TPR was essential for binding [371] and CGD-causing mutations in p67phox, such as deletion of K58, located in the second TPR, exhibited reduced binding of Rac [372]. These findings established with certainty that binding of Rac to p67phox was correlated with Nox2 activation and the proposal [369] that other members of the Rho family, such as Cdc42, also bind to p67phox but lack activating ability, was found to lack foundation.

9.3.1

Your Old Road Is Rapidly Agin ‘. . . For the Times They Are A-Changin’ From the song “The Times they Are A-Changin” by Bob Dylan The uncertainty about the specific residues in both Rac and p67phox participating in binding required a change in approach, which was provided by precise structural studies, using X-ray crystallography. This was also made necessary by data showing low affinity of the Rac1—p67phox interaction (Kd values of 1.7–2.7 μM [373]) and the need of a marked excess of p67phox over Rac in order to evidence binding [360, 362, 374]. In a ground-breaking X ray diffraction-based analysis of crystals of a complex of p67phox(1–203) and Rac1-GTP(1–184), Katrin Rittinger and coworkers determined the structure of the complex (with electron density shown up to residue 186 in p67phox) [373]. The binding surface of p67phox consisted of the β hairpin insertion between TPRs 3 and 4 and the loops

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

connecting TPRs 1 with 2, and 2 with 3. Residues in p67phox involved in direct hydrogen bonding were S37, D67, H69, R102, N104, L106, and D108. The p67phox binding sites on Rac belonged to a region corresponding roughly to switch I and to a C-terminal region, as first described by Hall and coworkers [362], and were located close to each other in the structure of the protein. Residues in Rac involved in direct hydrogen bonding were S22, T25, N26, F28, G30, E31, A159, L160, and Q162. Interestingly, some mutations in p67phox causing CGD, such as deletion of K58 [372], affected residues not directly involved in bonding with a Rac residue. Also, Rac residues A27, T35, D38, and Y40, the mutations of which were shown to cause loss of Nox2 activating ability and binding to p67phox [360, 364], were not mapping to the interface with Rac. Eva Pebay-Peyroula and coworkers determined the structure of the isolated p67phox(1–213), with electron density shown to residue 193, exceeding by 10 residues the p67phox in complex with Rac-GTP, studied by Rittinger [375]. The only difference between the two structures was the presence of an α helix, comprising residues 187–193, not visible in p67phox(1–203) complexed with Rac-GTP. Considering the present view of Rac–p67phox interaction as leading to a conformational change in p67phox [350], this was disappointing since it was hoped that such change will be visible in p67phox in complex with Rac. Possible reasons for this not being observed were that residues beyond 186 were not visible in complexed p67phox or that residues 204–210, forming part of the “activation domain” (AD), shown to be essential for Nox2 activation [376], were lacking in the truncated p67phox(1–203) used to generate the complex. An important feature of p67phox was the presence of reversible intramolecular bonds between residues G113 and Q115 in the β hairpin and downstream residues R184 and K181 [373, 375]. There is recent evidence that dissociation of these bonds mediated Rac-GTP-dependent conformational changes in p67phox, mandatory for interaction with Nox2 [377] (see Sect. 11.7). Rac1 was also required for the activation of Nox1 and, possibly, Nox3 [352]. Activation of Nox1 was dependent on interaction with NOXA1, and similarly to the Rac – p67phox interaction, Rac facilitated the membrane attachment of NOXA1 and, likely, induced a conformational change in NOXA1 [378]. Binding of Rac to NOXA1 resembled that of Rac to p67phox, as shown by the fact that NOXA1 with a R103A mutation (the equivalent of the R102E mutation in p67phox [371]) was incapable of activating Nox1. It is surprising that in spite of the in depth knowledge of the molecular mechanism of Rac – p67phox interaction, this was scarcely exploited for the design of Nox2 inhibitors for potential therapeutic use. Pick and coworkers found that a p67phox peptide corresponding to residues 106–120, thus, overlapping the β hairpin and containing the direct Rac-binding residues L106 and D108 [373], prevented

35

Nox2 activation in a cell-free system but the peptide was not explored for Nox2 inhibition in whole cells and organs [377]. An inhibitor of Nox2 activation was identified by Zheng and coworkers by carrying out a structure-based virtual screen of small molecules capable of interaction with the Rac1-binding interface of p67phox, bordered by residues R38 and R102, facing Rac1 residues T25, N26, and A30 [379]. The top hit compound, termed Phox-I1, abrogated the binding of Rac1 to p67phox and impeded Nox2 activation in a cell-free system and in murine and human neutrophils. The lack of emphasis on this direction of research in the design of Nox2-specific inhibitors is surprising considering the specificity of the Rac–p67phox interaction for Nox2 and the frequency of pathologies in which Nox2 is an etiologic or accessory agent [380].

10

An Induced Conformational Change in Nox2 Affects the Electron Flow: How and Where?

Cytosolic components are likely to have access only to the cytosolic DHR and to loops B and D of Nox2. The dominant hypothesis is that binding of one or more of these (the favored candidate being p67phox) cause(s) a conformational change(s) in the DHR of Nox2, accommodating the multi-site NADPH- and FAD-binding domains. The completion of writing this chapter coincided with solving the structure of the membranal domain of human Nox2 by cryo-EM, with only limited informartion being available on the structure of the cytosolic DHR (see Sect. 12). The crystal structure, based on X-ray diffraction, of part of the human Nox2 DHR (residues 385–570), comprising the NADPH-binding but not the FAD-binding domain, is known (unpublished; see Honbou K, Noda NN, Sumimoto H, Inagaki F (2009), PBD https://www.ebi.ac.uk/pdbe/entry/pdb/3A1F). 3A1F; Detailed structures are available for csNox5 (based on X-ray crystallography) [130]) and for the mouse [381] and human [382] DUOX1-DUOXA1 complexes (based on cryoEM) (see Chap. 30 by J-X Wu, J. Sun, and L. Chen). Nox5 and DUOX1 are regulated by Ca2+, are not complexed with p22phox, and none of the cytosolic components interacting with Nox2 participate in their activation; thus, they can serve only as partial models for Nox2. Further potential structural models for Nox2 are FNR [175, 176, 185], for the DHR, and the family of STEAP proteins [157], having evolved from a ferric reductase ancestor [156], for the transmembrane domain. Several proposals were put forward for ways in which a conformational change can facilitate the electron flow from NADPH to O2 along a multi-center redox gradient. One way in which this could occur was to make a previously inaccessible binding site for the NADPH substrate or the FAD

36

cofactor accessible. Taylor et al. [179] were the first to propose that an insertion sequence corresponding to residues 484–504, unique for Nox2 in the FNR superfamily, prevented access to the NADPH binding domain and that binding of a cytosolic component to the insertion sequence caused its displacement and allowed binding of NADPH. As discussed in Sect. 5.6, this proposal, although controversial, was presented as a possible mechanism in recent publications [180, 184] in spite of the finding that deletion of the insertion sequence did not result in constitutive activity of Nox2 [180]. Support for binding of NADPH representing the Nox activating element was also supplied by solving the structure of csNox5; regions in the DHR binding the mediators of Ca2+-dependent activation, the EF-hand and calmodulin, were in the proximity of the NADPH binding domain and might regulate binding of NADPH [130]. As far as Nox2 was concerned, labeling studies with NADPH analogues showed that cytochrome b558 of resting and ROS producing neutrophils was labeled with equal efficiency, indicating that enhanced affinity for NADPH was unlikely to be the initiator of Nox2 activation (see Sect. 5.2 and [124, 162, 163]). A model, focused on FAD, was proposed by Hirotada Fujii and coworkers, who reported that in a cell-free system, FAD was in an equilibrium between the free and Nox2bound state before the addition of the activating amphiphile but became tightly bound upon Nox activation by the amphiphile and was further stabilized by the binding of cytosolic components [172]. So far, no confirmatory evidence for such a process was published. A dominant hypothesis was that the conformational change resulted in the promotion of hydride transfer from NADPH to FAD. Cross and coworkers were the first to propose that a diaphorase activity, assessed by the reduction of a tetrazolium dye, was independent of p47phox, indicating that hydride transfer from NADPH to FAD was regulated exclusively by p67phox, the likely two-electrons donor to the dye being reduced FAD [383]. Further experimental proof for distinct roles of p67phox and p47phox was published by the same group, also comprising evidence for p47phox being required for electron transfer from reduced FAD to the hemes and, consequently, to O2 [384]. It should, however, be noted that the kinetics of electron flow in a cell-free system, consisting of purified cytochrome b558 and recombinant cytosolic components, were found identical whether p47phox was present or absent [187]. A major paradigm shift in Nox2 research was the finding by Lambeth and coworkers that a domain in p67phox, spanning residues 199–210, known as AD, was essential for Nox2 activation and was suggested to interact directly with Nox2 [376] (see Chap. 3 by J.D. Lambeth). The Discussion section of the paper contained one of the first enunciations of the concept of p47phox and Rac binding to and orienting p67phox

E. Pick

in a way to bring the AD in contact with a specific region in Nox2, responsible for initiating the electron flow. The relevant region in Nox2 was not identified and no proof was offered for the AD being directly engaged in protein–protein interaction with Nox2. Next, the use of sophisticated steady state kinetics of flavin reduction and kinetic deuterium isotope methodology led to the conclusion that p67phox regulated the electron flow from NADPH to FAD. The essential role of the AD in this process was proven by the fact that mutations in the AD lessening Nox2 activation also diminished steady state reduction of flavin [385]. It is remarkable that in the simplest model of cell-free Nox2 activation by anionic phospholipids and FAD, not involving cytosolic components, the dependence of activation on oxygen was an intrinsic property of Nox2 and due to the acceleration of FAD reduction [93]. Mutations in residues D484 and D500, within the Nox2 insertion sequence, were also suggested to interfere with the NADPH to FAD hydride transfer but the mediation of this effect by interference with making the NADPH binding site accessible could not be excluded [182]. Hydride transfer from NADPH to FAD was also defective in a PLB-985 cell line reproducing X91+ CGD caused by the deletion of residues 488–497 (within the insertion sequence) [333]. A much less popular proposal was that a conformational change in Nox2 facilitated electron flow from FAD to heme. At a time when the binary model of Nox was dominant, Kakinuma [46] suggested that in the plasma membrane of stimulated leukocytes, the flavin moved to the proximity of a transition metal, represented by the heme iron of cytochrome b558. A much later version of such a model was that a polybasic region in the B loop of Nox4 and Nox2, connecting transmembrane helices 2 and 3, interacted with residues in the N-terminal half of the NADPH-binding domain of the DHR [335]. This splendidly conceived and executed work by Lambeth and coworkers represented a significant paradigm shift, echoed by future structural data [130], by proposing that this interaction brought the FAD-binding domain of the DHR in close contact with the heme-binding transmembrane region and promoted electron transfer from FAD to the lower heme. p47phox was found to bind to the B loop of Nox2 (Sect. 9.1 and Ref. [331]) and mutations in the B loop that interfered with binding of p47phox also prevented interaction with the DHR. This could have meant that p47phox boosted the B loop—DHR interaction but a trial to prove this experimentally failed [335]. The fact that the B loop—DHR interaction was found in Nox4, the activity of which was not dependent on cytosolic components, suggested that it represented a static structural feature of Nox proteins, and the higher affinity of binding found in Nox4 might be one of the reasons for its constitutive activity. A role similar to that of loop B was attributed to loop D of Nox2 (connecting transmembrane helices 4 and 5), based on the finding that several mutants

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

of residues in the D loop caused loss of Nox activity in the presence of unchanged complex assembly and NADPH to FAD electron transfer [182]. On the contrary, expression of a [Nox4 D loop-Nox2] chimera in PLB-985 cells led to markedly enhanced Nox activity [182, 386]. The proximity of loop D to transmembrane helix 5, harboring two hemecoordinating histidines, was compatible with its participation in electron transfer from FAD to hemes. These functional studies gained support by the description of the extensive interdomain interactions involving residues in the C-terminal part of the DHR and loops B and D in the crystal structure of csNox5 [130]. Although the structural data are consistent with interactions of loops B and D with DHR contributing to Nox activity, direct evidence for their participation in the regulation of FAD to heme electron flow by cytosolic components in Nox2, is lacking. Resting on the elucidation of the structure of csNox5, Mattevi and coworkers proposed a Ca2+-initiated conformational change in the DHR, entailing binding of EF-hands and calmodulin to the C-terminal part of DHR, promoting the binding of NADPH and fostering electron transfer at the FAD–heme interface, involving loop D [130].

10.1

The Expected Paradigm Shift: A Conformational Change in Nox2 Affects Distances Between Redox Stations

I should be excused for digressing briefly from the description of past paradigm shifts to predicting a future occurrence. The fast-paced progress in solving the structures of individual proteins, protein–protein interactions, and assembled multimolecular complexes by X-ray crystallography and cryo-EM is very likely to lead soon to solving the structural conversion of Noxs from the resting to the activated state (see comment on the paradigm shift in structural biology [387]). In the case of Nox2, hydride transfer from NADPH to FAD and one-electron transfer from reduced FAD to the lower heme, are the two steps most likely to be regulated by cytosolic components. Electron flow from the lower to the upper heme as well as from the upper heme to dioxygen are unlikely targets for modulation due to the lack of direct accessibility to cytosolic components. The dominant concept is that distance between, and the geometry of the stacking of donor and acceptor, are the principal parameters that set the rate of electron transfer in proteins. However, two additional factors also participate; these are the “driving force”, represented by the relative redox potentials of the donor and acceptor, and the barrier, consisting of the insulating protein medium between the members of the pair. The distance dependence of the rate of electron transfer is roughly exponential, meaning that small

37

changes in distance result in large changes in the rate. Most of these parameters and their nomenclature are derived from the concept of “electron tunneling” (reviewed in [388, 389]). A thoroughly simplistic definition of this is that electrons get from one place to another by “tunneling” through the protein medium that separates one redox center from another. The “physics” behind electron tunneling are beyond the expected knowledge of the average biomedical investigator and require familiarity with quantum mechanics [390]. Although electron tunneling was suggested to apply to oxidoreductases [389], no publication reporting experimental data on a tunneling mechanism pertaining to Noxs was encountered. The overall redox potentials of the electron transfer stations in Nox2 are as follows (reviewed in [391]): NADPH=NADPþ FAD=FADH2 Nox2 Fe3þ =Nox2 Fe2þ O2 =O2 • - 320 mV - 280 mV - 245 mV - 160 mV

The -280 mV redox potential of the FAD/FADH2 couple is the overall value derived from the -304 mV value for the FADH2/FAD semiquinone couple and the -256 mV value for the FAD semiquinone/FAD couple. The -245 mV redox potential of the Nox2 Fe3+/Nox2 Fe2+ couple is the overall value derived from the -265 mV value for the lower heme and the -225 mV value for the upper heme. A more detailed rendition of the electron flow appears in Fig. 1.5. In the absence of an atomic structure of Nox2, the distances between the redox stations in the resting state are unknown and any activation-related changes are likely to become known when the structure of the assembled Nox2 complex will be solved. It might be useful to look at numerical data on distances between redox centers in Noxs other than Nox2 and in related oxidoreductases. A distance of 19.8 Å (Fe to Fe) and 6.4 Å (vinyl to vinyl) was measured between the two hemes in csNox5 but no information is available on NADPH to FAD and FAD to lower heme distances [130]. Preliminary work indicated a distance of 10 Å between NADPH and FAD in a human Nox5 structure derived by cryo-EM (Ji Sun, personal communication). A computer simulation of electron transfer in csNox5, quasiembedded in a lipid bilayer, with emphasis on heme to heme transfer, was published [392]. To the best of my knowledge, this was the first occasion in which a tunneling pathway was suggested to be involved in electron transfer in a Nox molecule, with the caveat that this was a modeling study. In a cryo-EM study of mouse Duox1, the distance between NADPH and FAD was 7.8 Å but since the electron density of nicotinamide was insufficient, the true distance was deemed to be shorter. The distance between FAD and the lower heme was 4.1 Å, and the lower heme to upper heme distances were 20.6 Å (Fe to Fe) and 8.9 Å (shortest interatomic distance) [381]. It is of interest that in mouse Duox1, a lipid molecule

38

E. Pick

Fig. 1.5 A contemporary scheme of the electron transfer pathways and mid-point potentials in cytochrome b558, from NADPH to O2•-. INT stands for iodonitrotetrazolium, used to assess the diaphorase activity [383] (reproduced from Ref. [391])

was detected, mediating the interaction between the transmembrane region and the phospho-ADP-ribose part of NADPH, the diphosphate group contacting the NADPHbinding domain on the DHR [381]. Phospholipid alkyl chains were also found along the helices of the transmembrane region of csNox5 [130]. The association of lipids with transmembrane regions of Noxs might be related to the absolute phospholipid dependence of solubilized phagocyte Nox and purified cytochrome b558 for activity in vitro [104, 246]. Not knowing the identity of the lipid, a PC molecule was modeled successfully into the cryo-EM of mouse Duox1. PC, with long-chain unsaturated fatty acids, was also the phospholipid most effective in relipidating phagocyte Nox preparations following removal of the native membrane lipids [246]. Detailed data on distances between redox centers were derived from a cryo-EM of human Duox1 and were similar to those reported for mouse Duox1; thus, the distance between NADPH and FAD was 8.2 Å, between FAD and the lower heme was 3.9 Å, and between the two hemes, 6.7 Å [382]. The possibility that lipids in the membrane bilayer might affect Duox1 structure and influence electron transfer was alluded to but no data were presented. On examining the procedures used to prepare samples for X-ray crystallography and cryo-EM, it appears that these varied widely. The transmembrane region of csNox5 was mixed with 1-oleoyl-racglycerol [130]; human Duox1 was assembled into peptidiscs [382, 393], and mouse Duox1 was used unmodified following purification [381]. It is certain that future structural endeavors will make use of the ever increasing variety of ways to purify Nox proteins mimicking the amphipathic

environment of a lipid bilayer while maintaining their structure in a physiologically relevant state. The main reason for lack of progress in identifying the key activation-related change in electron transfer resulting from a conformational change in Nox2 was the lack of pre-activation versus post-activation data originating in other Noxs, for which partial distance data between redox centers were known (Nox5 and Duox1). The dominant hypothesis that p67phox was the direct inducer of a conformational change in Nox2 (see Sect. 9) and the data attributing to p67phox the property of activating the hydride transfer from NADPH to FAD [383–385] made a shortening of the distance between bound NADPH and FAD an attractive mechanism. Unfortunately, the NADPH–FAD distance for csNox5 (the DHR of which was thought to have been crystallized in activated state) was not known [130], and distances for human Nox5 (Ji Sun, personal communication) and mouse [381] and human [382] Duox1 were in the 8–10 Å range, most likely in the resting state. Thus, in order to assess the NADPH–FAD distance in an active Nox-like oxidoreductase, preferably Nox2, its structure in the presence of NADPH and FAD and in complex with its positive modulators (individual cytosolic components or chimeric constructs) has to be “caught in action”. In the absence of data derived from Nox studies, a look at the Nox-parental FNR superfamily is worthwhile. In plant FNRs, the aromatic side chain of a C-terminal tyrosine occupies a site adjacent to the isoalloxazine N5 atom, to which the C4 atom of the nicotinamide ring of NADPH would stack in the course of productive NADPH – FAD binding, resulting in FAD reduction [176, 185]. In two

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

studies dealing with pea FNR [185] and corn root FNR [176], the C-terminal tyrosines were mutated to nonaromatic residues in order to prevent tyrosine from blocking the binding of the nicotinamide ring of NADPH to the tricyclic isoalloxazine ring of FAD. Mutant FNRs were capable of binding NADP(H) in a manner that assured the proximity of the C4 atom of nicotinamide and the N5 atom of the isoalloxazine, enabling hydride transfer. The distance between the C4 atom and the N5 atom in this productive mode was found to be 3 Å. A situation similar to that existing in FNRs was described in NADPH-cytochrome P450 reductase, in which the indole ring of the penultimate residue, tryptophan 677, is stacked to the isoalloxazine of FAD, preventing access of NADPH to FAD. Displacement of W677 by the nicotinamide ring enabled direct hydride transfer and catalysis [394, 395]. It is, therefore, to be expected that, assuming that Nox2 activation is linked to the initiation of hydride transfer from NADPH to FAD and FAD reduction [383– 385], a distance of 3 Å or less between NADPH and FAD should be found in activated Nox2.

10.2

Is the C-terminal F570 Involved in Nox2 Activation?

In light of the importance of C-terminal residues with an aromatic side chain in the function of FNRs [176, 185] and NADPH-cytochrome P450 reductase [394, 395], the question of whether the C-terminal F570 in Nox2 plays a similar role is legitimate. This question was raised in discussing the mechanism by which a Glu568Lys mutation, neighboring F570, affected Nox2 function [79], analyzing the role of the C-terminal F693 in csNox5 [130, 184], and attempting to extrapolate the aromatic placeholder model of plant FNRs to the Nox superfamily [176, 184]. What was missing in all these proposals was the link between the commonly accepted Nox2 activating mechanism (involving p67phox) and providing a competitive advantage to the nicotinamide ring of NADPH over F570 for binding to the isoalloxazine of FAD. In the FNR model, the occupancy of the nicotinamide binding pocket on the isoalloxazine by the tyrosine was seen as a dynamic process, with the tyrosine moving in and out, in the absence of NADPH; upon binding of NADPH (via the 2′-phospho-AMP part), the nicotinamide moved into the pocket, when the tyrosine was out (personal communication by P. A. Karplus). To the best of my knowledge, the only experimental approach to investigating the role of the C-terminus of Nox2 was by Dinauer and coworkers, who replaced the C-terminal F570 with A or deleted it and assessed O2•- production by induced myeloid leukemia cells, expressing wild type or mutant Nox2 [396]. Deletion or mutation of F570 resulted in only a two-fold reduction of O2•- production. Mutating the C-terminal Phe693 in csNox5

39

led to a two-fold increase in activity whereas its deletion had no effect but these results were derived from work with the isolated DHR and might not be representative of the fulllength protein [130]. These results contrasted with the 300-fold and 850-fold reduction of pea FNR activity by replacing the C-terminal tyrosine by glycine or its elimination [397]. The reason for the marked difference in the role of C-terminal aromatic residues in the activity of Noxs and FNRs is yet unexplained.

11

A Major Paradigm Shift: Relief of Autoinhibition as the Mechanism of Activation of Nox2 Organizer and Activator

11.1

Awakening the Dormant p47phox

The function of the tandem SH3 in p47phox was not addressed in the publications reporting the cloning of the gene [267, 270] (see Sect. 7.2). The literature on the structure of nonreceptor protein tyrosine kinases, characterized by the presence of SH3 domains, listed examples of autoinhibitory mechanisms [398]. In early 1994 Sumimoto and coworkers [143] and Leto and coworkers [144] were, independently, the originators of a major paradigm shift in our understanding of an essential step in the mechanism of Nox2 activation. This connected the existing information on the role of p47phox phosphorylation in the intact cell [192, 193, 262], the accumulating evidence that a PKC had a key role in signal transduction from certain membrane receptors to Nox2 (reviewed in [253, 399]), and cell-free Nox2 activation by anionic amphiphiles [197]. Both groups presented evidence for the SH3 tandem in p47phox being linked to an autoinhibitory region (AIR), maintaining p47phox in an inactive state. Anionic amphiphiles disconnected the SH3 tandem from the AIR and allowed interaction with a proline rich region (PRR) positioned in the cytosolic tail of p22phox. SH3 domains exhibited affinity for PRRs adopting a left-handed polyproline-II conformation, such as present at residues 151–160 in p22phox (reviewed in [400]). Disengagement of the SH3 from the AIR led to translocation of p47phox to the membrane and binding to p22phox. The ultimate result of this process was for the translocated p47phox to act as a carrier for p67phox to Nox2 by virtue of the binding of the C-terminal SH3 of p67phox to a C-terminal PRR in p47phox [401]. A Pro156Gln mutation in p22phox caused X91+ CGD, with normal expression of the cytochrome b558 dimer but lack of Nox activity in the cell-free system [142] and defective translocation of p47phox and p67phox [402]. With admirable foresight, Sumimoto suggested that, in vivo, the process leading to the relief of autoinhibition was p47phox phosphorylation, mimicking the

40

action of the anionic amphiphiles in the cell-free system. A peptide corresponding to p22phox residues 149–162 (KQPPSNPPPRPPAE) prevented binding of the p47phox SH3 tandem to the cytosolic tail of p22phox. Sumimoto proposed that, in the resting state, the p47phox SH3 tandem was bound to the autologous PRR at the C-terminus (residues 361–369) [143], and Leto proposed that the N-terminal SH3 was bound to the PRR region at the C-terminus, and the C-terminal SH3, to residues 73–80 in the Phox homology (PX) domain, comprising three prolines [144]. Further exploration of the subject pointed to W193, in the N-terminal SH3 of p47phox, as the key residue in the interaction with the PRR in p22phox [403]. A major advance in understanding the autoinhibition and its relief was the finding that the SH3 tandem was engaged in an intramolecular interaction with an arginine/lysine rich polybasic region, comprising residues 286–340, which lacked the canonical polyproline type II helix but contained a 299PPRR302 sequence, in which P299 and P300 were suggested to be responsible for binding [404]. Region 286–340 also contained three serines (S303, S304, and S328), which became phosphorylated upon Nox activation in the intact cell [405]. Phosphorylation of the three serines or replacement with the acidic residues aspartate or glutamate resulted in the disruption of the intramolecular bond and binding of p47phox to p22phox and consequent Nox2 activation [404, 406]. The unconventional nature of the intramolecular bond between the SH3 tandem and the polybasic region (lack of participation of a canonical PRR) motivated a reexamination based on the crystal structure of p47phox in the autoinhibited and activated forms [407]. This showed that the two SH3 domains formed a groove that was occupied by the N-terminal segment of the polybasic region, with residues 297 GAPPR301 adopting a polyproline type II helix, as proposed by Sumimoto [404]. An important finding was that residues in the C-terminal segment of the polybasic region also interacted with the SH3 tandem, a fact likely related to the role of serines 303, 304 and 328, located outside and at the border of the 297GAPPR301 helix [407]. The study showed that phosphorylation of all three serines caused structural changes that markedly disturbed the binding of residues outside the 297GAPPR301 helix to residues in the SH3 tandem and allowed p22phox to compete successfully for binding to the SH3 tandem [407]. The X-ray structure also revealed the details of the key interactions between W193 in the N-terminal SH3 and residues P152, P156, and N154 in the PRR of p22phox. Molecular dynamics simulation demonstrated the reversibility of the activation process; thus, dephosphorylation of the serines caused re-closure of the opened intramolecular bonds and a return to the resting state [408].

E. Pick

11.2

p47phox Acquires a Taste for Lipids by Abandoning an Autoinhibitory “Diet”

The PX domain of p47phox (residues 1–121) binds specifically to PtdIns(3,4)P2 and to PA or PS, by the intermediary of two distinct “pockets”, as revealed by X-ray diffraction [409]. Williams and coworkers found that interaction of the PX domain with the phospholipids was prevented by an intramolecular association with the C-terminal SH3 [409]. The PX domain contained a short polyproline motif consisting of residues 73PHLP76, which was part of a type I SH3-interacting sequence (RXqPXqP) (R is arginine; X represents any residue; q specifies proline or another hydrophobic residue) [400]. The C-terminal SH3 interacted with the polyproline motif in the PX domain, preventing the binding of p47phox to PtdIns(3,4)P2 [410]. A Trp263Arg mutation in the C-terminal SH3 led to binding of p47phox to PtdIns (3,4)P2, indicating that W263 participated in the autoinhibitory bond between SH3 and the PX domain, which was cut as a consequence of the mutation [409, 410]. Phosphorylation of serines in the polybasic region [411] or mimicry of serine phosphorylation by replacing serines by glutamates [410] or aspartates [411] caused the release of the PX domain from autoinhibition. Marcoux and coworkers suggested a different model in which the PX domain was anchored to the N-terminal SH3 residue R162 in synergy with D166; upon dissociation of the polybasic region from the SH3 tandem, destructuration of the N-terminal SH3 - PX domain interface caused release of the PX domain and binding to PtdIns(3,4)P2 [412]. The p47phox homologue NOXO1 is structurally similar to p47phox but lacks the polybasic AIR (corresponding to residues 292–340 in p47phox) and the PKC-targeted serines [413, 414]. Thus, the SH3 tandem is free to interact with p22phox and NOXO1 is, therefore, constitutively capable to cooperate with NOXA1 and to induce ROS production by Nox1. Surprisingly, an intramolecular autoinhibitory bond was identified in human NOXO1 between the SH3 tandem and the canonical PRR at the C-terminus, the disruption of which enhanced the constitutive activity [415, 416].

11.3

Targeting Nox2 Activation by p22phox Peptides

An attempt to map domains in p22phox participating in Nox activation by “peptide walking” [417], a method resting on inhibition of cell-free Nox activation by overlapping peptides (in the particular case, p22phox peptides) gave disappointing results [418]. None of the regions identified were involved in p22phox-related Nox2 activation and, paradoxically, peptides corresponding to residues 151–160, participating in interaction with p47phox, were not inhibitory. This result was in

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

agreement with the failure of p22phox peptide 149–162 to exert an inhibitory effect in a cell-free assay, reported by another group [402]. p22phox peptides were also tested for binding p47phox in an assay in which the peptides were surface-attached and the protein was in the fluid phase [419, 420]. p47phox was found to bind to a cluster of peptides centered on residues 151–160, but, surprisingly, binding was not dependent on disengagement of the SH3 tandem from the polybasic domain [418]. The critical influence of methodological choices was demonstrated by experiments in which binding of a fluorescent p22phox peptide, corresponding to residues 151–165, to p47phox was assessed when both components were in solution. Under these conditions, the peptide was bound exclusively to p47phox truncated at residue 286 (lacking the AIR) and a Trp193Arg mutant of p47phox(1–286) did not bind the peptide (E. Pick, unpublished results). Recently, fragment-based drug discovery identified small molecular compounds exhibiting high affinity to the p47phox SH3 tandem [421]. By screening 2500 fragments, 8 hits were found and optimization led to a dimeric compound which inhibited the p22phox–p47phox interaction with a Ki of 20 μM. In a different approach, an undecapeptide mimetic of the p22phox sequence 151–161 was synthesized, in which the central 155PPP157 trio was replaced by the synthetic pseudotriproline mimetic 155Pro-Pro- cyclopentanecarboxamide (Cyp)157 [422]. The synthetic triproline mimetic exhibited submicromolar in vitro activity in preventing binding of p47phox to p22phox.

11.4

Mimicking PKC Action In Vitro: An Alternative to Amphiphile-Dependent Cell-Free NADPH Oxidase Activation

Motivated by the voluminous evidence on the role of PKC-mediated phosphorylation in the activation of Nox (reviewed in [253]), attempts were made to activate Nox by PKC in vitro by the groups of Tauber and of Babior. Tauber and coworkers made use of the fact that PMA can replace 1,2-diacylglycerol (DAG) as a direct activator of PKC and reported that PMA elicited Nox2 activation in a mixture of neutrophil membranes, cytosol, PS, ATP and NADPH, in the presence of Ca2+ [423]. Purified PKC could replace cytosol and the authors suggested that PKC acted on a regulatory component, a more than fair prediction of specific PKC isoforms acting on p47phox [424]. A subsequent study showed that the PMA-activated cell-free system was essentially different from the canonical AA- [194] and SDS- [233] activated systems by the fact that the latter two did not involve PKC activation, as shown by the independence of PKC cofactors and the resistance to a PKC inhibitor [425]. Activation of PKC by PMA or DAG causes

41

proteolysis of PKC resulting in cleavage into a regulatory and a catalytic domain, the latter being active in the absence of cofactors. The catalytic segment of PKC activated Nox2 in neutrophil membranes in the absence of PMA and Ca2+ but, unexpectedly, PS was still necessary [426]. A decade after the publication of the work of Tauber, a different PKC-activated cell-free system was described by Babior and coworkers. The activation mixture consisted of neutrophil membranes, cytosol, recombinant p47phox, purified PKC, DAG, PS, Ca2+, GTP, and ATP, and, following brief incubation, the catalytic step commenced by the addition of NADPH [427]. Nox2 activation was dependent on ATP and prevented by two PKC inhibitors. These results made the claim by Tauber that neutrophil cytosol could be replaced by PKC [423] untenable since cytosol was also expected to contribute the essential cytosolic components, most notably the PKC target, p47phox. It was next found that Nox2 could be activated by p47phox pre-phosphorylated by PKC and separated from PKC [428]. A C-terminally truncated form of p47phox, missing residues 348–390, did not support PKC-elicited Nox2 activation in a cell-free system comprising all the other components (purified PKC, cytosol, membrane, DAG, PS, Ca2+, and ATP), consistent with the missing serines being required targets for phosphorylation by PKC [429]. The deleted residues comprised Ser379, the phosphorylation of which was essential for Nox2 activation [430, 431]. More compelling proof for PKC-dependent Nox2 activation being mediated by phosphorylation of serine residues in p47phox was offered by the ability of a semi-recombinant system, consisting of neutrophil membrane, recombinant p67phox, Rac-GTP, and p47phox phosphorylated by purified PKC and separated from the kinase, to generate O2•- upon addition of NADPH [432]. Nox2 activation in the PKC-dependent systems were always found of inferior magnitude compared to anionic amphiphile-dependent activation, possibly due to the fact that a more complete p47phox phosphorylation required the participation of kinases other than PKC, such as mitogenactivated protein (MAP) kinase (discussed in [433]).

11.5

Physico-Chemical Evidence for an Induced Conformational Change in p47phox

The idea that disengagement of the SH3 tandem from the C-terminal AIR in p47phox resulted in a conformational change was substantiated by an innovative approach by Quinn and coworkers [434]. Exposure of p47phox to the anionic amphiphiles, AA or SDS, at cell-free Nox2 activating concentrations, caused dose-dependent quenching of intrinsic tryptophan fluorescence and it was significant that five of the seven tryptophans were located in the SH3 tandem.

42

E. Pick

Recombinant p47phox, phosphorylated by PKC in the presence of PS, DAG, Ca2+, and ATP, exhibited a significant reduction of tryptophan fluorescence in comparison to nonphosphorylated p47phox. A change in the circular dichroism spectrum of p47phox exposed to SDS was also recorded. Phosphorylation of p47phox by PKC or exposure to AA resulted in reduced accessibility of Cys378 to the alkylating agent N-ethylmaleimide, representing a different expression of a conformational change [435]. The location of Cys378 next to Ser379, known to be an important target of p47phox phosphorylation [430], was significant. In another study, all four cysteines in p47phox (including Cys378) were labeled covalently with a thiol-reactive fluorescent probe and exposed to anionic amphiphiles (AA, other long-chain unsaturated fatty acids, SDS) at concentrations optimal for cellfree Nox2 activation. This resulted in an increase in fluorescence compatible with lesser exposure to the protein surface [436]. In a more recent study, a contradictory result was obtained, expressed in enhanced accessibility of all four cysteines in p47phox in response to AA, with the caveat that measurements were performed in the presence of p67phox [437]. Changes in the circular dichroism spectrum of p47phox exposed to AA resembled the pattern seen with p47phox exposed to SDS, as described by Quinn [434]. Searching for an explanation for the less effective Nox2 activation by phosphorylation of p47phox, compared to that by amphiphiles, Sumimoto suggested that phosphorylation of p47phox caused an incomplete severance of the intramolecular bond, requiring the participation of an amphiphile, such as AA, at concentrations lower than those seen with AA acting alone, for a stronger effect [438]. He concluded that in vivo, AA and phosphorylation act synergistically to cause a conformational change in p47phox. Yet another example for the independent birth of paradigms is that Lambeth and coworkers described a decade earlier synergism between the PKC activator DAG, and anionic amphiphiles in causing Nox2 activation and p47phox phosphorylation [439].

11.6

Alternative Pathways of In Vitro Nox2 Activation

In the preceding text we focused on the role of p47phox phosphorylation and its mimicry by anionic amphiphiles in breaking intramolecular autoinhibitory bonds in p47phox, followed by p47phox carrying p67phox to the proximity of Nox2. In accordance with the proverb “All roads lead to Rome” and with p67phox representing Rome, we shall discuss alternative in vitro models of p67phox-mediated Nox2 activation. The fact that p67phox has an overall negative charge and lacks elements with affinity for specific membrane phospholipids (such as the PX domains present in p47phox and p40phox) makes its translocation to the membrane

dependent on the carriers, p47phox, Rac-GTP, or p40phox. The in vitro models to be discussed are centered on different ways to achieve productive interaction between p67phox and Nox2, resulting in Nox2 activation. The reductionist character of the approaches is evident in the narrowing down of variables and the precise definition of experimental parameters, expected to allow a more rigorous understanding of Nox2 activation in the whole cell.

11.6.1 Truncation of Cytosolic Components In a pioneering study, Sumimoto and coworkers found that combining a C-terminally truncated p47phox(1–286) with C-terminally truncated p67phox(1–242) in a cell-free system containing membrane, Rac-GTP, and NADPH, resulted in O2•- generation in the absence of an anionic amphiphile [440]. The role of the p47phox truncation was explained by the absence of the intramolecular link between the SH3 tandem and the deleted C-terminal polybasic region, enabling interaction with p22phox. The requirement for the truncation in p67phox was evocative of the enhanced p47phox-independent Nox2 activation by p67phox truncated at residue 246, which appears to have been the first time that an autoinhibitory effect was attributed to the C-terminal region of p67phox [344]. A rather unexpected result was that amphiphile-independent activation depended upon the C-terminal truncation of both p47phox and p67phox, suggesting that p67phox, too, has to undergo a structural reorganization in order to participate in Nox2 assembly. Kleinberg and coworkers described an amphiphile-independent cell-free system being the consequence of alanine substitution of anionic residues in p47phox, in the region linking the N-terminal SH3 to the C-terminal SH3, or the replacement of serines 310 and 328 by aspartate [441]. Both substitutions were supposed to dissociate the SH3 tandem from the polybasic region and represented alternatives to C-terminal truncation of p47phox. Unlike the requirement for truncations of both p47phox and p67phox, reported by Sumimoto [440], Nox2 activation occurred with full-length p67phox in the absence of amphiphile. 11.6.2 Prenylation-Dependent Amphiphile-Independent Nox2 Activation Pick and coworkers first reported that a semi-recombinant cell-free system, consisting of phagocyte membrane liposomes or purified relipidated cytochrome b558, p67phox and prenylated Rac1, generated O2•- upon addition of NADPH in the absence of an amphiphilic activator [289]. Prenylated Rac was not carboxymethylated and not subject to proteolysis at Cys189. The presence of p47phox was not required and its addition did not augment Nox2 activation. Prenylated Rac1 bound to membrane liposomes and artificial protein-free PC liposomes independently of p67phox, which was incapable of unmediated attachment,

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

proof for membrane attachment of p67phox being fully dependent on its interaction with Rac [289]. Tamura [349] and Pick [340] showed that Nox2 activation was induced in an amphiphile-stimulated cell-free system in which p67phox and Rac1 were replaced by a chimera consisting of p67phox, truncated at residue 210 or 212, fused with full-length Rac1. Prenylation of the [p67phox(1–212)Rac1(1–192)] chimera on the Rac1 moiety enabled it to activate Nox2 in the absence of amphiphile and p47phox and direct evidence was provided for binding of the chimera to the membrane being dependent on prenylation [290]. However, even though the dominant parameter governing binding of prenylated Rac and prenylated [p67phox-Rac1] chimeras to membranes was the hydrophobic interaction between the prenyl group and the membrane, charge-based interaction between the polybasic C-terminal region of the Rac moiety and the negatively charged interior aspect of the membrane was important [290]. A prenylated tripartite chimera (trimera), consisting of fused p47phox(1–286), phox p67 (1–212), and Rac1Q61L(1–192), was also a potent Nox2 activator in an amphiphile-independent system [442]. Interestingly, the presence in the trimera of a truncated p47phox moiety, lacking the AIR, allowed distinguishing between binding to the membrane, dependent on prenylation and charge, and binding to cytochrome b558 and activation of Nox2, which entailed the participation of the p47phox moiety, and of W193 in that moiety, in particular.

11.6.3 Replacing the Anionic Amphiphile with Anionic Membrane Phospholipids The importance of the negative charge of membrane phospholipids in Nox2 activation was first brought into prominence by Lambeth, when defining the requirements of membrane attachment of Rac [288], as an application of the principle of synergy between hydrophobic and electrostatic forces in anchoring proteins to membranes, elaborated by Aderem [443] (see Ref. [293] for a commentary). Tamura and coworkers were influenced by the work of Sumimoto on amphiphile-independent Nox2 activation by combining C-terminally truncated p47phox and p67phox [440] and designed a chimeric protein consisting of p67phox(1–210) fused to p47phox(1–286). When the chimera was combined with neutrophil membrane or purified relipidated cytochrome b558 and Rac-GTP, NADPH elicited O2•- generation in the absence of anionic amphiphile [444]. Tamura and Lambeth have reported in the past that PS enhanced Nox activity of solubilized enzyme derived from PMA-stimulated neutrophils [445]. Consistent with this finding, a cell-free system, consisting of the [p67phox(1-210)phox p47 (1-286)] chimera, purified cytochrome b558 supplemented in the course of relipidation with the anionic phospholipids PS or PI, and Rac-GTP, was found to exhibit

43

markedly enhanced NADPH-dependent O2•- generation in the absence of amphiphile [444]. Supplementation of neutrophil membranes with phosphoinositides or the anionic phospholipids PS and PG, but not with the neutral PC, markedly enhanced amphiphileindependent cell-free Nox2 activation by p47phox mutants with a severed intramolecular autoinhibitory bond, described by Kleinberg [441]. This report was also the first to relate minor but significant amphiphile-independent Nox2 activation by wild type p47phox, conditional on supplementation of membranes with PG or PA [441]. In an in debth study on the role of membrane phospholipids in Nox2 activation, the trimera [p47phox(1–286)-p67phox(1–212)-RacQ61L(1–192)] supported vigorous amphiphile-independent Nox2 activation by membrane liposomes enriched with anionic (PA, PG, PS, PI) but not neutral (PC) phospholipids [446]. Supplementation of membrane with anionic phospholipids affected mainly the EC50 of the trimera, with values in the order PA < PG < PS < PI; supplementation with PA resulted in the lowest EC50 (3.8 nM). Reducing the positive charge of the Rac1 moiety of the trimera by deleting Rac1 residues 179–192 or by replacing the all-basic Rac1 residues 183–188 with glutamines or with the less basic Rac2 residues 183–188, caused a pronounced loss of binding to the membrane [442] and Nox2 activating ability [446]. As in the case of the prenylated trimera, binding to cytochrome b558 involved the p47phox moiety proper, with a central role for residue W193. In the course of these studies it became apparent that enrichment of the membrane with anionic phospholipids, as an alternative to amphiphile-induced Nox2 activation, was not limited to chimeric constructs but also applied to conventional cell-free systems consisting of individual cytosolic components [197, 446]. Another example of amphiphileindependent Nox2 activation enhanced by enrichment of membranes with anionic phospholipids was that by [Rac1RhoGDI] complexes and p67phox [324]. Stephens and Hawkins and coworkers introduced a novel amphiphile-independent cell-free Nox2 activation method building on the property of phosphatidylinositol 3-phosphate (PtdIns(3)P) to bind to the PX domain of p40phox. O2•- generation was induced in a mixture of relipidated cytochrome b558 supplemented with PtdIns(3)P, a p40phox–p67phox complex, p47phox, and Rac-GTP [447]. Deletion of the PX domain of p40phox prevented Nox2 activation. To the best of my knowledge, this was the only cell-free system dependent on p40phox and its interaction with PtdIns(3)P. Since PtdIns(3)P was one of the products of PI(3)K, which was activated downstream of binding to membrane receptors of opsonized particles or chemotactic factors, such as fMLP, complement factor 5a, or interleukin 8 [448], this cell-free system represented a close equivalent of in vivo

44

events. It might be worth mentioning that in the amphiphileindependent cell-free system supported by a [p47phoxp67phox-Rac] trimera, there was no evidence for the participation of the PX domain of the p47phox moiety and its ligands in the process of Nox2 activation [442, 446]. The phospholipid composition of the cytosolic aspect of the plasma membrane conferred it a negative charge, due primarily to PS and PI but multivalent anionic phosphoinositides also contributed to it [292]. Although phospholipid remodeling of the membrane was studied mainly in association with phagocytosis [449], some “membrane receptors to Nox” signal transduction pathways might also be associated with increases in the negative charge of the cytosolic aspect of the membrane, making it likely that the in vitro models described above have their in vivo equivalents.

11.6.4 GEF-Assisted Nox2 Activation In a novel cell-free system, the GEF Trio, known to stimulate GDP to GTP exchange on Rac1 [450], enhanced O2•- generation by phagocyte membrane combined with prenylated Rac-GDP, p67phox, and GTP [301]. A rarely noticed additional and rather unusual pathway of Nox2 activation was by reducing the concentration of Mg2+ to the low micromolar range. This approach originated in the finding of Zheng and coworkers that GEFs act in part by displacement of Mg2+ bound to Rac [451]. Consequently, O2•- production could be induced in a cell-free system consisting solely of membrane, prenylated Rac-GDP, p67phox, and GTP, in the presence of 4 μM free Mg2+ and in the absence of a GEF and an amphiphile [301]. This finding was antedated by Cross and coworkers who described Nox2 activation by purified cytochrome b558 combined with p47phox, p67phox, nonprenylated Rac2, and p40phox, provided that Mg2+ was absent in the reaction medium [118]. This report stood out by its publication prior to the availability of the data on the role of Mg2+ in nucleotide exchange on Rac [451], although the possibility of the involvement of a G protein was mentioned in the paper. GEF-assisted cell-free Nox2 activation was further investigated in a rigorously controlled assay, composed of macrophage membrane, prenylated Rac1-GDP, p67phox, GTP, and Trio [298]. O2•- production did not require amphiphile and p47phox, and Trio could be replaced by another GEF, the Rac-specific Tiam1 [452]. The presence of tryptophan at position 56 in Rac was critical for discrimination by Rac-specific GEFs, comprising Trio and Tiam1 [453]. Replacing wild-type prenylated Rac1-GDP with a W56F mutant prevented Trio- and Tiam1-dependent activation [298]. An unexpected finding was that GTP could be replaced by ATP, the latter serving as a γ-phosphoryl donor for a membrane localized nucleoside diphosphate kinase converting GDP, derived by GEF-mediated dissociation from Rac-GDP, to GTP [298].

E. Pick

Considering that the physiological path of Rac-mediated Nox2 activation was initiated by Rac(GDP)–RhoGDI complexes, the in vitro model in which membrane, Rac (GDP)-RhoGDI, p67phox, GEF and GTP participated, was the most relevant (see Sect. 8.2.1).

11.7

Autoinhibition in p67phox – “Et Tu Brute?”

William Shakespeare, Julius Caesar; Act 3, Scene 1

To the best of my knowledge, the first experimental evidence for an autoinhibitory mechanism in p67phox was the finding by Lambeth and Freeman that the EC50 of p67phox truncated at residue 246, assayed in a cell-free assay in the absence of p47phox, was much lower than that of the full-length protein [344]. The authors suggested that a segment C-terminal to residue 246 masked a binding domain in the N-terminal half; removal of the C-terminal segment augmented the affinity of the truncated protein for Nox2. Using serial deletions, Babior and coworkers showed that p67phox truncated at residues 198 or 210 was bound to cytochrome b558 with higher affinity than the full-length protein [343]. In addition to its significance as an indicator for the existence of an intramolecular bond in p67phox, this work was the first to draw attention to the fact that structural requirements for binding to cytochrome b558 (with high probability, to Nox2) were different from those for supporting Nox2 activation. Thus, p67phox truncated at residues 246, 235, 221, 216, and 210 was active in cell-free assays whereas truncation at residue 198 yielded an inactive protein, which, nevertheless, bound to Nox2 [376]. This indicated that binding of p67phox to Nox2 by a domain N-terminal to residue 199 was not sufficient for activation, which was dependent on a second signal requiring the participation of p67phox residues 199–210, defined as the AD [376]. In the course of the serial truncations leading to the definition of the AD, it was noted that p67phox(1–246) was half as active as p67phox(1–235), suggesting that a segment consisting of residues 235–246 acted as an additional AIR [376]. p67phox truncated at residue 243 was also bound by cytochrome b558 with lower affinity than proteins truncated at residues 210 and 199 [343]. Jamel El Benna and coworkers were the first to provide direct experimental proof for the existence of intramolecular regulatory bonds in p67phox. In the course of studies on phosphorylation of p67phox, they identified a domain in the C-terminal part of the protein (residues 244–526) which was phosphorylated in the isolated C-terminal segment but not in the intact protein, suggesting that the site might be masked by a domain in the N-terminal half [454]. Indeed, incubation of the C-terminal segment with N-terminal segments corresponding to residues 1–243, 1–210, and 1–199

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

prevented phosphorylation of the C-terminal segment. This located the autoinhibitory sequence in the TPR-rich region, as opposed to its positioning in the C-terminal half, described in the earlier reports. Stringent structural evidence supporting the existence of intramolecular bonds in p67phox is lacking. Sedimentation velocity studies indicated that the shape of p67phox deviated considerably from that of a globular protein and that it had an extended and potentially flexible structure, explaining the fact that its molecular mass on gel filtration exceeded the theoretical value [455]. An investigation of the behavior of p67phox by small angle X-ray scattering (SAXS) revealed that it was an elongated molecule with multiple domains connected by flexible linkers reminiscent of an image of beads on a string [456]. Although the unstructured nature of the linkers was most compatible with the elongated shape, the data did not exclude the possible existence of intramolecular bonds, the most likely being between the N-terminal SH3 domain and an upstream region, accommodating the TPRs. Yet another source of the search for an intramolecular bond(s) in p67phox was the lack of a clear functional assignment for the N-terminal SH3 domain. Deletion of the N-terminal SH3 led to an impairment of Nox2 activation in whole cells transfected with the mutant p67phox [361, 457], which could, however, be overcome by increasing the expression of the deletion mutant [457]. In sharp contrast, p67phox lacking the N-terminal SH3 was fully competent in a cell-free system [361]. Solving this puzzle was made even more difficult by the finding that mutating tryptophan 277 in the N-terminal SH3 of p67phox, a residue essential for interaction with a PRR, did not affect the ability of p67phox to activate Nox2. This raised the possibility that the N-terminal SH3 might be engaged in an “illegal” intramolecular interaction with a region lacking a canonical PRR. The participation of the N-terminal SH3 in an intramolecular bond in p67phox was first proposed by Leto in a chapter published in 1999 [250]. The SH3 domain was supposed to interact in the resting state of the cell with the upstream PRR, consisting of residues 224–235, containing the canonical polyproline motif. In an earlier publication, the existence of a second PRR in p67phox, downstream from the N-terminal SH3 domain, was mentioned but not elaborated on [458]. This consisted of residues 318 PAPPSSKAPGRPQLSP333 and offered a theoretical alternative target for an intramolecular bond (personal communication from Dr. T.L. Leto). Unfortunately, after more than two decades from the time that these proposals were put forward, no experimental proof was provided in their support. Rather paradoxically, indirect evidence for the existence of an autoinhibitory mechanism in p67phox came from work on the p67phox homologue, NOXA1, which lacks the N-terminal SH3 domain [413, 414]. In a study by Clark and

45

coworkers the function of NOXA1 was investigated by expression in a cellular set up also expressing Nox1 and NOXO1. Truncation of NOXA1 (human NOXA1 is 476 residues long) at residue 273, yielding a protein lacking the only SH3 and the PB1 domains, resulted in enhanced PMA-stimulated O2•- production [459]. This was due primarily to the absence of the SH3 domain, as shown by enhanced Nox activity supported by NOXA1 truncated at residue 393, leaving the PB1 domain intact. Tamura and coworkers made use of the ability of NOXA1 to activate Nox2 in a cell-free system consisting of purified cytochrome b558, NOXA1, NOXO1, Rac and amphiphile [460]. Truncation of NOXA1 at residues 393 (removing the only SH3 domain); 273 (removing the SH3 and PB1 domains), or 225 (removing the C-terminal half, downstream of the AD) caused an enhancement of Nox activity (Vmax), proportional to the extent of truncation [461]. Truncations did not cause a change in EC50 in spite of the expectation that unmasking of site(s) occupied by C-terminal segments would enhance binding of NOXA1 to Nox2. A surprising finding was that the degree of NOXA1 truncation (from full-length to 1–225) correlated with an increase in the affinity of FAD for Nox2 (expressed in a decrease in EC50), as assayed in the NOXA1-supported cell-free system.

11.8

A Serendipitous Finding Reveals a Hidden Target in Nox2

Renewed focusing on the N-terminal SH3 domain of p67phox was generated by the serendipitous finding, made in the course of “peptide walking” through p67phox, that three overlapping p67phox peptides, 259–273, 262–276, and 265–279, inhibited Nox2 activation in a cell-free assay [462]. The common interpretation of such a result, competition between peptides and p67phox, was not applicable because the peptides also inhibited Nox2 activation supported by p67phox truncated at residue 212 and by a [p67phox(1–212)-Rac1(1–192)] chimera, both of which lacked the 259–279 sequence. The location of sequence 259–279 at the center of the N-terminal SH3 domain, the function of which was uncertain [361, 457], led to constructing a recombinant p67phox with residues 259–279, or the entire N-terminal SH3 domain, deleted. Both deletion mutants were fully competent to support Nox2 activation in vitro. Based on the concept of p67phox being the component engaged in direct interaction with Nox2, binding of the p67phox deletion mutants to a range of overlapping peptides, covering the full length of Nox2 DHR (residues 288–570), was assessed. Wild type p67phox was found before to bind to two Nox2 peptides corresponding to residues 357–371 (peptide 24) and 369–383 (peptide 28), located at a strategically important position, between the FAD and

46

E. Pick

Peptide 28 (369-383) Shared CysGlyCys (369-371) Peptide 24 (357-371)

Fig. 1.6 Predicted structure of human Nox2 by AlphaFold. The location of residues corresponding to Nox2 peptides 28 (369–383) (orange) and 24 (357–371) (yellow), and that of the overlapping residues 369–371 (purple) are pointed out. The orientation of the model is with the DHR on top and the transmembrane region towards the bottom of the figure

NADPH multi-site binding domains [463] (see Sect. 5.5). Although the two peptides shared a 369Cys-Gly-Cys371 sequence, binding of p67phox to peptide 357–371 was cysteine-dependent whereas binding to peptide 369–383 did not involve cysteines. The deletion mutant p67phox(1–526)Δ(259–279) was found to bind to peptide 369–383 with an affinity markedly exceeding that of the wild type protein, whereas binding to peptide 357–371 was unchanged [462]. Binding to Nox2 peptide 369–383 involved hydrophobic forces, was cysteine independent, Nox2 sequence specific, and focused on residues 375–383 [462]. The three-dimensional structure of residues 369–383 in the context of the whole Nox2 molecule, predicted by AlphaFold [464, 465], was that of an unstructured loop, “sticking out” of Nox2, compatible with accessibility to external proteins (Fig. 1.6) (E. Pick, unpublished results). The more N-terminal residues 357–368, corresponding to most of peptide 24, were structured as an α helix in the context of Nox2. Residues 357–361 participate in binding FAD [79, 175] and it is conceivable that binding of p67phox to residues 369–383 exerts an allosteric effect on the upstream FAD binding region, repositioning FAD to optimal distances from NADPH (as a hydride acceptor) or to the lower heme (as an electron donor) (see Sects. 10.1 and 12). Deletion of lesser segments of the 259–279 sequence, such as the sequence 265–270, and even deletion of single residues within the latter sequence was sufficient for specific high affinity binding to Nox2 peptide 369–383. The totality

of these results led to the conclusion that residues within the 259–279 sequence were engaged in an intramolecular bond with yet unknown residues, possibly positioned C-terminal to the N-terminal SH3 domain. Experimental evidence in support of the existence of an intramolecular bond was provided by the finding that p67phox peptides, corresponding to residues 259–279 and shown to inhibit Nox2 activation, were capable of auto-binding p67phox [462]. Truncation of the peptides indicated that the minimal sequence required for auto-binding comprised residues 265–273, nearly identical to the residues the deletion of which led to specific binding of p67phox to Nox2 peptide 369–383. An unsolved issue remained the mechanism responsible for the disengagement of the intramolecular bond. A study dealing with this question was triggered by the observation that, in addition to the inhibition of Nox2 activation by sequence 259–279 peptides, p67phox peptides 106–120 and 181–195 were also inhibitory. This was significant because residues within the 106–120 sequence (the β hairpin) were essential to the interaction of p67phox with Rac [360, 371, 373] and comprised residues G113 and Q115, shown to be linked by intramolecular hydrogen bonds to downstream residues R184 and K181, respectively [373], present in peptide 181–195. Deletion of residues 181–193 or point mutations Gln115Arg or Lys181Glu caused an increase in the binding affinity of the deleted or mutated p67phox to Nox2 peptide 369–383, similar to that seen following deletion of residues 259–279 [377]. This appeared to be the consequence of a conformational change in p67phox, as evidenced by an increase in the apparent molecular mass of the deleted and mutant proteins, probably unmasking a previously inaccessible binding site for Nox2. A [p67phox(1–212)-Rac1(1–192)Q61L] chimera also exhibited enhanced specific binding to Nox2 peptide 369–383, similar by all characteristics to that seen with p67phox deletion and point mutants [377, 462]. This finding and the involvement of residues in the β hairpin, neighboring residues G113 and Q115, in binding Rac-GTP, led to the proposal that Rac-GTP was responsible for converting p67phox from the autoinhibited to the activated form, capable of binding to Nox2 [377]. There is no better example for the way new paradigms are born, forgotten, and, then, reborn than the proposal by Babior made back in 2001 that “at least one of the oxidase-related functions of Rac1 is to promote the interaction between p67phox and cytochrome b558” (literal rendering of the sentence in the Abstract) [342]). What makes this historical detail actual is that the first author of the cited publication by Babior et al. (P.M-C. Dang) is a coauthor with J. El Benna of a recent Editorial Commentary related to the role of Rac in “activating” p67phox [466]. As always, many questions remain open, one of the most important being whether binding of Rac-modified p67phox to Nox2 is sufficient for the induction of electron flow or a

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

second signal is required, the obvious candidate for contributing it being the p67phox AD [376, 385] (see Editorial Commentary [466]).

12

“Curtain Up” Again

One could not wish for a more glorious ending to this chapter than the solving of a major part of the structure of Nox2 by two groups of investigators, using cryo-EM [467, 468]. Sigrid Noreng and coworkers co-expressed recombinant Nox2 and p22phox in mammalian cells, and following solubilization, purification, and supplementation with the Fab of an antibody against an extracellular site on Nox2, subjected it to cryo-EM [467]. The transmembrane domain (TMD) and the three extracellular loops were well resolved, whereas the cytosolic DHR, harboring the NADPH- and FAD-binding domains, was poorly visualized. The TMD of p22phox and both extracellular and intracellular loops were discernible but the cytosolic C-terminus was disordered. The architecture of the Nox2 TMD resembled that of csNox5 [130] and human DUOX1 [382] (see Chap. 30 by J-X. Wu, J. Sun, and L. Chen), with six transmembrane α helices, and two hemes orthogonal to the plane of the membrane, linked to four coordinating histidines on helices 3 and 5. The distance between the two hemes was 19.8 Å, with the shortest interatomic distance of 7.4 Å. A cluster of hydrophobic residues was located between the hemes, with Phe215 likely facilitating electron transfer between the two hemes. The oxygen reducing center was located between the outer heme and basic residues, including Arg54. A density feature compatible with a lipid alkyl chain, also detected in csNox5 and murine DUOX1, was in contact with helix 3 and might represent a phospholipid coordination site conserved in the Nox family. Detailed information on an extracellular structure comprising loops A, C, and E, covering the outer heme, became available, also comprising the three canonical glycosylation sites. Four potential tunnels for the entrance/exit of O2/O2•- were detected. p22phox was found to contain 4 transmembrane helices, two short extracellular loops, an intracellular loop, and a mostly unstructured C-terminal tail. Nox2 and p22phox interacted strongly by electrostatic and hydrophobic forces and shape complementarity, mediated by three interaction regions, consisting of Nox2 transmembrane helices 3, 4, and 5 and p22phox transmembrane helices 1 and 4, and contacts between loop D of Nox2 and the intracellular loop of p22phox. The low resolution of the DHR by cryo-EM indicated that the region was dynamic and loosely attached to the TMD in the resting state Nox2. Stabilizing the interactions between specific domains in the DHR and TMD was proposed as the

47

likely mechanism of Nox2 activation by the cytosolic components (see Sect. 12.2). In a similar study by Lei Chen and coworkers, Nox2 and p22phox were incorporated in phospholipid nanodiscs and the structure determined by cryo-EM [468]. A distinguishing feature of this investigation was the demonstration of the ability of the Nox2–p22phox heterodimer in nanodiscs to produce ROS when exposed to a [p47phox-p67phox-Rac] trimera, indicating that the enzyme in the resting state possessed all the components required for activation. The structural data confirmed independently those of Noreng and coworkers, including the high mobility of the DHR and the lack of docking to the TMD when Nox2 was in the resting state. FAD bound to DHR was observed at a distance of 5.6 Å from the lower heme.

12.1

Back to the Future

The “real” structural data provided by the two cryo-EM studies revealed how surprisingly accurate were the predictions resting on biochemistry, genetics and rather rudimentary bioinformatics. A search of the literature led to what appears to have been the first published model of Nox2 and p22phox by Jesaitis and coworkers [469], resting on prediction of secondary structure from the amino acid sequence [470]. The model was only partially correct; Nox2 was predicted to comprise five transmembrane helices and p22phox, three helices, a version also proposed by Segal and coworkers [124]. A six transmembrane helices archetype became dominant starting with a publication by Wallach and Segal [471], based on hydropathy profiles derived from the amino acid sequence [470, 472]. More detailed representations of what became the canonical architecture of Nox2 of several species, with minor differences in the location of residues, were published between 1998 and 2010 [79, 151, 152, 332, 473] and reviewed in [474, 475]. The finding of the presence of phospholipid at critical sites in Nox2 was a pleasing confirmation of earlier work showing the absolute dependence of Nox2 activity in vitro on a certain phospholipid milieu [246, 476]. The cryo-EM results finally settled the long-going controversy about the number of transmembrane helices in p22phox (from two to four) by confirming the prediction of four helices linked by two extracellular and one intracellular loop [418]. The strength of the Nox2–p22phox assembly and the multiplicity of Nox2–p22phox interaction interfaces explained the purification of cytochrome b558 as a heterodimer in nonionic detergents [94, 95, 98, 104] and the requirement of a strong anionic detergent (SDS) to dissociate the heterodimer [135]. Finally, the topography of the two hemes and the coordinating histidines on transmembrane helices 3 and

48

E. Pick

A

B

C

TM4

TM5 loop D

p67phox

Nox2 357-383

Nox2 357-383 DUOX1 1338-1362

Nox2 357-383 DUOX1 1338-1362

Fig. 1.7 (a) Model of human Nox2 by AlphaFold prediction, with the DHR at the bottom of the figure. DHR residues 357–383, consisting of an N-terminal α helix (residues 357–368) and a C-terminal unstructured region (residues 369–383), are colored gold. p67phox is represented by a block arrow pointing to residues 369–383. (b) Superposition of the Nox2 model by AlphaFold, supposed to represent the resting state enzyme (colored light blue), with human Duox1, based on the structure derived by cryo-EM in high Ca2+ state (PDB 7D3F) [382] (colored pink). Nox2 residues 357–383 are colored gold; the corresponding residues in Duox1 (1338–1362), identified by sequence alignment [467], are colored purple. The DHR of Nox2 is displaced compared to

the DHR of DUOX1, suggesting that activation of Nox2 is associated with the movement of the DHR of Nox2 into a position similar to that of the DHR of DUOX1 when activated by Ca2+ (see curved red arrow). (c) Superposition of the DHR of Nox2 with the DHR of DUOX1 in high Ca2+ state. This is based on the assumption that the DHR of Nox2, following activation by cytosolic components (p67phox), will be in a position similar to that of the DHR of DUOX1, following activation by Ca2+. The locations of transmembrane helices 4 (TM4) and 5 (TM5), and of loop D are indicated. This figure was generated based on an idea of Dr. Sigrid Noreng, who also provided the structural data

5, as evidenced by both cryo-EM studies, were proof for the excellence and predictive power of a long list of preliminary investigations and clinical correlates performed at a time when structural information was absent. This comprised the prediction of the histidines participating in the binding of the lower heme [148, 149], the ground-breaking discovery that cytochrome b558 contained two nonidentical hemes [90], and the findings that Nox2 was the heme-binding subunit [151], together with the identification of the four heme-liganding histidines [152] (see Sect. 4.3). The structural findings made by cryo-EM also served as a confirmation of the significance of the paradigm shifting realization of the similarity of the bis-heme motif and hydropathy profile of the FRE1 ferric reductase of Saccharomyces cerevisiae and the TMD of Nox2, culminating in the prediction of the four histidines coordinating the hemes in Nox2 [153–155]. The solved structure of the oxygen reduction center located between the outer heme and three basic residues, including Arg54, also revealed the surprising accuracy of the predictions made in the pre-structural era by Cross and coworkers [90], generated by investigating the mechanism of impaired O2•production caused by an Arg54 to Ser mutation in a CGD patient [477].

12.2

The Missing Finale

In both reports of the structure of Nox2, the lesser quality of the cryo-EM map of the DHR was suggested to reflect its flexibility in relation to the TMD, a state that was defined as “dynamic” [467] or “undocked” [468]. The flexibility of the DHR was attributed to Nox2-p22phox being in the resting state in the absence of the cytosolic components, and the authors of both reports proposed that stabilizing the interaction between the DHR and TMD in a “docked” conformation results in Nox2 activation. Since no structure of Nox2p22phox assembled with cytosolic components was available, the structure of human DUOX1 in the high Ca2+ state [382] served as a model for activated Nox2. An overlay of Nox2 in the resting state with the high Ca2+ state DUOX1 aligned by their TMD indicated that there was a displacement of the DHR of Nox2 compared to the DHR of DUOX1 and that Nox2 activation was likely associated with movement of the DHR of Nox2 to the position of the DHR in activated DUOX1 [467, 468]. Within the DHR of Nox2, a segment consisting of residues 369–383 was proposed to serve as a binding site for activated p67phox, the result of relief from an autoinhibited state [377, 462] (see Sect. 11.8). The segment is

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

49

2 O2.-

2 O2

p22phox

Nox2 Heme

Heme

anionic phospholipid

PX

Binding site to p67phox

C

AD TPR1 TPR2 TPR3 TPR4

switch I

++++++

polybasic

prenyl

SH3-N P R R

Binding site to Nox2

GTP/GDP

N

N

NADPH

C

Rac1

RhoGDI

anionic phospholipid

insert

switch II

++++++

C

prenyl

polybasic

anionic phospholipid

SH3-C

PRR SH3-N p67phox

p47phox SH3-C

C

N

C

Fig. 1.8 Representation of the assembled Nox2 complex. Emphasis is on protein—protein interactions discussed in the chapter. Although its role is acknowledged in the text of the chapter, p40phox is absent from the figure. The C-terminus of Rac appears twice: complexed to RhoGDI and attached to the membrane. All abbreviations are listed in the body of

the chapter (revised version of a figure which appeared in Dahan I, Pick E (2012) Strategies for identifying synthetic peptides to act as inhibitors of NADPH oxidases or “All that you did and did not want to know about inhibitory peptides”. Cell Mol Life Sci 69:2283–2305)

C-terminal to an α helix (residues 357–368), with residues 357–361 participating in binding FAD. The model illustrated in Fig. 1.7 rests on the hypothesis that a central role in Nox2 activation is the movement of a DHR segment, consisting of residues 357–383, from its position in the resting state to a position coinciding with that of the homologous region in DUOX1, in the Ca2+-activated state. This is supposed to bring the FAD binding site and, thus, FAD, to the proximity of loop D and of the lower heme, facilitating FAD → lower heme electron transfer. Binding of activated p67phox to residues 369–383 is either responsible for causing the movement or for stabilizing the DHR in the “docked” position. This might also bring the Nox2 α helix (357–368) and loop D to the proximity of the intracellular loop of p22phox, at the Nox2-p22phox interface. Only solving the structure of the functionally active Nox2, assembled with cytosolic components, will be able to confirm the veracity of this or

other models and offer the missing and much expected Finale to this work-intensive research effort. A very schematic representation of the assembled Nox2 complex, intended as a summary of the major protein–protein interactions leading to assembly, is displayed in Fig. 1.8.

13

Epilogue

History, as nearly no one seems to know, is not merely something to be read. And it does not refer merely, or even principally, to the past. On the contrary, the great force of history comes from the fact that we carry it within us, are unconsciously controlled by it in many ways, and history is literally present in all that we do. It could scarcely be otherwise, since it is to history that we owe our frames of reference, our identities, and our aspirations James Baldwin, The White Man’s Guilt (1965)

50

This chapter is not a review of Nox2 and its regulators. It is rather an attempt to tell the hisstory of the main discoveries as they emerged on the tortuous path leading to the present state of knowledge. Although I attempted to be as unbiased as possible, the story is, by definition “filtered” through the personality, background, and tastes of the author. This, I am certain, led to omissions of and overemphasis on certain subjects and a lack of uniformity in the level of detail offered in the narration of the various subjects. The emphasis was on the description of the fascinating and mysterious process of discovery. The names of those who made the discoveries were frequently but not always mentioned and some readers may feel that the choice was somewhat haphazard. Scientists have a strong ego and I apologize for lapses in objectivity in the selection of the names mentioned or not. I was also bothered by being forced to use the term “and coworkers” in association with the scientist who, usually, was the head of a group. As somebody who for more than five decades never left “the bench”, I am a very much aware of how difficult it is to know who in the group contributed what and how much and the little addendum at the end of papers, defined as “Author contributions”, is not of real help. It was also difficult to decide where to stop and I did not put too much thought into it because this is an open-ended story. In the course of my lifetime, sophisticated biochemistry, molecular biology, molecular genetics, and bioinformatics have transformed Nox investigation from its descriptive phase into an almost exact science. While this chapter was in preparation an extraordinary revolution in both determining the structure of Noxs and in predicting their threedimensional architecture took place. The manuscript of this chapter was submitted to the Publishers when the heartening news of the solving of the structure of Nox2 in the resting state proved that reality, sometimes, surpasses the most optimistic expectations [467, 468]. Based on the speed of progress in this field, solving the structure of Nox2 assembled with the cytosolic components as a functionally active complex looks like a goal attainable soon. In addition to the splendor of the achievement, it will offer a unique opportunity to assess which in the long Odyssey of Nox2 research was the right way to Ithaca. Acknowledgements I am using this opportunity to thank the colleagues, postdoctoral fellows, students, and research associates, with whom I had the privilege to work and collaborate in the course of five decades, for their key role in solving some of the mysteries of Nox2. Here is the list in alphabetical order: Arie Abo, Iris Aharoni, Natalie Alloul, Maya Amichay, Edna Bechor, Yevgeny Berdichevsky, Yael Bromberg, Iris Dahan, Valery Diatchuk, Iris Dotan, Sharon Engel, Aya Federman, Tanya Fradin, Maya Freund, Rahamim Gadba, Yara Gorzalczany, Miriam Hirshberg, Irina Issaieva, Michal Itan, Gili Joseph, Yona Keisari, Pablo Kiselstein, Sarah Knoller, Vasilij Koshkin, Taly Kroizman, Ofra Lotan, Ariel Mizrahi, Shahar Molshanski-Mor, Igor Morozov, Yael Nakash, Eyal Ozeri, Nimrod Pik, Meirav Rafalowski, Rive Sarfstein, Doron Sha’ag, Sally Shpungin, Natalia Sigal, Amir

E. Pick Toporik, Yelena Ugolev, and Anat Zahavi. They are, to paraphrase Churchill, “the few to whom I owe so much”. I thank Dr. Sigrid Noreng for most valuable advice on the structural aspects of Nox2 and for help in conceiving the model illustrated in Fig. 1.7.

References 1. Holmes B, Quie PG, Windhorst DB, Good RA (1966) Fatal granulomatous disease of childhood. Lancet 1:1225–1228 2. Baldridge CW, Gerard RW (1932) The extra respiration of phagocytes. Am J Phys 103:235–236 3. Ado AD (1933) Über den Verlauf der oxidativen und glykolitischen Prozesse in den Leukocyten des entzűndeten Gewebes während der Phagocytose. Z Ges Exp Med 87:473–480 4. Stähelin H, Suter E, Karnovsky ML (1956) Studies on the interaction between phagocytes and tubercle bacilli. I. Observations on the metabolism of guinea pig leukocytes and the influence of phagocytosis. J Exp Med 104:121–136 5. Sbarra AJ, Karnovsky ML (1959) The biochemical basis of phagocytosis. I. Metabolic changes during ingestion of particles by polymorphonuclear leukocytes. J Biol Chem 234:1355–1362 6. Cohn ZA, Morse SI (1960) Functional and metabolic properties of polymorphonuclear leucocytes. I. Observations on the requirements and consequences of particle ingestion. J Exp Med 111:667–687 7. Cagan RH, Karnovsky ML (1964) Enzymatic basis of the respiratory stimulation during phagocytosis. Nature 204:255–257 8. Segal BH, Leto TL, Gallin JI, Malech HL, Holland SM (2000) Genetic, biochemical, and clinical features of chronic granulomatous disease. Medicine 79:170–200 9. Karnovsky ML (1973) Chronic granulomatous disease – pieces of a cellular and molecular puzzle. Fed Proc 32:1527–1533 10. Badwey JA, Curnutte JT, Karnovsky ML (1979) The enzyme of granulocytes that produces superoxide and peroxide. An elusive Pimpernel. N Engl J Med 300:1157–1160 11. Badwey JA, Karnovsky ML (1986) Production of superoxide by phagocytic leukocytes: a paradigm for stimulus – response phenomena. Curr Top Cell Regul 28:183–208 12. Iyer GYN, Islam DMF, Quastel JH (1961) Biochemical aspects of phagocytosis. Nature 192:535–541 13. Iyer GYN, Quastel JH (1963) NADPH and NADH oxidation by guinea pig polymorphonuclear leucocytes. Can J Biol Physiol 41: 427–434 14. Kuhn TS (1970) The structure of scientific revolutions, 2nd edn. University of Chicago Press, Chicago 15. Rossi F, Zatti M (1964) Changes in the metabolic pattern of polymorphonuclear leucocytes during phagocytosis. Br J Exp Pathol 45:548–559 16. Rossi F, Zatti M (1964) Biochemical aspects of phagocytosis in polymorphonuclear leucocytes. NADH and NADPH oxidation by the granules of resting and phagocytosing cells. Experientia 20:21– 23 17. Borregaard N, Heiple JM, Simons ER, Clark RA (1983) Subcellular localization of the b-cytochrome component of the human microbicidal oxidase: translocation during activation. J Cell Biol 97:52–61 18. Clark RA, Leidal KG, Pearson DW, Nauseef WM (1987) NADPH oxidase of human neutrophils. Subcellular localization and characterization of an arachidonate activatable superoxide-generating system. J Biol Chem 262:4065–4074 19. Karnovsky ML, Wallach DFH (1961) The metabolic basis of phagocytosis. III. Incorporation of inorganic phosphate into various classes of phosphatides during phagocytosis. J Biol Chem 236: 1895–1901

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

20. Patriarca P, Dri P, Kakinuma K, Tedesco F, Rossi F (1975) Studies on the mechanism of metabolic stimulation in polymorphonuclear leucocytes during phagocytosis. I. Evidence for superoxide anion involvement in the oxidation of NADPH2. Biochim Biophys Acta 385:380–386 21. Curnutte JT, Karnovsky NL, Babior BM (1975) Manganesedependent NADPH oxidation by granulocyte particles. The role of superoxide and the nonphysiological nature of the manganese requirement. J Clin Invest 57:1059–1967 22. Nisimoto Y, Tamura M, Lambeth JD (1998) A menadionestimulated pyridine nucleotide oxidase from resting bovine neutrophil membranes. Purification, properties, and immunochemical cross-reactivity with the human neutrophil NADPH oxidase. J Biol Chem 263:11657–11663 23. Fridovich I (1978) The biology of oxygen radicals. Science 201: 875–880 24. McCord JM, Fridovich I (1969) Superoxide dismutase: an enzymatic function for erythrocuprein (hemocuprein). J Biol Chem 244: 6049–6055 25. McCord JM, Fridovich I (1971) An enzyme-based theory of obligate anaerobiosis: the physiological function of superoxide dismutases. Proc Natl Acad Sci U S A 68:1024–1027 26. Klebanoff SJ (1967) Iodination of bacteria. A bactericidal mechanism. J Exp Med 126:1063–1078 27. Lehrer RI, Cline MJ (1969) Leukocyte myeloperoxidase deficiency and disseminated candidiasis: the role of myeloperoxidase in resistance to Candida infections. J Clin Invest 48:1478–1488 28. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanism. The production by leukocytes of superoxide, a potential bactericidal agent. J Clin Invest 52:741–743 29. Babior BM (1978) Oxygen-dependent microbial killing by phagocytes (First of two parts). N Engl J Med 298:659–668 30. Segal AW, Meshulam T (1979) Production of superoxide by neutrophils: a reappraisal. FEBS Lett 100:27–32 31. Babior BM (1979) Superoxide production by phagocytes. Another look at the effect of cytochrome c on oxygen uptake by stimulated neutrophils. Biochem Biophys Res Commun 91:222–226 32. Root RK, Metcalf JA (1977) H2O2 release from human granulocytes during phagocytosis: Relationship to superoxide anion formation and cellular catabolism of H2O2: studies with normal and cytochalasin B treated cells. J Clin Invest 60:1266– 1279 33. Babior BM, Curnutte JT, McMurrich BJ (1976) The particulate superoxide-forming system from human neutrophils. Properties of the system and further evidence supporting its participation in the respiratory burst. J Clin Invest 58:989–996 34. Curnutte JT, Kipnes RB, Babior BM (1975) Defect in pyridine nucleotide dependent superoxide production by a particulate fraction from the granulocytes of patients with chronic granulomatous disease. N Engl J Med 293:628–632 35. Babior BM, Kipnes RS (1977) Superoxide-forming enzyme from human neutrophils: evidence for a flavin requirement. Blood 50: 517–524 36. Gabig TG, Kipnes RS, Babior BM (1978) Solubilization of the O2- – forming activity responsible for the respiratory burst in human neutrophils. J Biol Chem 253:6663–6665 37. Gabig TG, Babior BM (1979) The O2- – forming oxidase responsible for the respiratory burst in human neutrophils. Properties of the solubilized enzyme. J Biol Chem 254:9070–9074 38. Babior BM, Peters WA (1981) The O2- – producing enzyme of human neutrophils. Further properties. J Biol Chem 256:2321– 2323 39. DeChatelet LR, McCall C, Shirley PS (1980) Activation by dialysis of NAD(P)H oxidase (s) from human neutrophils. J Reticuloendothel Soc 28:533–545

51

40. Light DR, Walsh C, O’Callaghan AM, Goetzl EJ, Tauber AI (1981) Characteristics of the cofactor requirements for the superoxide-generating NADPH oxidase of human polymorphonuclear leukocytes. Biochemistry 20:1468–1476 41. Gabig TG (1983) The NADPH-dependent O2•- – generating oxidase from human neutrophils. Identification of a flavoprotein component that is deficient in a patient with chronic granulomatous disease. J Biol Chem 258:6352–6356 42. Gabig TG, Lefker BA (1984) Deficient flavoprotein component of the NADPH-dependent O2•- – generating oxidase in neutrophils from three male patients with chronic granulomatous disease. J Clin Invest 73:701–705 43. Gabig TG, Lefker BA (1984) Catalytic properties of the resolved flavoprotein and cytochrome b components of the NADPH dependent O2•- generating oxidase from human neutrophils. Biochem Biophys Res Commun 118:430–436 44. Wakeyama H, Takeshige K, Minakami S (1983) NADPHdependent reduction of 2,6-dichlorophenol-indophenol by the phagocytic vesicles of pig polymorphonuclear leucocytes. Biochem J 210:577–581 45. Murakami M, Nakamura M, Minakami S (1986) 2.6-dichlorophenolindophenol-reducing activity of phagocytosis-associated NADPH oxidase. J Biochem 100:1493–1497 46. Kakinuma K, Kaneda M, Chiba T, Ohnishi T (1986) Electron spin resonance studies on a flavoprotein in neutrophil plasma membranes. Redox potentials of the flavin and its participation in NADPH oxidase. J Biol Chem 261:9426–9432 47. Kakinuma K, Fukuhara Y, Kaneda M (1987) The respiratory burst oxidase of neutrophils. Separation of an FAD enzyme and its characterization. J Biol Chem 262:12316–12322 48. Fleck L (1979) Genesis and development of a scientific fact. The University of Chicago Press, Chicago 49. Borregaard N, Tauber AI (1984) Subcellular localization of the human neutrophil NADPH oxidase. b-cytochrome and associated flavoprotein. J Biol Chem 259:47–52 50. Parkinson JF, Gabig TG (1988) Isolation of the respiratory burst oxidase: the role of a flavoprotein component. J Bioenerg Biomembr 20:653–677 51. Cross AR, Jones OTG, Garcia R, Segal AW (1982) The association of FAD with cytochrome b-245 of human neutrophils. Biochem J 208:759–763 52. Segal AW, Jones OTG (1978) Novel cytochrome b system in phagocytic cells. Nature 276:515–517 53. Nisimoto Y, Otsuka-Murakami H, Lambeth DJ (1995) Reconstitution of flavin-depleted neutrophil flavocytochrome b558 with 8-mercapto-FAD and characterization of the flavin-reconstituted enzyme. J Biol Chem 270:16428–16434 54. Crawford DR, Schneider DL (1982) Identification of ubiquinone50 in human neutrophils and its role in microbicidal events. J Biol Chem 257:6662–6668 55. Cunningham CC, DeChatelet LR, Spach PI et al (1982) Identification and quantitation of electron transport components in human polymorphonuclear neutrophils. Biochim Biophys Acta 682:430– 435 56. Crawford DR, Schneider DL (1983) Ubiquinone content and respiratory burst activity of latex-filled phagolysosomes isolated from human neutrophils and evidence for the probable involvement of a third granule. J Biol Chem 258:5363–5367 57. Gabig TG, Lefker BA (1985) Activation of the human neutrophil NADPH oxidase results in coupling of electron carrier function between ubiquinone-10 and cytochrome b559. J Biol Chem 260: 3991–3995 58. Cross AR, Jones OTG, Garcia R, Segal AW (1983) The subcellular localization of ubiquinone in human neutrophils. Biochem J 216: 765–768

52 59. Lutter R, van Zwieten R, Weening RS, Hamers MN, Roos D (1984) Cytochrome b, flavins, and ubiquinone-50 in enucleated human neutrophils (polymorphonuclear leukocyte cytoplasts). J Biol Chem 259:9603–9606 60. Cross AR, Jones OTG, Harper AM, Segal AW (1981) Oxidationreduction properties of the cytochrome b found in the plasma membrane fraction of human neutrophils. A possible oxidase in the respiratory burst. Biochem J 194:599–606 61. Tauber AI, Goetzl EJ (1979) Structural and catalytic properties of the solubilized superoxide-generating activity of human polymorphonuclear leukocytes. Solubilization, stabilization in solution and partial characterization. Biochemistry 18:5576–5584 62. Pick E, Bromberg Y, Shpungin S, Gadba R (1987) Activation of the superoxide forming NADPH oxidase in a cell-free system by sodium dodecyl sulfate. Characterization of the membraneassociated component. J Biol Chem 262:16476–16483 63. Bellavite P, Cross AR, Serra MC, Davoli A, Jones OTG, Rossi F (1983) The cytochrome b and flavin content and properties of the O2- - forming NADPH oxidase solubilized from activated neutrophils. Biochim Biophys Acta 746:40–47 64. Bellavite P, Jones OTG, Cross AR, Papini E, Rossi F (1984) Composition of partially purified NADPH oxidase from pig neutrophils. Biochem J 223:639–648 65. Serra MC, Bellavite P, Davoli A, Rossi F (1984) Isolation from neutrophil membranes of a complex containing active NADPH oxidase and cytochrome b-245. Biochim Biophys Acta 788:138– 146 66. Berton G, Papini E, Cassatella MA, Bellavite P, Rossi F (1985) Partial purification of the superoxide-generating system of macrophages. Possible association of the NADPH oxidase activity with a low-potential (-245 mV) cytochrome b. Biochim Biophys Acta 810:164–173 67. Cross AR, Parkinson JF, Jones OTG (1984) The superoxidegenerating oxidase of leucocytes. NADPH-dependent reduction of flavin and cytochrome b in solubilized preparations. Biochem J 223:337–344 68. Wakeyama H, Takeshige K, Takayanagi R, Minakami S (1982) Superoxide-forming NADPH oxidase preparation of pig polymorphonuclear leucocytes. Biochem J 205:593–601 69. Umei T, Takeshige K, Minakami S (1986) NADPH binding component of neutrophil superoxide-generating oxidase. J Biol Chem 261:5229–5232 70. Kojima H, Takahashi K, Sakane F, Koyama J (1987) Purification and characterization of a NADPH-cytochrome c reductase from porcine polymorphonuclear leukocytes. J Biochem 102:1083– 1088 71. Sakane F, Kojima H, Takahashi K, Koyama J (1987) Porcine polymorphonuclear leukocyte NADPH-cytochrome c reductase generates superoxide in the presence of cytochrome b559 and phospholipid. Biochem Biophys Res Commun 147:71–77 72. Doussière J, Vignais PV (1985) Purification and properties of an O2•- – generating oxidase from bovine polymorphonuclear neutrophils. Biochemistry 24:7231–7239 73. Morel F, Doussière J, Stasia M-J, Vignais PV (1985) The respiratory burst of bovine neutrophils. Role of a b type cytochrome and coenzyme specificity. Eur J Biochem 152:669–679 74. Markert M, Glass GA, Babior BM (1985) Respiratory burst oxidase from human neutrophils: Purification and some properties. Proc Natl Acad Sci U S A 82:3144–3148 75. Glass GA, DeLisle DM, DeTogni P, Gabig TG, Magee BH, Markert M, Babior BM (1986) The respiratory burst oxidase of human neutrophils. Further studies of the purified enzyme. J Biol Chem 261:13247–13251 76. Bellavite P, Casatella MA, Papini E, Magyeri P, Rossi F (1986) Presence of cytochrome b-245 in NADPH oxidase preparations from human neutrophils. FEBS Lett 199:159–163

E. Pick 77. Green TR, Pratt KL (1988) Purification of the solubilized NADPH: O2 oxidoreductase of human neutrophils. Isolation of its catalytically inactive cytochrome b and flavoprotein redox centers. J Biol Chem 263:5617–5623 78. Rae J, Newburger PE, Dinauer MC et al (1998) X-linked chronic granulomatous disease: mutations in the CYBB gene encoding the gp91phox component of the respiratory-burst oxidase. Am J Hum Genet 62:1320–1331 79. Debeurme F, Picciochi A, Dagher M-C et al (2010) Regulation of NADPH oxidase activity in phagocytes. Relationship between FAD/NADPH binding and oxidase complex assembly. J Biol Chem 285:33197–33208 80. Reis J, Massari M, Marchese S et al (2020) A closer look into NADPH oxidase inhibitors: Validation and insight into their mechanism of action. Redox Biol 32:101466 81. Hattori H (1961) Studies on the labile, stable Nadi oxidase and peroxidase staining reactions in the isolated particles of horse granulocyte. Nagoya J Med Sci 23:362–378 82. Ehrlich P (1877) Beiträge zur Kenntniss der Anilinfärbungen und ihrer Verwendung in der mikroskopischen Technik. Archiv für Mikrosk Anat 13:263–277 83. Shinagawa Y, Tanaka C, Teraoka A, Shinagawa Y (1966) A new cytochrome in neutrophilic granules of rabbit leukocyte. J Biochem 59:622–624 84. Segal AW, Jones OTG, Webster D, Allison AC (1978) Absence of a newly described cytochrome b from neutrophils of patients with chronic granulomatous disease. Lancet II:446–449 85. Borregaard N, Staehr Johansen K, Taudorff E, Wandall JH (1979) Cytochrome b is present in neutrophils from patients with chronic granulomatous disease. Lancet I:949–951 86. Segal AW, Jones OTG (1979) Neutrophil cytochrome b in chronic granulomatous disease. Lancet I:1036–1037 87. Segal AW, Jones OTG (1980) Absence of cytochrome b reduction in stimulated neutrophils form both female and male patients with chronic granulomatous disease. FEBS Lett 110:111–114 88. Roos D, van Leeuwen K, Hsu AP et al (2021) Hematologically important mutations: The autosomal forms of chronic granulomatous disease (third update). Blood Cells Mol Dis 92:102596 89. Schröder K, Weissmann N, Brandes RP (2017) Organizers and activators: cytosolic proteins impacting on vascular function. Free Radic Biol Med 109:22–32 90. Cross AR, Rae J, Curnutte JT (1995) Cytochrome b-245 of the neutrophil superoxide-generating system contains two nonidentical hemes. Potentiometric studies of a mutant form of gp91phox. J Biol Chem 270:17075–17077 91. Cross AR, Higson FK, Jones OT, Harper AM, Segal AW (1982) The enzymic reduction and kinetic of oxidation of cytochrome b-245 of neutrophils. Biochem J 204:479–485 92. Cross AR, Parkinson JF, Jones OTG (1985) Mechanism of the superoxide-producing oxidase of neutropils. O2 is necessary for the fast reduction of cytochrome b-254 by NADPH. Biochem J 226: 881–884 93. Koshkin V (1995) Aerobic and anaerobic functioning of superoxide-producing cytochrome b-559 reconstituted with phospholipids. Biochim Biophys Acta 1232:225–229 94. Harper AM, Dunne MJ, Segal AW (1984) Purification of cytochrome b-245 from human leukocytes. Biochem J 219:519–527 95. Segal AW (1987) Absence of both cytochrome b-245 subunits from neutrophils in X-linked chronic granulomatous disease. Nature 326:88–91 96. Harper AM, Chaplin MF, Segal AW (1985) Cytochrome b-245 from human neutrophils is a glycoprotein. Biochem J 227:783–788 97. Jesaitis AJ, Heners PR, Hertel R (1977) Characterization of a membrane fraction containing a b-type cytochrome. Plant Physiol 59:941–947

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

98. Parkos CA, Allen RA, Cochrane CG, Jesaitis AJ (1987) Purified cytochrome b from human granulocyte plasma membrane is comprised of two polypeptides with relative molecular weights of 91,000 and 22,000. J Clin Invest 80:732–742 99. Parkos CA, Allen RA, Cochrane CG, Jesaitis AJ (1988) The quaternary structure of the plasma membrane b-type cytochrome of human granulocytes. Biochim Biophys Acta 932:71–83 100. Pember SO, Heyl BL, Kinkade JM Jr, Lambeth JD (1984) Cytochrome b558 from (bovine) granulocytes. Partial purification from Triton X-114 extracts and properties of the isolated cytochrome. J Biol Chem 259:10590–10595 101. Lutter R, van Schalk MLJ, van Zwieten R, Wever R, Roos D, Hamers MN (1985) Purification and partial characterization of the b-type cytochrome from human polymorphonuclear leukocytes. J Biol Chem 260:2237–2244 102. Bellavite P, Papini E, Zeni L, Della Bianca V, Rossi F (1985) Studies on the nature and activation of O2- – forming NADPH oxidase of leukocytes. Identification of a phosphorylated component of the active enzyme. Free Radic Res Commun 1:11–29 103. Morel F, Vignais PV (1987) Purification of cytochrome b558 from bovine polymorphonuclear neutrophils. Biochem Biophys Res Commun 149:46–55 104. Knoller S, Shpungin S, Pick E (1991) The membrane-associated component of the amphiphile-activated, cytosol-dependent superoxide-forming NADPH oxidase of macrophages is identical to cytochrome b559. J Biol Chem 266:2795–2804 105. Rotrosen D, Yeung CL, Leto TL, Malech HL, Kwong CH (1992) Cytochrome b558: the flavin-binding component of the phagocyte NADPH oxidase. Science 256:1459–1462 106. Abo A, Boyhan A, West I, Thrasher AJ, Segal AW (1992) Reconstitution of neutrophil NADPH oxidase activity in the cell-free system by four components: p67phox, p47phox, p21rac1 and cytochrome b-245. J Biol Chem 267:16767–16770 107. Miki T, Yoshida LS, Kakinuma K (1992) Reconstitution of superoxide-forming NADPH oxidase activity with cytochrome b558 purified from porcine neutrophils. Requirement of a membrane-bound flavin enzyme for reconstitution of activity. J Biol Chem 267:18695–18701 108. Koshkin V, Pick E (1993) Generation of superoxide by purified and relipidated cytochrome b559 in the absence of cytosolic activators. FEBS Lett 327:57–62 109. Koshkin V, Pick E (1994) Superoxide production by cytochrome b559. Mechanism of cytosol-independent activation. FEBS Lett 338:285–289 110. Seifert R, Rosenthal W, Schultz G (1986) Guanine nucleotides stimulate NADPH oxidase in membranes of human neutrophils. FEBS Lett 205:161–165 111. Gabig TG, English D, Akard LP, Schell MJ (1987) Regulation of neutrophil NADPH oxidase activation in a cell-free system by guanine nucleotides and fluoride. Evidence for participation of a pertussis and cholera toxin sensitive G protein. J Biol Chem 262: 1685–1690 112. Quinn MT, Parkos CA, Walker L, Orkin SH, Dinauer MC, Jesaitis AJ (1989) Association of a Ras-related protein with cytochrome b of human neutrophils. Nature 342:198–200 113. Bokoch GM, Quilliam LA, Bohl BP, Jesaitis AJ, Quinn MT (1991) Inhibition of Rap1A binding to cytochrome b558 of NADPH oxidase by phosphorylation of Rap1A. Science 254:1794–1796 114. Quinn MT, Curnutte JT, Parkos CA et al (1992) Reconstitution of defective respiratory burst activity with partially purified human neutrophil cytochrome b in two genetic forms of chronic granulomatous disease: possible role of Rap1A. Blood 79:2438–2445 115. Eklund EA, Marshall M, Gibbs JB, Crean CD, Gabig TG (1991) Resolution of a low molecular weight G protein in neutrophil cytosol required for NADPH oxidase activation and reconstitution by recombinant Krev-1 protein. J Biol Chem 266:13964–13970

53

116. Gabig TG, Crean CD, Mantel PL, Rosli R (1995) Function of wildtype or mutant Rac2 and Rap1a GTPases in differentiated HL60 cell NADPH oxidase activation. Blood 85:804–811 117. Li Y, Yan J, De P et al (2007) Rap1A null mice have altered myeloid cell functions suggesting distinct roles for the closely related Rap1a and 1b proteins. J Immunol 179:8322–8331 118. Cross A, Erickson RW, Ellis BA, Curnutte JT (1999) Spontaneous activation of NADPH oxidase in a cell-free system: unexpected multiple effects of magnesium ion concentrations. Biochem J 338: 229–233 119. Shiro Y, Isogai Y, Nakamura H, Iizuka T (2002) Physiological functions and molecular structures of new types of hemoproteins. In: Endo I et al (eds) Molecular anatomy of cellular systems. Elsevier, Amsterdam, pp 189–204 120. Iizuka T, Kanegasaki S, Makino R, Tanaka T, Ishimura Y (1985) Studies on neutrophil b- type cytochrome in situ by low temperature absorption spectroscopy. J Biol Chem 260:12049–12053 121. Hurst JK, Loehr TM, Curnutte JT, Rosen H (1991) Resonance Raman spectra and electron paramagnetic resonance structural investigations of neutrophil cytochrome b558. J Biol Chem 266: 1627–1634 122. Ueno I, Fujii S, Ohya-Nishiguchi H, Iizuka T, Kanegasaki S (1991) Characterization of neutrophil b-type cytochrome on situ by electron paramagnetic spectroscopy. FEBS Lett 281:130–132 123. Miki T, Fujii H, Kakinuma K (1992) EPR signals of cytochrome b558 purified from porcine neutrophils. J Biol Chem 267:19673– 19675 124. Segal AW, West I, Wientjes F et al (1992) Cytochrome b-245 is a flavocytochrome containing FAD and NADPH-binding site of the microbicidal oxidase of phagocytes. Biochem J 284:781–788 125. Fujii H, Johnson M, Finnegan MG, Miki T, Yoshida LS, Kakinuma K (1995) Electron spin resonance studies on neutrophil cytochrome b558. Evidence that low-spin heme iron is essential for O2•- generating activity. J Biol Chem 270:12685–12689 126. Doussière J, Gaillard J, Vignais PV (1996) Electron transfer across the O2- generating flavocytochrome b of neutrophils. Evidence for a transition from a low-spin state to a high-spin state of the heme iron component. Biochemistry 35:13400–13410 127. Fujii H, Finnegan MG, Johnson MK (1999) The active form of the ferric heme in neutrophil cytochrome b558 is low-spin in the reconstituted cell-free system in the presence of amphophil. J Biochem 126:708–714 128. Isogai Y, Iizuka T, Makino R, Iyanagi T, Orii Y (1993) Superoxide-producing cytochrome b. Enzymatic and electron paramagnetic resonance properties of cytochrome b558 purified from neutrophils. J Biol Chem 268:4025–4031 129. Isogai Y, Iizuka T, Shiro T (1995) The mechanism of electron donation to molecular oxygen by phagocytic cytochrome b558. J Biol Chem 270:7853–7857 130. Magnani F, Nenci S, Fananas EM et al (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci U S A 114:6764–6769 131. Orkin SH (1986) Reverse genetics and human disease. Cell 47: 845–850 132. Baehner RL, Kunkel LM, Monaco AP et al (1986) DNA linkage analysis of X chromosome-linked chronic granulomatous disease. Proc Natl Acad Sci U S A 83:3398–3401 133. Royer-Pokora B, Kunkel LM, Monaco AP et al (1986) Cloning the gene for an inherited human disorder – chronic granulomatous disease – on the basis of its chromosomal location. Nature 322: 32–38 134. Orkin SH (1989) Molecular genetics of chronic granulomatous disease. Annu Rev Immunol 7:277–307 135. Teahan C, Rowe P, Parker P, Totty N, Segal AW (1987) The X-linked chronic granulomatous disease gene codes for the β-chain of cytochrome b-245. Nature 327:720–721

54 136. Dinauer MC, Orkin SH, Brown R, Jesaitis AJ, Parkos CA (1987) The glycoprotein encoded by the X-linked chronic granulomatous disease locus is a component of the neutrophil cytochrome b complex. Nature 327:717–720 137. Parkos CA, Dinauer MC, Walker LE, Allen RA, Jesaitis AJ, Orkin SH (1988) Primary structure and unique expression of the 22-kilodalton light chain of human neutrophil cytochrome b. Proc Natl Acad Sci U S A 85:3319–3323 138. Dinauer MC, Pierce EA, Bruns GA, Curnutte JT, Orkin SH (1990) Human neutrophil cytochrome b light chain (p22phox). Gene structure, chromosomal location and mutations in cytochrome-negative autosomal recessive chronic granulomatous disease. J Clin Invest 86:1729–1737 139. Parkos CA, Dinauer MC, Jesaitis AJ, Orkin SH, Curnutte JT (1989) Absence of both the 91kD and 22kD subunits of human neutrophil cytochrome b in two genetic forms of chronic granulomatous disease. Blood 73:1416–1420 140. Yu L, Zhen L, Dinauer MC (1997) Biosynthesis of the phagocyte NADPH oxidase cytochrome b558. Role of heme incorporation and heterodimer formation in maturation and stability of gp91phox and p22phox subunits. J Biol Chem 272:27288–27294 141. DeLeo FR, Burritt JB, Yu L, Jesaitis AJ, Nauseef WM (2000) Processing and maturation of flavocytochrome b558 include incorporation of heme as a prerequisite for heterodimer assembly. J Biol Chem 275:13986–13993 142. Dinauer MC, Pierce EA, Ericson RW et al (1991) Point mutation in the cytoplasmic domain of the neutrophil p22phox cytochrome b subunit is associated with a nonfunctional NADPH oxidase and chronic granulomatous disease. Proc Natl Acad Sci U S A 88: 11231–11235 143. Sumimoto H, Kage Y, Nunoi H et al (1994) Role of Src homology 3 domains in assembly and activation of the phagocyte NADPH oxidase. Proc Natl Acad Sci U S A 91:5345–5349 144. Leto TL, Adams AG, De Mendez I (1994) Assembly of the phagocyte NADPH oxidase: binding of Src homology 3 domains to proline-rich targets. Proc Natl Acad Sci U S A 91:10650–10654 145. Stasia MJ (2016) CYBA encoding p22phox, the cytochrome b558 alpha polypeptide: gene structure, expression, role in physiology. Gene 586:27–35 146. Nugent JHA, Gratzer W, Segal AW (1989) Identification of the haem-binding subunit of cytochrome b-245. Biochem J 264:921– 924 147. Quinn MT, Mullen M, Jesaitis AJ (1992) Human neutrophil cytochrome b contains multiple hemes. Evidence for heme associated with both subunits. J Biol Chem 267:7303–7309 148. Fujii H, Yonetani T, Miki T, Kakinuma K (1995) Modulation of the heme environment of neutrophil cytochrome b558 to a “cytochrome P450-like” structure by pyridine. J Biol Chem 270:3193– 3196 149. Bolscher BGJM, de Boer M, de Klein A, Weening RS, Roos D (1991) Point mutations in the β-subunit of cytochrome b558 leading to X-linked chronic granulomatous disease. Blood 77:2482–2487 150. Dinauer MC (2019) Insights into the NOX NADPH oxidases using heterologous whole cell assays. In: Knaus UG, Leto TL (eds) NADPH oxidases Methods and protocols. Springer Science+Business Media, New York, pp 139–151 151. Yu L, Quinn MT, Cross AR, Dinauer MC (1998) gp91phox is the heme binding subunit of the superoxide-generating NADPH oxidase. Proc Natl Acad Sci U S A 95:7993–7998 152. Biberstine-Kinkade KJ, DeLeo FR, Epstein RI, LeRoy BA, Nauseef WM, Dinauer MC (2001) Heme-ligating histidines in flavocytochrome b558. Identification of specific histidines in gp91phox. J Biol Chem 276:31105–31112 153. Dancis A, Roman DG, Anderson GJ, Hinnebusch AG, Klausner RD (1992) Ferric reductase of Saccharomyces cerevisiae:

E. Pick molecular characterization, role in iron uptake, and transcriptional control by iron. Proc Natl Acad Sci U S A 89:3869–3873 154. Shatwell KP, Dancis A, Cross AR, Klausner RD, Segal AW (1996) The FRE1 ferric reductase of Saccharomyces cerevisiae is a cytochrome b similar to that of NADPH oxidase. J Biol Chem 271: 14240–14244 155. Finegold AA, Shatwell KP, Segal AW, Klausner RD, Dancis A (1996) Intramembrane bis-heme motif for transmembrane electron transport conserved in a yeast iron reductase and the human NADPH oxidase. J Biol Chem 271:31021–31024 156. Zhang X, Krause K-H, Xenarios I, Soldati T, Boeckmann B (2013) Evolution of the ferric reductase (FRD) superfamily: modularity, functional diversification, and signature motifs. PLoS One 8: e58126 157. Oosterheert W, Reis J, Gros P, Mattevi A (2020) An elegant fourhelical fold in NOX and STEAP enzymes facilitates electron transport across biomembranes – similar vehicle, different destination. Acc Chem Res 53:1969–1980 158. Chiba T, Kaneda M, Fujii H, Clark RA, Nauseef WM (1990) Two cytosolic components of the neutrophil NADPH oxidase, p47phox and p67phox, are not flavoproteins. Biochem Biophys Res Commun 173:376–381 159. Leusen JHW, Meischl C, Eppink MHM et al (2000) Four novel mutations in the gene encoding gp91phox of human NADPH oxidase; consequences for oxidase assembly. Blood 95:666–673 160. Umei T, Takeshige K, Minakami S (1987) NADPH-binding component of the superoxide-generating oxidase in unstimulated neutrophils and the neutrophils from patients with chronic granulomatous disease. Biochem J 243:467–472 161. Doussière J, Laporte F, Vignais PV (1986) Photolabeling of a O2•generating protein in bovine polymorphonuclear neutrophils by an arylazido NADP+ analog. Biochem Biophys Res Commun 29:85– 93 162. Doussière J, Brandolin G, Derrien V, Vignais PV (1993) Critical assessment of the presence of an NADPH binding site on neutrophil cytochrome b558 by photoaffinity and immunochemical labeling. Biochemistry 32:8880–8887 163. Ravel P, Lederer F (1993) Affinity-labeling of an NADPH-binding site on the heavy subunit of flavocytochrome b558 in particulate NADPH oxidase from activated neutrophils. Biochem Biophys Res Commun 196:543–552 164. Sha’ag D, Pick E (1988) Macrophage-derived superoxidegenerating NADPH in an amphiphile-activated cell-free system; partial purification of the cytosolic component and evidence that it may contain the NADPH binding site. Biochim Biophys Acta 952: 213–219 165. Sha’ag D, Pick E (1990) Nucleotide binding properties of cytosolic components required for expression of activity of the superoxide generating NADPH oxidase. Biochim Biophys Acta 1037:405– 412 166. Smith RM, Curnutte JT, Babior BM (1989) Affinity labelling of the cytosolic and membrane components of the respiratory burst oxidase by the 2′3′-dialdehyde derivative of NADPH. Evidence for a cytosolic location of the nucleotide-binding site in the resting cell. J Biol Chem 264:1958–1962 167. Smith RM, Curnutte JT, Mayo LA, Babior BM (1989) Use of an affinity label to probe the function of the NADPH binding component of the respiratory burst oxidase of human neutrophils. J Biol Chem 264:12243–12248 168. Okamura N, Babior BM, Mayo LA, Peveri P, Smith RM, Curnutte JT (1990) The p67phox cytosolic peptide of the respiratory burst oxidase from human neutrophils. Functional aspects. J Clin Invest 85:1583–1587 169. Umei T, Babior BM, Curnutte JT, Smith RM (1991) Identification of the NADPH-binding subunit of the respiratory burst oxidase. J Biol Chem 266:6019–6022

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

170. Dang PM-C, Johnson JL, Babior BM (2000) Binding of nicotinamide dinucleotide phosphate to the tetratricopeptide repeat domains at the N-terminus of p67phox, a subunit of the leukocyte nicotinamide adenine dinucleotide phosphate oxidase. Biochemistry 39:3069–3075 171. Ge F, Guillory RJ (1994) NADPH-binding protein of the neutrophil superoxide-generating oxidase of guinea pigs. Proc Natl Acad Sci U S A 91:8622–8626 172. Hashida S, Yuzawa S, Suzuki NS et al (2004) Binding of FAD to cytochrome b558 is facilitated during activation of the phagocyte NADPH oxidase, leading to superoxide production. J Biol Chem 279:26378–26386 173. Sumimoto H, Sakamoto N, Nozaki M, Sakaki Y, Takeshige K, Minakami S (1992) Cytochrome b558, a component of the phagocyte NADPH oxidase, is a flavoprotein. Biochem Biophys Res Commun 186:1368–1375 174. Doussière J, Buzenet G, Vignais PV (1995) Photoaffinity labeling and photoinactivation of the O2- – generating oxidase of neutrophils by an azido derivative of FAD. Biochemistry 34: 1760–1770 175. Karplus PA, Daniels MJ, Herriott JR (1991) Atomic structure of ferredoxin-NADP+ reductase: prototype for a structurally novel flavoenzyme family. Science 251:60–66 176. Kean KM, Carpenter RA, Pandini V et al (2017) High-resolution studies of hydride transfer in the ferredoxin: NADP+ reductase. FEBS J 284:3302–3319 177. Dinauer MC, Curnutte JT, Rosen H, Orkin SH (1989) A missense mutation in the neutrophil cytochrome b heavy chain in cytochrome-positive X-linked chronic granulomatous disease. J Clin Invest 84:2012–2016 178. Yoshida LS, Saruta F, Yoshikawa K, Tatsuzawa O, Tsunawaki S (1998) Mutation at histidine 338 of gp91phox depletes FAD and affects expression of cytochrome b558 of the human NADPH oxidase. J Biol Chem 273:27879–27886 179. Taylor WR, Jones DT, Segal AW (1993) A structural model for the nucleotide binding domains of the flavocytochrome b-245 β-chain. Protein Sci 2:1675–1685 180. Beaumel S, Picciocchi A, Debeurne F et al (2017) Downregulation of Nox2 activity in phagocytes mediated by ATM-kinase dependent phosphorylation. Free Radic Biol Med 113:1–15 181. Leusen JW, de Boer M, Bolscher BGJM et al (1994) A point mutation in gp91phox of cytochrome b558 of the human NADPH oxidase leading to defective translocation of the cytosolic proteins p47phox and p67phox. J Clin Invest 93:2120–2126 182. Li XJ, Grunwald D, Mathieu J, Morel F, Stasia M-J (2005) Crucial role of two potential cytosolic regions of Nox2, 191 TSSTKTIRRS200 and 484DESQANHFAVHHDEEKD500, on NADPH oxidase activation. J Biol Chem 280:14962–14973 183. Kleinberg ME, Malech HL, Rotrosen D (1990) The phagocyte 47-kilodalton cytosolic oxidase protein is an early reactant in activation of the respiratory burst. J Biol Chem 265:15577–15583 184. Magnani F, Mattevi A (2019) Structure and mechanism of ROS generation by NADPH oxidases. Curr Opin Struct Biol 59:91–97 185. Deng Z, Aliverti A, Zanetti G et al (1999) A productive NADP+ binding mode of ferredoxin-NADP+ reductase revealed by protein engineering and crystallographic studies. Nat Struct Biol 6:847– 853 186. Babior BM (1999) NADPH oxidase: an update. Blood 93:1464– 1476 187. Koshkin V, Lotan O, Pick E (1997) Electron transfer in the superoxide-generating NADPH oxidase complex reconstituted in vitro. Biochim Biophys Acta 1319:139–146 188. Cross A (1999) The participation of the hemes of flavocytochrome b245 in the electron transfer process in NADPH oxidase (Letter to the Editor). Blood 93:4449

55

189. Babior BM (1999) Response (Letter to the Editor). Blood 93:4449 190. Nauseef WM, Clark RA (2019) Intersecting stories of the phagocyte NADPH oxidase and chronic granulomatous disease. In: Knaus UG, Leto TL (eds) NADPH oxidases methods and protocols. Springer Science+Business Media, New York, pp 3–16 191. Hamers MN, de Boer M, Meerhof LJ, Weening RS, Roos D (1984) Complementation in monocyte hybrids revealing genetic heterogeneity in chronic granulomatous disease. Nature 307:553–555 192. Segal AW, Heyworth PG, Cockroft S, Barrowman MM (1985) Stimulated neutrophils from patients with autosomal recessive chronic granulomatous disease fail to phosphorylate a Mr-44,000 protein. Nature 306:547–549 193. Heyworth PG, Segal AW (1986) Further evidence for the involvement of a phosphoprotein in the respiratory burst oxidase of neutrophils. Biochem J 239:723–731 194. Bromberg Y, Pick E (1984) Unsaturated fatty acids stimulate NADPH- dependent superoxide generation by cell-free system in macrophages. Cell Immunol 88:213–221 195. Heyneman RA, Vercauteren RE (1984) Activation of a NADPH oxidase from horse poymorphonuclear leukocytes in a cell-free system. J Leukoc Biol 36:751–759 196. Pick E, Keisari Y (1981) Superoxide anion production and hydrogen peroxide production by chemically elicited macrophages – induction by multiple nonphagocytic stimuli. Cell Immunol 59: 301–318 197. Pick E (2020) Cell-free NADPH oxidase activation assay: a triumph of reductionism. In: Quinn MT, DeLeo FR (eds) Neutrophil methods and protocols, 3rd edn. Springer Science+Business Media, New York, pp 325–411 198. Kakinuma K (1974) Effects of fatty acids on the oxidative metabolism of leukocytes. Biochim Biophys Acta 348:76–85 199. Kakinuma K, Minakami S (1978) Effects of fatty acids on superoxide radical generation in leukocytes. Biochim Biophys Acta 538: 50–59 200. Badwey JA, Curnutte JT, Karnovsky ML (1981) cis-polyunsaturated fatty acids induce high levels if superoxide production by human neutrophils. J Biol Chem 256:12640–12643 201. Bromberg Y, Pick E (1983) Unsaturated fatty acids as second messengers of superoxide generation by macrophages. Cell Immunol 79:243–252 202. Scott WA, Zrike JM, Hamill AL, Kempe J, Cohn ZA (1980) Regulation of arachidonic acid metabolites in macrophages. J Exp Med 152:324–335 203. Smolen JE, Shohet SB (1974) Remodeling of granulocyte membrane fatty acids during phagocytosis. J Clin Invest 53:726–734 204. McPhail LC, Clayton CC, Snyderman R (1984) A potential second messenger role for unsaturated fatty acids: activation of Ca2+dependent protein kinase. Science 224:622–625 205. Heyneman R (1981) Activation by oleate of a NADPH oxidase from polymorphonuclear leukocytes. Arch Int Physiol Biochim 89: B107 206. Heyneman RA, Bauwens-Monbaliou D (1981) Kinetics of nicotinamide adenine dinucleotides in oleate-stimulated polymorphonuclear leukocytes. FEBS Lett 127:87–90 207. Heyneman RA (1983) Subcellular localization and properties of the NAD(P)H oxidase from equine polymorphonuclear leukocytes. Enzyme 29:198–207 208. Maridonneau-Parini I, Tringale SM, Tauber AI (1986) Identification of distinct activation pathways of the human neutrophil NADPH-oxidase. J Immunol 137:2925–2929 209. Maridonneau-Parini I, Tauber AI (1986) Activation of NADPHOxidase by arachidonic acid involves phospholipase A2 in intact human neutrophils but not in the cell-free system. Biochem Biophys Res Commun 138:1099–1105 210. Henderson LM, Chappell JB, Jones OTG (1989) Superoxide generation is inhibited by phospholipase A2 inhibitors. Role of

56 phospholipsase A2 in the activation of the NADPH oxidase. Biochem J 264:249–255 211. Sakata A, Ida E, Tominaga M, Onoue K (1987) Arachidonic acid acts as an intracellular activator of NADPH oxidase in Fcγ receptor-mediated superoxide generation in macrophages. J Immunol 138:4353–4359 212. Dabrai D, van den Bogaart G (2021) The roles of phospholipase A2 in phagocytes. Front Cell Dev Biol 9:673502 213. Shmelzer Z, Haddad N, Admon E et al (2003) Unique targeting of cytosolic phospholipase A2 to plasma membranes mediated by the NADPH oxidase of phagocytes. J Cell Biol 162:683–692 214. Zhao X, Bey EA, Wientjes FB, Cathcart MK (2002) Cytosolic phospholipase A2 (cPLA2) regulation of human monocyte NADPH oxidase activity. cPLA2 affects translocation but not phosphorylation of p67phox and p47phox. J Biol Chem 277:25385–25392 215. Petry A, Weitnauer M, Görlach A (2010) Receptor activation of NADPH oxidases. Antioxid Redox Signal 13:467–487 216. McPhail LC, Snyderman R (1983) Activation of the respiratory burst enzyme in human polymorphonuclear leukocytes by chemoattractants and other soluble stimuli. Evidence that the same oxidase is activated by different transductional mechanisms. J Clin Invest 72:192–200 217. McPhail LC, Clayton CC, Snyderman R (1984) Evidence that activation of human neutrophil NADPH oxidase involves association of cytosolic factor with membrane components. Clin Res 32: 315a 218. McPhail LC, Shirley PS, Clayton CC, Snyderman R (1985) Activation of the respiratory burst enzyme from human neutrophils in a cell-free system. J Clin Invest 75:1735–1739 219. Curnutte JT (1984) Activation of human neutrophil NADPH oxidase by arachidonic acid in a cell-free system. Blood 64:66a 220. Curnutte JT (1985) Activation of human neutrophil nicotinamide adenine dinucleotide phosphate reduced (triphosphopyridine nucleotide, reduced) oxidase by arachidonic acid in a cell-free system. J Clin Invest 75:1740–1743 221. Curnutte JT, Kuver R, Scott PJ (1987) Activation of neutrophil NADPH oxidase in a cell-free system. Partial purification of the components and characterization of the activation process. J Biol Chem 262:5563–5569 222. Curnutte JT, Berkow RL, Roberts RL, Shurin SB, Scott PJ (1988) Chronic Granulomatous Disease due to a defect in the cytosolic factor required for nicotinamide dinucleotide phosphate oxidase activation. J Clin Invest 81:606–610 223. Curnutte PJ, Scott PJ, Babior BM (1989) Functional defect in neutrophil cytosols from two patients with autosomal recessive cytochrome-positive Chronic Granulomatous Disease. J Clin Invest 83:1236–1240 224. Tsunawaki S, Nathan CF (1986) Release of arachidonate and reduction of oxygen. Independent metabolic bursts of the mouse peritoneal macrophage. J Biol Chem 261:11563–11570 225. Badwey JA, Curnutte JT, Robinson JM, Berde CB, Karnovsky MJ, Karnovsky ML (1984) Effects of fatty acids on release of superoxide and on change of shape by human neutrophils. Reversibility by albumin. J Biol Chem 259:7870–7877 226. Steinbeck MJ, Robinson JM, Karnovsky MJ (1991) Activation of the neutrophil NADPH-oxidase by free fatty acids requires the ionized carboxyl group and partitioning into membrane lipid. J Leukoc Biol 49:360–368 227. Zatti M, Rossi F (1967) Relationship between glycolysis and respiration in surfactant-treated leucocytes. Biochim Biophys Acta 148:553–555 228. Graham RC, Karnovsky MJ, Shafer AW, Glass EA, Karnovsky ML (1967) Metabolic and morphological observations on the effect of surface-active agents on leukocytes. J Cell Biol 32:629–647 229. Cohen HJ, Chovaniec ME (1978) Superoxide generation by digitonin-stimulated guinea pig granulocytes. A basis for a

E. Pick continuous assay for monitoring superoxide production and for the study of the activation of the generating system. J Clin Invest 61:1081–1087 230. Pick E, Mizel D (1982) Role of transmethylation in the elicitation of an oxidative burst in macrophages. Cell Immunol 72:277–285 231. Kakinuma K, Hatae T, Minakami S (1976) Effect of ionic sites of surfactants on leukocyte metabolism. J Biochem 79:795–802 232. Washida N, Sagawa A, Tamoto K, Koyama J (1980) Comparative studies on superoxide production by polymorphonuclear leukocytes stimulated with various agents. Biochim Biophys Acta 631:371–379 233. Bromberg Y, Pick E (1985) Activation of NADPH-dependent superoxide production in a cell-free system by sodium dodecyl sulfate. J Biol Chem 260:13539–13545 234. Seifert R, Schultz G (1987) Fatty acid-induced activation of NADPH oxidase in plasma membranes of human neutrophils depends on neutrophil cytosol and is potentiated by stable guanine nucleotides. Eur J Biochem 162:563–569 235. Tanaka T, Kanegasaki S, Makino R, Iizuka T, Ishimura Y (1987) Saturated and trans-unsaturated fatty acids elicit high levels of superoxide generation in intact and cell-free preparations of neutrophils. Biochem Biophys Res Commun 144:606–612 236. Tanaka T, Makino R, Iizuka T, Ishimura Y, Kanegasaki S (1988) Activation by saturated and monounsaturated fatty acids of the O2--generating system in a cell-free preparation from neutrophils. J Biol Chem 263:13670–13676 237. Nishida S, Yoshida LS, Shimoyama T, Nunoi H, Kobayashi T, Tsunawaki S (2005) Fungal metabolite gliotoxin targets flavocytochrome b558 in the activation of the human neutrophil NADPH oxidase. Infect Immun 73:235–244 238. Souabni H, Thoma V, Bizouarn T et al (2012) trans arachidonic acid acid isomers inhibit NADPH-oxidase activity by direct interaction with enzyme components. Biochim Biophys Acta 1818: 2314–2324 239. Yamaguchi T, Kaneda M, Kakinuma K (1986) Effect of saturated and unsaturated fatty acids on the oxidative metabolism of human neutrophils. The role of calcium ion in the extracellular medium. Biochim Biophys Acta 861:440–446 240. Ligeti E, Doussière J, Vignais PV (1988) Activation of the O2- generating oxidase in plasma membrane from bovine polymorphonuclear neutrophils by arachidonic acid, a cytosolic factor of protein nature, and nonhydrolyzable analogues of GTP. Biochemistry 27:193–200 241. Pilloud (Dagher) M-C, Doussière J, Vignais PV (1989) Parameters of activation of the membrane-bound O2- – generating oxidase from bovine neutrophils in a cell-free system. Biochem Biophys Res Commun 159:783–790 242. Aharoni I, Pick E (1990) Activation of the superoxide-generating NADPH oxidase of macrophages by sodium dodecyl sulfate in a soluble cell-free system: evidence for involvement of a G protein. J Leukoc Biol 48:107–115 243. Dagher M-C, Pick E (2007) Opening the black box: learning from cell-free systems on the phagocyte NADPH oxidase. Biochimie 89: 1123–1132 244. Souabni H, Wien F, Bizouarn T, Houée-Levin C, Réfrégiers M, Baciou L (2017) The physicochemical properties of membranes correlate with the NADPH oxidase activity. Biochim Biophys Acta 1861:3520–3530 245. Massoud R, Bizouarn T, Houée-Levin C (2014) Cholesterol: a modulator of the phagocyte NADPH oxidase – a cell-free study. Redox Biol 3:16–24 246. Shpungin S, Dotan I, Abo A, Pick E (1989) Activation of the superoxide forming NADPH oxidase in a cell-free system by sodium dodecyl sulfate. Absolute lipid dependence of the solubilized enzyme. J Biol Chem 264:9195–9203

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

247. Foubert TR, Burritt JB, Taylor RM, Jesaitis AJ (2002) Structural changes are induced in human neutrophil cytochrome b by NADPH oxidase activators, LDS, SDS and arachidonate: Intermolecular resonance energy transfer between trisulfopyrenyl-wheat germ agglutinin and cytochrome b558. Biochim Biophys Acta 1567:221–231 248. Taylor RM, Foubert TR, Burritt JB et al (2004) Anionic amphiphile and phospholipid-induced conformational changes in human neutrophil flavocytochrome b observed by fluorescence resonance energy transfer. Biochim Biophys Acta 1663:201–213 249. Bellavite P, Corso F, Dusi S, Grzeskowiak M, Della-Bianca V, Rossi F (1988) Activation of NADPH-dependent superoxide production in plasma membrane extracts of pig neutrophils by phosphatidic acid. J Biol Chem 263:8210–8214 250. Leto TL (1999) The respiratory burst oxidase. In: Gallin JI, Snyderman R (eds) Inflammation: basic principles and clinical correlates, 3rd edn. Lippincott Williams and Wilkins, Philadelphia, pp 769–786 251. Groemping Y, Rittinger K (2005) Activation and assembly of the NADPH oxidase: a structural perspective. Biochem J 386:401–416 252. Sumimoto H (2008) Structure, regulation and evolution of Nox-family NADPH oxidases that produce reactive oxygen species. FEBS J 275:3249–3277 253. Tauber AI (1987) Protein kinase C and the activation of the human neutrophil NADPH-oxidase. Blood 69:711–720 254. Volpp BD, Nauseef WM, Clark RA (1988) Two cytosolic neutrophil oxidase components absent in autosomal chronic granulomatous disease. Science 242:1295–1297 255. Levy R, Rotrosen D, Nagauker O et al (1990) Induction of the respiratory burst in HL-60 cells. Correlation of function and protein expression. J Immunol 145:2595–2601 256. Fuchs A, Dagher M-C, Jouan A, Vignais PV (1994) Activation of the O2- – generating NADPH oxidase in a semi-recombinant cellfree system. Assessment of the function of Rac in the activation process. Eur J Biochem 226:587–595 257. Clark RA, Malech HL, Gallin JI et al (1989) Genetic variants of chronic granulomatous disease: prevalence of deficiencies of two cytosolic components of the NADPH oxidase system. N Engl J Med 321:647–652 258. Nunoi H, Rotrosen D, Gallin JI, Malech HL (1988) Two forms of autosomal chronic granulomatous disease lack distinct neutrophil cytosol factors. Science 242:1298–1301 259. Clark RA, Volpp BD, Leidal KG, Nauseef WM (1990) Two cytosolic components of the human neutrophil respiratory burst oxidase translocate to the plasma membrane during cell activation. J Clin Invest 85:714–721 260. Bolscher BGJM, van Zwieten R, Kramer IJM, Weening RS, Verhoeven AJ, Roos D (1989) A phosphoprotein of Mr 47,000, defective in autosomal chronic granulomatous disease, copurifies with one of two soluble components required for NADPH:O2 oxidoreductase activity in human neutrophils. J Clin Invest 83: 757–763 261. Bolscher BGJM, Denis SW, Verhoeven AJ, Roos D (1990) The activity of one soluble component of the cell-free NADPH:O2 oxidoreductase of human neutrophils depends on guanosine 5’-O-(3-thio)triphosphate. J Biol Chem 265:15782–15787 262. Teahan CG, Totty N, Casimir CM, Segal AW (1990) Purification of the 47 kDa phosphoprotein associated with the NADPH oxidase of human neutrophils. Biochem J 267:485–489 263. Tanaka T, Makino R, Ishimura Y (1992) Cytosolic components involved in porcine neutrophil oxidase activation. Purification of a 47-kilodalton protein and reconstitution of the activation system. J Biol Chem 267:1239–1244 264. Pilloud-Dagher M-C, Jouan A, Vignais PV (1992) Purification and properties of a functional 47-kilodalton cytosolic factor required

57

for NADPH-oxidase activation in bovine neutrophils. Biochem Biophys Res Commun 186:731–738 265. Tanaka T, Imajoh-Ohmi S, Kanegasaki S et al (1990) A 63-kilodalton cytosolic polypeptide involved in superoxide generation in porcine neutrophils. J Biol Chem 265:18717–18720 266. Pilloud-Dagher M-C, Vignais PV (1991) Purification and characterization of an oxidase activating factor of 63 kilodaltons from bovine neutrophils. Biochemistry 30:2753–2760 267. Lomax KJ, Leto TL, Nunoi H et al (1989) Recombinant 47-kilodalton cytosol factor restores NADPH oxidase in chronic granulomatous disease. Science 245:409–412 268. Vogel US, Dixon RA, Schaber MD et al (1988) Cloning of bovine GAP and its interaction with oncogenic ras p21. Nature 335:90–93 269. Duchesne M, Schweighoffer F, Parker F et al (1993) Identification of the SH3 domain GAP as an essential sequence for Ras-GAPmediated signaling. Science 259:525–528 270. Volpp BD, Nauseef WM, Donelson JE et al (1989) Cloning of the cDNA and functional expression of the 47-kilodalton cytosolic component of human neutrophil respiratory burst oxidase. Proc Natl Acad Sci U S A 86:7195–7199 271. Musacchio A, Gibson T, Lehto V-P, Saraste M (1992) SH3 – an abundant protein domain in search of a function. FEBS Lett 307: 55–61 272. Leto TL, Lomax KJ, Volpp BD et al (1990) Cloning of a 67-kD neutrophil oxidase factor with similarity to a noncatalytic region of p60c-src. Science 248:727–730 273. Ligeti E, Tardif M, Vignais PV (1989) Activation of O2•-generating oxidase of bovine neutrophils in a cell-free system. Interaction of a cytosolic factor with the plasma membrane and control by G nucleotides. Biochemistry 28:7116–7123 274. Pick E, Kroizman T, Abo A (1989) Activation of the superoxideforming NADPH oxidase of macrophages requires two cytosolic components – one of them is also present in certain nonphagocytic cells. J Immunol 143:4180–4187 275. Fujita I, Takeshige K, Minakami S (1987) Characterization of the NADPH-dependent superoxide production activated by sodium dodecyl sulfate in a cell-free system of pig neutrophils. Biochim Biophys Acta 931:41–48 276. Abo A, Pick E (1991) Purification and characterization of a third cytosolic component of the superoxide-generating NADPH oxidase of macrophages. J Biol Chem 266:23577–23585 277. Abo A, Pick E, Hall A, Totty N, Teahan CG, Segal AW (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21 rac 1. Nature 353:668–670 278. Pick E, Gorzalczany Y, Engel S (1993) Role of the rac1 p21 - rho GDI heterodimer in the activation of the superoxide forming NADPH oxidase of macrophages. Eur J Biochem 217:441–455 279. Didsbury J, Weber RF, Bokoch GM, Evans T, Snyderman R (1989) rac, a novel ras-related family of proteins that are botulinum toxin substrates. J Biol Chem 264:16378–16382 280. Ohga N, Kikuchi A, Ueda T et al (1989) Rabbit intestine contains a protein that inhibits the dissociation of GDP from and the subsequent binding of GTP to rhoB p20, a ras p21-like-GTPbinding protein. Biochem Biophys Res Commun 163:1523–1533 281. Olofsson B (1999) Rho guanine dissociation inhibitors: pivotal molecules in cellular signaling. Cell Signal 1:545–554 282. Knaus UG, Heyworth PG, Evans T, Curnutte JT, Bokoch GM (1991) Regulation of phagocyte oxygen radical production by the GTP-binding protein Rac 2. Science 254:1512–1515 283. Knaus UG, Heyworth PG, Kinsella BT, Curnutte JT, Bokoch GM (1992) Purification and characterization of Rac2. A cytosolic GTP-binding protein that regulates human neutrophil oxidase. J Biol Chem 267:23575–23582 284. Zhao X, Carnevale KA, Cathcart MK (2003) Human monocytes use Rac1, not Rac2, in the NADPH oxidase complex. J Biol Chem 278:40788–40792

58 285. Kim C, Dinauer MC (2001) Rac2 is an essential regulator of neutrophil nicotinamide adenine dinucleotide phosphate oxidase activation in response to specific signaling pathways. J Immunol 166:1223–1232 286. Zhang FL, Casey PJ (1996) Protein prenylation: molecular mechanisms and functional consequences. Annu Rev Biochem 54:241–269 287. Didsbury JR, Uhling RJ, Snyderman R (1990) Isoprenylation of the low molecular mass GTP-binding proteins Rac1 and Rac2: possible role in membrane localization. Biochem Biophys Res Commun 171:804–812 288. Kreck ML, Freeman JL, Abo A, Lambeth JD (1996) Membrane association of Rac is required for high activity of the respiratory burst oxidase. Biochemistry 35:15683–15692 289. Gorzalczany Y, Sigal N, Itan M, Lotan O, Pick E (2000) Targeting of Rac1 to the phagocyte membrane is sufficient for the induction of NADPH oxidase assembly. J Biol Chem 275:40073–40081 290. Gorzalczany Y, Alloul N, Sigal N, Weinbaum C, Pick E (2002) A prenylated p67phox-Rac1 chimera elicits NADPH-dependent superoxide production by phagocyte membranes in the absence of an activator and p47phox. Conversion of a pagan NADPH oxidase to monotheism. J Biol Chem 277:18605–18610 291. Sigal N, Gorzalczany Y, Pick E (2003) Two pathways of activation of the superoxide-generating NADPH oxidase of phagocytes in vitro – distinctive effects of inhibitors. Inflammation 27:147– 159 292. Yeung T, Grinstein S (2007) Lipid signaling and the modulation of surface charge during phagocytosis. Immunol Rev 219:17–36 293. Pick E (2010) Editorial: when charge is in charge – “Millikan” for leukocyte biologists. J Leukoc Biol 87:537–540 294. Mizuno T, Kaibuchi K, Ando S et al (1992) Regulation of the superoxide-generating NADPH oxidase by a small GTP-binding protein and its stimulatory and inhibitory GDP/GTP exchange proteins. J Biol Chem 267:10215–10218 295. Ando S, Kaibuchi K, Sasaki T et al (1992) Post-translational processing of rac p21s is important both for their interaction with the GDP/GTP exchange proteins and for their activation of NADPH oxidase. J Biol Chem 267:25709–25713 296. Shimizu H, Toma-Fukai S, Kontani K et al (2018) GEF mechanism revealed by the structure of SmgGDS-558 and farnesylated RhoA complex and its implication for a chaperone mechanism. Proc Natl Acad Sci U S A 115:9563–9568 297. Xu X, Wang Y, Barry DC et al (1997) Guanine nucleotide binding properties of Rac2 mutant proteins and analysis of the responsiveness to guanine nucleotide dissociation stimulator. Biochemistry 36:626–632 298. Mizrahi A, Molshanski-Mor S, Weinbaum C et al (2005) Activation of the phagocyte NADPH oxidase by Rac guanine nucleotide exchange factors in conjunction with ATP and nucleoside diphosphate kinase. J Biol Chem 280:3802–3811 299. Heyworth PG, Knaus UG, Xu X et al (1993) Requirement for posttranslational processing of Rac GTP-binding proteins for activation of human neutrophil NADPH oxidase. Mol Biol Cell 4:261– 269 300. Bokoch GM, Bohl BP, Chuang T-H (1994) Guanine nucleotide exchange regulates membrane translocation of Rac/Rho GTP-binding proteins. J Biol Chem 269:31674–31679 301. Sigal N, Gorzalczany Y, Sarfstein R et al (2003) The guanine nucleotide exchange factor Trio activates the phagocyte NADPH oxidase in the absence of GDP to GTP exchange – “The emperor’s new clothes”. J Biol Chem 278:4854–4861 302. Ugolev Y, Berdichevsky Y, Weinbaum C, Pick E (2008) Dissociation of Rac1(GDP)-RhoGDI complexes by the cooperative action of anionic liposomes containing phosphatidylinositol 3,4,5trisphosphate, Rac guanine nucleotide exchange factor, and GTP. J Biol Chem 283:22257–22271

E. Pick 303. Zheng Y (2001) Dbl family guanine nucleotide exchange factors. Trends Biochem Sci 26:724–732 304. Welch HCE, Coadwell WJ, Stephens LR, Hawkins PT (2003) Phosphoinositide 3-kinase-dependent activation of Rac. FEBS Lett 546:93–97 305. Pick E (2014) Role of the Rho GTPase Rac in the activation of the phagocyte NADPH oxidase. Outsourcing a key task. Small GTPases 5:e-27952:1–e-2795223 306. Ambruso DR, Knall C, Abell AN, Panepinto J et al (2000) Human neutrophil immunodeficiency syndrome is associated with Rac2 mutation. Proc Natl Acad Sci U S A 97:4654–4659 307. Williams DA, Tao W, Yang F et al (2000) Dominant negative mutation of the hematopoietic Rho GTPase, Rac2, is associated with human phagocytic immunodeficiency. Blood 96:1646–1654 308. Hsu A, Donkó A, Arrington ME et al (2019) Dominant activating Rac mutation with lymphopenia, immunodeficiency, and cytoskeletal defects. Blood 133:1977–1988 309. Stebbins CE, Galán JE (2000) Modulation of host signaling by a bacterial mimic: structure of the Salmonella effector SptP bound to Rac1. Mol Cell 6:1449–1460 310. Chuang T-H, Bohl BP, Bokoch GM (1993) Biologically active lipids are regulators of Rac-GDI complexation. J Biol Chem 268: 26206–26211 311. Dovas A, Couchman JR (2005) RhoGDI: multiple functions in the regulation of Rho family GTPases. Biochem J 390:1–9 312. Garcia-Mata R, Boulter E, Burridge K (2011) The “invisible hand”: regulation of Rho GTPases by RHOGDIs. Nat Rev Mol Cell Biol 12:493–504 313. Sasaki T, Kato M, Takai Y (1993) Consequences of weak interaction of rho GDI with the GTP-bound forms of rho p21 and rac p21. J Biol Chem 268:23959–23963 314. Grizot S, Faure J, Fieschi F et al (2001) Crystal structure of the Rac1-RhoGDI complex involved in NADPH oxidase activation. Biochemistry 40:10007–10013 315. Di-Poi N, Faure J, Grizot S et al (2001) Mechanism of NADPH oxidase activation by the Rac/Rho-GDI complex. Biochemistry 40: 10014–10022 316. Sawai T, Asada M, Nunoi H et al (1993) Combination of arachidonic acid and guanosine 5’-O-(3-thiotriphosphate) induce translocation of rac p21s to membrane and activation of NADPH oxidase in a cell-free system. Biochem Biophys Res Commun 195: 264–269 317. Abo A, Webb MR, Grogan A, Segal AW (1994) Activation of NADPH oxidase involves the dissociation of p21rac from its inhibitory GDP/GTP exchange protein (rhoGDI) followed by its translocation to the plasma membrane. Biochem J 298:585–591 318. Quinn MT, Evans T, Loetterle LR, Jesaitis AJ, Bokoch GM (1991) Translocation of Rac correlates with NADPH oxidase activation. Evidence for equimolar translocation of oxidase components. J Biol Chem 268:20983–20987 319. Fauré J, Vignais PV, Dagher M-C (1999) Phosphoinositidedependent activation of Rho A involves partial opening of the RhoA/Rho-GDI complex. Eur J Biochem 262:879–889 320. Philips MR, Pillinger MH, Staud R et al (1993) Carboxyl methylation of Ras-related proteins during signal transduction in neutrophils. Science 259:977–980 321. Heyworth PG, Bohl BP, Bokoch GM, Curnutte JT (1994) Rac translocates independently of the neutrophil NADPH oxzidase components p47phox and p67phox. Evidence for its interaction with flavocytochrome b558. J Biol Chem 269:30749–30752 322. Robbe K, Otto-Bruc A, Chardin P, Antonny B (2003) Dissociation of GDP dissociation inhibitor and membrane translocation are required for efficient activation of Rac by the Dbl homologypleckstrin homology region of Tiam. J Biol Chem 278:4756–4762 323. Dansart E, Olofsson B, Cherfils J (2005) RhoGDI revisited: novel roles in Rho regulation. Traffic 6:957–966

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

324. Ugolev Y, Molshanski-Mor S, Weinbaum C, Pick E (2006) Liposomes comprising anionic but not neutral phospholipids cause dissociation of Rac (1 or 2)-RhoGDI complexes and support amphiphile-independent NADPH oxidase activation by such complexes. J Biol Chem 281:19204–19219 325. Perisic O, Wilson MI, Karathanasis D et al (2004) The role of phosphoinositides and phosphorylation in regulation of NADPH oxidase. Adv Enzyme Regul 44:279–298 326. Missy K, Van Poucke V, Raynal P et al (1998) Lipid products of phosphoinositide 3-kinase interact with Rac1 GTPase and stimulate GDP dissociation. J Biol Chem 273:30279–30286 327. Bromberg Y, Shani E, Joseph G et al (1994) The GDP-bound form of the small G protein Rac1 p21 is a potent activator of the superoxide-forming NADPH oxidase of macrophages. J Biol Chem 269:7055–7058 328. Toporik A, Gorzalczany Y, Hirschberg M et al (1998) Mutational analysis of novel effector domains in Rac1 involved in the activation of nicotinamide adenine dinucleotide phosphate (reduced) oxidase. Biochemistry 37:7147–7156 329. Rotrosen D, Kleinberg ME, Nunoi H et al (1990) Evidence for a functional cytoplasmic domain of phagocyte oxidase cytochrome b558. J Biol Chem 265:8745–8750 330. Nakanishi A, Imajoh-Ohmi S, Fujinawa T et al (1992) Direct evidence for interaction between COOH-terminal regions of cytochrome b558 subunits and cytosolic 47-kDa protein during activation of an O2- – generating system in neutrophils. J Biol Chem 267: 19072–19074 331. DeLeo FR, Yu L, Burritt JB et al (1995) Mapping sites of interaction of p47phox and flavocytochrome b with random-sequence peptide phage display libraries. Proc Natl Acad Sci U S A 92: 7110–7114 332. Davis A, Mascolo PL, Bunger PL et al (1998) Cloning and sequencing of the bovine flavocytochrome b subunit proteins, gp91phox and p22phox: comparison with other known flavocytochrome b sequences. J Leukoc Biol 64:114–123 333. Yu L, Cross AR, Zhen L, Dinauer MC (1999) Functional analysis of NADPH oxidase in granulocytic cells expressing a Δ488-497 gp91phox deletion mutant. Blood 94:2497–2504 334. Rey FE, Cifuentes ME, Kiarash A et al (2001) Novel competitive inhibitor of NAD(P)H oxidase assembly attenuates vascular O2and systolic blood pressure in mice. Circ Res 89:408–414 335. Jackson H, Kawahara T, Nisimoto Y, Smith SME, Lambeth JD (2010) Nox4 B-loop creates an interface between the transmembrane and dehydrogenase domains. J Biol Chem 285:10281–10290 336. Diebold BA, Bokoch GM (2001) Molecular basis for Rac2 regulation phagocyte NADPH oxidase. Nat Immunol 2:211–215 337. Kao Y-Y, Gianni D, Bohl B, Taylor RM, Bokoch GM (2008) Identification of a conserved Rac-binding site on NADPH oxidases supports a direct GTPase regulatory mechanism. J Biol Chem 283: 12736–12746 338. Diebold BA, Fowler B, Lu J, Dinauer MC, Bokoch GM (2004) Antagonistic cross-talk between Rac and Cdc42 GTPase regulates generation of reactive oxygen species. J Biol Chem 279:28136– 28142 339. Bokoch GM, Zhao T (2006) Regulation of the phagocyte NADPH oxidase by Rac GTPase. Antioxid Redox Signal 8:1533–1548 340. Alloul N, Gorzalczany Y, Itan M, Sigal N, Pick E (2001) Activation of the superoxide-generating NADPH oxidase by chimeric proteins consisting of segments of the cytosolic component p67phox and the small GTPase Rac1. Biochemistry 40:14557– 14566 341. Miyano K, Koga H, Minakami R, Sumimoto H (2009) The insert region of the Rac GTPases is dispensable for activation of superoxide-producing NADPH oxidases. Biochem J 422:373–382 342. Dang PM-C, Cross AR, Babior BM (2001) Assembly of the neutrophil respiratory burst oxidase: A direct interaction between

59

p67phox and cytochrome b558. Proc Natl Acad Sci U S A 98:3001– 3005 343. Dang PM-C, Cross AR, Quinn MT, Babior BM (2002) Assembly of the neutrophil respiratory burst oxidase: a direct interaction between p67phox and cytochrome b558 II. Proc Natl Acad Sci U S A 99:4262–4265 344. Freeman JL, Lambeth JD (1996) NADPH oxidase activity is independent of p47phox in vitro. J Biol Chem 271:22578–22582 345. Koshkin V, Lotan O, Pick E (1996) The cytosolic component p47phox is not a sine qua non participant in the activation of NADPH oxidase but is required for optimal superoxide production. J Biol Chem 271:30326–30329 346. Kuhns DB, Hsu AP, Sun D et al (2019) NCF1 (p47phox)-deficient chronic granulomatous disease: comprehensive genetic and flow cytometric analysis. Blood Adv 3:136–147 347. Patino P, Rae J, Noack D et al (1999) Molecular characterization of autosomal recessive chronic granulomatous disease caused by a defect of the nicotinamide adenine dinucleotide phosphate (reduced form) oxidase component p67phox. Blood 94:2505–2514 348. Köker MY, Camcioğlu Y, van Leeuwen K et al (2013) Clinical, functional, and genetic characterization of chronic granulomatous disease in 89 Turkish patients. J Allergy Clin Immunol 132:1156– 1163 349. Miyano K, Ogasawara S, Han C-H, Fukuda H, Tamura M (2001) A fusion protein between Rac and p67phox (1-210) reconstitutes NADPH oxidase with higher activity and stability than the individual components. Biochemistry 40:14089–14097 350. Sarfstein R, Gorzalczany Y, Mizrahi A et al (2004) Dual role of Rac in the assembly of NADPH oxidase: Tethering to the membrane and activation of p67phox. A study based on mutagenesis of p67phox-Rac1 chimeras. J Biol Chem 279:16007–16016 351. Mizrahi A, Berdichevsky Y, Ugolev Y et al (2006) Assembly of the phagocyte NADPH oxidase complex: chimeric constructs derived from the cytosolic components as tools for exploring structure-function relationships. J Leukoc Biol 79:881–895 352. Bedard K, Krause K-H (2007) The Nox family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313 353. Daniels RH, Bokoch GM (1999) p21-activated kinase: a crucial component of morphological signaling? Trends Biochem Sci 24: 350–355 354. Manser E, Leung T, Salihuddin H et al (1994) A brain serine/ threonine protein kinase activated by cdc42 and Rac1. Nature 367:40–46 355. Bagrodia S, Taylor SJ, Creasy CL et al (1995) Identification of a mouse p21Cdc42/Rac activated kinase. J Biol Chem 270:22731– 22737 356. Burbelo PD, Drechsel D, Hall A (1995) A conserved binding motif defines numerous candidate target proteins for both Cdc42 and Rac GTPases. J Biol Chem 270:29071–29074 357. Aspenström P, Lindberg U, Hall A (1996) Two GTPases, CDC42 and Rac, bind directly to a protein implicated in the immunodeficiency disorder Wiskott-Aldrich syndrome. Curr Biol 6:70–75 358. Kim AS, Kakalis LT, Abdul-Manan N et al (2000) Autoinhibition and activation mechanisms of the Wiskott-Aldrich syndrome protein. Nature 404:151–158 359. Rudolph MG, Bayer P, Abo A et al (1998) The Cdc42/Rac interactive binding region motif of the Wiskott Aldrich syndrome protein (WASP) is necessary but not sufficient for tight binding to Cdc42 and structure formation. J Biol Chem 273:18067–18076 360. Diekmann D, Abo A, Johnston C, Segal AW, Hall A (1994) Interaction of Rac with p67phox and regulation of phagocytic NADPH oxidase activity. Science 265:531–533 361. de Mendez I, Garrett MC, Adams AG, Leto TL (1994) Role of p67phox SH3 domains in assembly of the NADPH oxidase system. J Biol Chem 269:16326–16332

60 362. Diekmann D, Nobes CD, Burbelo PD, Abo A, Al H (1995) Rac GTPase interacts with GAPs and target proteins through multiple effector sites. EMBO J 14:5297–5305 363. Pik N, Toporik A, Pugach N, Pick E, Lotan O (1999) The p21-binding domain of p21-activated kinase (PAK-3) inhibits NADPH oxidase activation by competition with p67phox. Eur J Clin Investig 29(Suppl 1):42 364. Kwong CH, Adams AG, Leto TL (1995) Characterization of the effector-specifying domain of Rac involved in NADPH oxidase activation. J Biol Chem 270:19868–19872 365. Freeman JLR, Kreck ML, Uhlinger DJ, Lambeth JD (1994) Ras effector-homologue region on Rac regulates protein association in the neutrophil respiratory oxidase complex. Biochemistry 33: 13431–13435 366. Nisimoto Y, Freeman JLR, Motalebi SZ et al (1997) Rac binding to p67phox. Structural basis for interactions of the Rac1 effector region and insert region with components of the respiratory burst oxidase. J Biol Chem 272:18834–18841 367. Xu X, Barry DC, Settleman J, Schwartz MA, Bokoch GM (1994) Differing structural requirements for GTPase-activating protein responsiveness and NADPH oxidase activation by Rac. J Biol Chem 269:23569–23574 368. Sehr P, Joseph G, Genth H et al (1998) Glucosylation and ADP ribosylation of Rho proteins: effects on nucleotide binding, GTPase activity, and effector coupling. Biochemistry 37:5296– 5304 369. Prigmore E, Ahmed S, Best A et al (1995) A 68-kDa kinase and NADPH oxidase component p67phox are targets for CDC42Hs and Rac1 in neutrophils. J Biol Chem 270:10717–10722 370. Ahmed S, Prigmore E, Govind S et al (1998) Cryptic Rac-binding and p21Cdc42HS/Rac-activated kinase phosphorylation sites of NADPH oxidase component p67phox. J Biol Chem 273:15693– 15701 371. Koga H, Terasawa H, Nunoi H, Takeshige K, Inagaki F, Sumimoto H (1999) Tetratricopeptide repeat (TPR) motifs on p67phox participate in interaction with the small GTPase Rac and activation of the NADPH oxidase. J Biol Chem 274:25051–25060 372. Leusen JHW, de Klein A, Hilarius PM et al (1996) Disturbed interaction of p21-rac with mutated p67phox causes chronic granulomatous disease. J Exp Med 289:1243–1249 373. Lapouge K, Smith SM, Walker PA et al (2000) Structure of the TPR domain of p67phox in complex with Rac-GTP. Mol Cell 6: 899–907 374. Diekmann D, Hall A (1995) In vitro binding assay for interactions of Rho add Rac with GTPase-activating proteins and effectors. Meth Enzymol 256:207–215 375. Grizot S, Fieschi F, Dagher M-C, Pebay-Peyroula E (2001) The active N-terminal region of p67phox. Structure at 1.8 Å resolution and biochemical characterizations of the A128V mutant implicated in chronic granulomatous disease. J Biol Chem 276:21627–21631 376. Han C-H, Freeman JLR, Le T, Motalebi SA, Lambeth JD (1998) Regulation of the neutrophil respiratory burst oxidase. Identification of an activation domain in p67phox. J Biol Chem 273:16663– 16668 377. Bechor E, Zahavi A, Berdichevsky Y, Pick E (2021) The molecular basis of Rac-GTP action – promoting binding of p67phox to Nox2 by disengaging the β hairpin from downstream residues. J Leukoc Biol 110:219–237 378. Miyano K, Ueno N, Takeya R, Sumimoto H (2006) Direct involvement of the small GTPase Rac in activation of the superoxideproducing NADPH oxidase Nox1. J Biol Chem 281:21857–21868 379. Bosco E, Marchioni F, Kumar S et al (2012) Rational design of small molecule inhibitors targeting the Rac GTPase – p67phox signaling axis in inflammation. Chem Biol 19:228–242

E. Pick 380. Diebold BA, Smith SMS, Li Y, Lambeth JD (2015) NOX2 as a target for drug development: indications, possible complications, and progress. Antioxid Redox Signal 23:375–405 381. Sun J (2020) Structures of mouse DUOX1-DUOXA1 provide mechanistic insights into enzyme activation and regulation. Nat Struct Mol Biol 27:1086–1093 382. Wu J-X, Liu R, Song K, Chen L (2021) Structures of human dual oxidase 1 complex in low-calcium and high-calcium states. Nat Commun 12:155 383. Cross AR, Yarchover JL, Curnutte JT (1994) The superoxidegenerating system of human neutrophils possesses a novel diaphorase activity. Evidence for distinct regulation of electron flow within NADPH oxidase by p47phox and p67phox. J Biol Chem 269:21448–21454 384. Cross AR, Curnutte JT (1995) The cytosolic activating factors p47phox and p67phox have distinct roles in the regulation of electron flow in NADPH oxidase. J Biol Chem 270:6543–6548 385. Nisimoto Y, Motalebi S, Han C-H, Lambeth JD (1999) The p67phox activation domain regulates electron flow from NADPH to flavin in flavocytochrome b558. J Biol Chem 274:22999–23005 386. Carrichon L, Picciocchi A, Debeurne F et al (2011) Characterization of superoxide overproduction by the D-loopNox4-Nox2 cytochrome b558 in phagocytes – differential sensitivity to calcium and phosphorylation events. Biochim Biophys Acta 1808:78–90 387. Subramaniam S, Kleywegt GJ (2022) A paradigm shift in structural biology. Nat Methods 19:20–23 388. Moser CC, Chobot SE, Page CC, Dutton PL (2008) Distance metrics for heme protein electron tunneling. Biochim Biophys Acta 1777:1032–1037 389. Moser CM, Anderson JLR, Dutton PL (2010) Guidelines for tunneling in enzymes. Biochim Biophys Acta 1787:1573–1586 390. Trixler F (2013) Quantum tunneling to the origin and evolution of life. Curr Org Chem 17:1758–1770 391. Cross AR, Segal AW (2004) The NADPH oxidase of professional phagocytes – prototype of the NOX electron transport system. Biochim Biophys Acta 1657:1–22 392. Wu X, Hénin J, Baciou L et al (2021) Mechanistic insights on heme-to-heme transmembrane electron transfer within NADPH oxydases from atomistic simulations. Front Chem 9:650651 393. Carlson ML, Young JW, Zhao Z et al (2018) The peptidisc, a simple method for stabilizing membrane proteins in detergentfree solution. elife 7:e34085 394. Wang M, Roberts DL, Paschke R et al (1997) Three-dimensional structure of NADPH-cytochrome P450 reductase: prototype for FMN- and FAD-containing enzymes. Proc Natl Acad Sci U S A 94:8411–8416 395. Hubbard PA, Shen AL, Paschke R et al (2001) NADPHcytochrome P450 oxidoreductase – structural basis for hydride and electron transfer. J Biol Chem 276:29163–29170 396. Zhen L, Yu L, Dinauer MC (1998) Probing the role of the carboxyl terminus of the gp91phox subunit of neutrophil flavocytochrome b558 using site-directed mutagenesis. J Biol Chem 273:6575–6581 397. Orellano EG, Calcaterra NB, Carrillo N, Ceccarelli EA (1993) Probing the role of the carboxyl-terminal region of ferredoxinNADP+ reductase by site-directed mutagenesis and deletion analysis. J Biol Chem 268:19267–19273 398. Xu W, Harison SC, Eck MJ (1997) Three-dimensional structure of the tyrosine kinase c-Src. Nature 385:595–602 399. Belambri SA, Rolas L, Raad H et al (2018) NADPH oxidase activation in neutrophils: role of the phosphorylation of its subunits. Eur J Clin Investig 48(Suppl 2):e12951 400. Mayer BC, Eck MJ (1995) Minding your p’s and q’s. Curr Biol 5: 364–367 401. Finan P, Shimizu Y, Gout I et al (1994) An SH3 domain and proline-rich sequence mediate interaction between two

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

components of the phagocyte NADPH oxidase complex. J Biol Chem 269:13752–13755 402. Leusen JHW, Bolscher BGJM, Hilarius PM et al (1994) 156Pro → Gln substitution in the light chain of cytochrome b558 of the human NADPH oxidase (p22phox) leads to defective translocation of the cytosolic proteins p47phox and p67phox. J Exp Med 180:2329–2334 403. Sumimoto H, Hata K, Mizuki K et al (1996) Assembly and activation of the phagocyte NADPH oxidase. Specific interaction of the N-terminal Src homology domain 3 of p47phox with p22phox is required for activation of the NADPH oxidase. J Biol Chem 271: 22152–22158 404. Ago T, Nunoi H, Ito T, Sumimoto H (1999) Mechanism for phosphorylation –induced activation of the phagocyte NADPH oxidase protein p47phox. Triple replacement of serines 303, 304, and 328 with aspartates disrupts the SH3 domain-mediated intramolecular interaction in p47phox, thereby activating the oxidase. J Biol Chem 274:33644–33653 405. Inanami O, Johnson JL, McAdara JK et al (1998) Activation of the leukocyte NADPH oxidase by phorbol ester requires the phosphorylation pf p47phox on serine 303 or 304. J Biol Chem 273:9539– 9543 406. Huang J, Kleinberg ME (1999) Activation of the phagocyte NADPH oxidase protein p47phox. Phosphorylation controls SH3 domain-dependent binding to p22phox. J Biol Chem 274:19731– 19737 407. Groemping Y, Lapouge K, Smerdon SJ, Rittinger K (2003) Molecular basis of phosphorylation-induced activation of the NADPH oxidase. Cell 113:343–355 408. Autore F, Pagano B, Fomili A et al (2010) In silico phosphorylation of the autoinhibited form of p47phox: insights into the mechanism of activation. Biophys J 99:3716–3725 409. Karathanassis D, Stahelin R, Bravo J et al (2002) Binding of the PX domain of p47phox to phosphatidylinositol 3,4-bisphosphate and phosphatidic acid is masked by an intramolecular interaction. EMBO J 21:5057–5068 410. Hiroaki H, Ago T, Ito T et al (2001) Solution structure of the PX domain, a target of the SH3 domain. Nat Struct Biol 8:526–530 411. Ago T, Kuribayashi F, Hiroaki H et al (2003) Phosphorylation of p47phox directs phox homology domain from SH3 toward phosphoinositides, leading to phagocyte NADPH oxidase activation. Proc Natl Acad Sci U S A 100:4474–4479 412. Marcoux J, Man P, Petit-Haertlein I et al (2010) p47phox molecular activation for assembly of the neutrophil NADPH oxidase complex. J Biol Chem 285:28980–28990 413. Bánfi B, Clark RA, Steger K, Krause K-H (2003) Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 278:3510–3513 414. Takeya R, Ueno N, Kami K et al (2003) Novel human homologues of p47phox participate in activation of superoxide-producing NADPH oxidases. J Biol Chem 278:25234–25246 415. Yamamoto A, Kami K, Takeya R, Sumimoto H (2007) Interaction between the SH3 domains and C-terminal proline-rich region in NADPH oxidase organizer 1 (Noxo1). Biochem Biophys Res Commun 352:560–565 416. Dutta S, Rittinger K (2010) Regulation of NOXO1 activity through reversible interactions with p22phox and NOXA1. PLoS One 5: e10478 417. Joseph G, Pick E (1995) “Peptide walking” is a novel method of mapping functional domains in proteins. Its application to the Rac1-dependent activation of NADPH oxidase. J Biol Chem 270: 29079–29082 418. Dahan I, Issaeva I, Gorzalczany Y et al (2002) Mapping of functional domains in the p22phox subunit of flavocytochrome b559 participating in the assembly of the NADPH oxidase complex by “peptide walking”. J Biol Chem 277:8421–8432

61

419. Morozov I, Lotan O, Joseph G et al (1998) Mapping of functional domains in p47phox involved in the activation of NADPH oxidase by “pepide walking”. J Biol Chem 273:15435–15444 420. Pick E (2019) Using synthetic peptides for exploring proteinprotein interactions in the assembly of the NADPH oxidase complex. In: Knaus UG, Leto TL (eds) NADPH oxidases methods and protocols. Springer Science+Business Media, New York, pp 377–415 421. Solbak SMØ, Zang J, Narayanan D et al (2020) Developing inhibitors of the p47phox-p22phox protein interaction by fragmentbased drug discovery. J Med Chem 63:1156–1177 422. Garsi JB, Komjáti B, Cullia G et al (2022) Targeting NOX2 via p47/phox - p22/phox inhibition with novel triproline mimetics. ACS Med Chem Lett 13:949–954 423. Cox JA, Jeng AY, Sharkey NA et al (1985) Activation of the human neutrophil nicotinamide dinucleotide phosphate (NADPH)-oxidase by protein kinase C. J Clin Invest 76:1932– 1938 424. Fontayne A, Dang PM-C, Gougerot-Pocidalo MA, El Benna J (2002) Phosphorylation of p47phox sites by PKC α, βII, δ, and ζ: effect on binding to p22phox and on NADPH oxidase activation. Biochemistry 41:7743–7750 425. Cox J, Jeng AY, Blumberg PM, Tauber AI (1987) Comparison of subcellular activation of the human neutrophil NADPH-oxidase by arachidonic acid, sodium dodecyl sulfate (SDS) and phorbol myristate acetate (PMA). J Immunol 138:1884–1888 426. Tauber AI, Cox JA, Curnutte JT et al (1989) Activation of human neutrophil NADPH oxidase in vitro by the catalytic fragment of protein kinase C. Biochem Biophys Res Commun 158:884–890 427. El Benna J, Park J-W, Ruedi JM, Babior BM (1995) Cell-free activation of the respiratory burst oxidase by protein kinase C. Blood Cells Mol Dis 21:201–206 428. Park J-W, Hoyal CR, El Benna J, Babior BM (1997) Kinasedependent activation of the leukocyte NADPH oxidase in a cellfree system. Phosphorylation of membranes and p47phox during oxidase activation. J Biol Chem 272:11035–11043 429. Park J-W, Scott KE, Babior BM (1998) Activation of the leukocyte NADPH oxidase in a cell-free system: phosphorylation vs. amphiphiles. Exp Hematol 26:37–44 430. Faust LRP, El Benna J, Babior BM, Chanock SJ (1995) The phosphorylation targets of p47phox, a subunit of the respiratory burst oxidase. Functions of the individual target serines as evaluated by site-directed mutagenesis. J Clin Invest 96:1499– 1505 431. Meijles DN, Fan LM, Howlin BJ, Li J-M (2014) Molecular insights of p47phox phosphorylation dynamics in the regulation of NADPH oxidase activation and superoxide production. J Biol Chem 289: 22759–22770 432. Rossetti Lopes L, Hoyal CR, Knaus UG, Babior BM (1999) Activation of the leukocyte NADPH oxidase by protein kinase C in a partially recombinant cell-free system. J Biol Chem 274: 15533–15537 433. El Benna J, Faust LRP, Babior BM (1984) The phosphorylation of the respiratory burst oxidase component p47phox during neutrophil activation. Phosphorylation of sites recognized by protein kinase C and by proline-directed kinases. J Biol Chem 269:23431–23436 434. Swain SD, Helgerson SL, Davis AR, Nelson LK, Quinn MT (1997) Analysis of activation-induced conformational changes in p47phox using tryptophan fluorescence spectroscopy. J Biol Chem 272:29502–29510 435. Park J-W, Babior BM (1998) Activation of the leukocyte NADPH oxidase subunit p47phox by protein kinase C. A phosphorylationdependent change in the conformation of the C-terminal end of p47phox. Biochemistry 36:7474–7480

62 436. Park H-S, Park J-W (1998) Fluorescent labeling of the leukocyte NADPH oxidase subunit p47phox: evidence for amphiphile-induced conformational changes. Arch Biochem Biophys 360:165–172 437. Bizouarn T, Karimi G, Masoud R et al (2016) Exploring the arachidonic acid-induced structural changes in phagocyte NADPH oxidase p47phox and p67phox via thiol accessibility and SCRD spectroscopy. FEBS J 283:2896–2910 438. Shiose A, Sumimoto H (2000) Arachidonic acid and phosphorylation induce a conformational change of p47phox to activate the phagocyte NADPH oxidase. J Biol Chem 275:13793–13801 439. Burnham DN, Uhlinger DJ, Lambeth JD (1990) Diradylglycerol synergizes with an anionic amphiphile to activate superoxide generation and phosphorylation of p47phox in a cell-free system from human neutrophils. J Biol Chem 265:17550–17559 440. Hata K, Ito T, Takeshige K, Sumimoto H (1998) Anionic amphiphile-independent activation of the phagocyte NADPH oxidase in a cell-free system by p47phox and p67phox, both in C terminally truncated forms. Implications for regulatory Src homology 3 domain-mediated interactions. J Biol Chem 273:4232–4236 441. Peng G, Huang J, Boyd M, Kleinberg ME (2003) Properties of phagocyte NADPH oxidase p47phox mutants with unmasked SH3 (Src homology 3) domains: full reconstitution of oxidase activity in a semi-recombinant cell-free system lacking arachidonic acid. Biochem J 373:221–229 442. Mizrahi A, Berdichevsky Y, Casey PJ, Pick E (2010) A prenylated p47phox-p67phox-Rac1 chimera is a quintessential NADPH oxidase activator. Membrane association and functional capacity. J Biol Chem 285:25485–25499 443. McLaughlin S, Aderem A (1995) The myristoyl-electrostatic switch: a modulator of reversible protein-membrane interactions. Trends Biochem Sci 20:272–276 444. Ebisu K, Nagasawa T, Watanabe K, Kakinuma K, Miyano K, Tamura M (2001) Fused p47phox and p67phox truncations efficiently reconstitute NADPH oxidase with higher activity than the individual components. J Biol Chem 276:24498–24505 445. Tamura M, Tamura T, Tyagi SR, Lambeth JD (1988) The superoxide-generating respiratory burst oxidase of human neutrophil plasma membrane. Phosphatidylserine as an effector of the activated enzyme. J Biol Chem 263:17621–17626 446. Berdichevsky Y, Mizrahi A, Ugolev Y, Molshanski-Mor S, Pick E (2007) Tripartite chimeras comprising functional domains derived from the three cytosolic components p47phox, p67phox and Rac1 elicit activator-independent superoxide production by phagocyte membranes. Role of membrane lipid charge and of specific residues in the chimeras. J Biol Chem 282:22122–22139 447. Ellson DE, Gobert-Gosse S, Anderson KE et al (2001) PtdIns(3)P regulates the neutrophil oxidase complex by binding to the PX domain of p40phox. Nat Cell Biol 3:679–682 448. Matute JD, Arias AA, Dinauer MC, Patiño PJ (2005) p40phox: the last NADPH oxidase subunit. Blood Cells Mol Dis 35:291–302 449. Yeung T, Terebiznik M, Yu L et al (2006) Receptor activation alters inner surface potential during phagocytosis. Science 313: 347–351 450. Debant A, Sera-Pagès C, Seipel K et al (1996) The multidomain protein Trio binds the LAR transmembreane tyrosine phosphatase, contains a protein kinase domain, and has separate rac-specific and rho-specific guanine nucleotide exchange factor domains. Proc Natl Acad Sci U S A 93:5466–5471 451. Zhang B, Zhang Y, Wang Z-X, Zheng Y (2000) The role of Mg2+ cofactor in the guanine nucleotide exchange and GTP hydrolyisis reactions of Rho family GTP-binding proteins. J Biol Chem 275: 25299–25307 452. Worthylake DK, Rossman KL, Sondek J (2000) Crystal structure of Rac1 in complex with the guanine nucleotide exchange region of Tiam1. Nature 408:682–688

E. Pick 453. Gao Y, Xing J, Streuli M, Leto TL, Zheng Y (2001) Trp56 specifies interaction with a subset of guanine nucleotide exchange factors. J Biol Chem 276:47530–47541 454. Dang PM-C, Morel F, Gougerot-Pocidalo M-A, El Benna J (2003) Phosphorylation of the NADPH oxidase component p67phox by ERK2 and P38MAPK: selectivity of phosphorylated sites and existence of an intramolecular regulatory domain in the tetratricopeptide-rich region. Biochemistry 42:4520–4526 455. Lapouge K, Smith SJM, Groemping Y, Rittinger K (2002) Architecture of the p40-p47-p67phox complex in the resting state of the NADPH oxidase. A central role for p67phox. J Biol Chem 277: 10121–10128 456. Durand D, Vivès C, Cannella D et al (2010) NADPH oxidase activator p67phox behaves in solution as a multidomain protein with semi-flexible linkers. J Struct Biol 169:45–53 457. Maehara Y, Miyano K, Sumimoto H (2009) Role of the first SH3 domain of p67phox in activation of superoxide-producing NADPH oxidases. Biochem Biophys Res Commun 379:589–593 458. de Mendez I, Adams AG, Sokolic RA et al (1996) Multiple SH3 domain interactions regulate NADPH oxidase assembly in whole cells. EMBO J 15:1211–1220 459. Valente AJ, El Jamali A, Epperson TK et al (2007) NOX1 NADPH oxidase regulation by the NOXA1 SH3 domain. Free Radic Biol Med 43:384–396 460. Kawano M, Miyamoto K, Kaito Y et al (2012) Noxa1 as a moderate activator of Nox2-based NADPH oxidase. Arch Biochem Biophys 519:1–7 461. Kawano M, Ishii R, Yoshioka Y et al (2013) C-terminal truncation of NoxA1 greatly enhances its ability to activate Nox2 in a pure reconstitution system. Arch Biochem Biophys 538:164–170 462. Bechor E, Zahavi A, Amichay M et al (2020) p67phox binds to a newly identified site in Nox2 by following disengagement of an intramolecular bond – Canaan sighted? J Leukoc Biol 107:509– 528 463. Dahan I, Smith SME, Pick E (2015) A Cys-Gly-Cys triad in the dehydrogenase region of Nox2 plays a key role in the interaction with p67phox. J Leukoc Biol 98:859–874 464. Jumper J, Evans R, Pritzel A et al (2021) Highly accurate protein structure prediction with AlphaFold. Nature 596:583–596 465. Varadi M, Anyango S, Deshpande M et al (2022) AlphaFold protein structure database: massively expanding the structural coverage of protein-sequence space with high accuracy models. Nucleic Acids Res 50:D439–D444 466. El-Bena J, Dang PM-C (2021) Starting-NOX2-Up: Rac unrolls p67phox. J Leukoc Biol 110:213–215 467. Noreng S, Ota N, Sun Y et al (2022) Structure of the core human NADPH oxidase NOX2. Nat Commun 13:6079 468. Liu R, Song K, Wu J-X et al (2022) Structure of human phagocyte NADPH oxidase in the resting state. elife 11:e83743 469. Parkos CA, Quinn MT, Sheets S, Jesaitis AJ (1992) The structure of the human neutrophil plasma membrane b-type cytochrome involved in superoxide production. In: Jesaitis AJ, Dratz EA (eds) Molecular basis of oxidative damage by leukocytes. CRC Press, Boca Raton, FL, pp 45–56 470. Chou PY, Fasman GD (1974) Prediction of protein conformation. Biochemistry 13:222–245 471. Wallach T, Segal AW (1997) Analysis of glycosylation sites on gp91phox, the flavocytochrome of the NADPH oxidase, by sitedirected mutagenesis and translation in vitro. Biochem J 321:583– 585 472. Kyte J, Doolittle RF (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157:105–132 473. Taylor RM, Baniulis D, Burritt JB et al (2006) Analysis of human phagocyte flavocytochrome b558 by mass spectrometry. J Biol Chem 281:37045–37056

1

Paradigm Shifts in the History of Nox2 and Its Regulators: An Appreciative Critique

474. Vignais P (2002) The superoxide-generating NADPH oxidase: structural aspects and activation mechanism. Cell Mol Life Sci 59:1428–1459 475. Heyworth PG, Cross AR, Curnutte JT (2003) Chronic granulomatous disease. Curr Opin Immunol 15:578–584 476. Nozaki M, Takeshige K, Sumimoto H, Minakami S (1990) Reconstitution of the partially purified membrane component of the

63

superoxide-generating NADPH oxidase of pig neutrophils with phospholipid. Eur J Biochem 187:335–340 477. Cross AR, Heyworth PG, Rae J, Curnutte JT (1995) A variant X-linked chronic granulomatous disease patient (X91+) with partially functionl cytochrome b. J Biol Chem 270:8194–8200

2

The Phagocyte Oxidase: The Early Years John T. Curnutte and Alfred I. Tauber

Abstract

This autobiographical dialogue describes the scientific interchanges during the 1970s of the Boston-based investigators studying the respiratory burst associated with phagocytosis. The early studies to characterize the responsible enzyme are reviewed and placed in the context of the initial studies conducted by Bernard Babior and John Curnutte that led to identification of superoxide as the product of oxygen reduction generated by phagocyte activation. The posit of NADPH as the electron donor contested other co-factors, a hypothesis that proved correct. By the end of the decade (following elucidation of artefacts, demonstrating flavin dependence, and discovering the absence of a b-cytochrome in X-linked chronic granulomatous disease patients), the basic enzymatic architecture that produced reactive oxygen species mediating the inflammatory response had been elucidated. Keywords

Respiratory burst · NADPH oxidase · NADH oxidase · Superoxide · Neutrophil

1

Prologue

The following dialogue is a reconstruction of conversations held from the summer of 2021 into the spring of 2022. John and Fred had not seen each other for 30 years and while each kept tabs on the other, their lives had diverged by 1991. J. T. Curnutte Samsara BioCapital, Palo Alto, CA, USA Pliant Therapeutics, South San Francisco, CA, USA Orchard Therapeutics, London, UK A. I. Tauber (✉) Center for Philosophy and History of Science, Boston University, Boston, MA, USA e-mail: [email protected]

Fred was moving from the laboratory to history and philosophy of science, and by 1998 he was tenured in the Department of Philosophy at Boston University, where he led the Center for Philosophy and History of Science. His scholarship focused on the theoretical development of immunology from the late nineteenth century into the current era. During this “middle period,” of their respective careers, John moved his laboratory from The Scripps Research Institute to Genentech in 1993 as well as his chronic granulomatous disease (CGD) clinic at Scripps to Stanford Medical School. This was the beginning of his career in drug development in biotechnology, beginning with Director of Immunology at Genentech and progressing to President/CEO of the DNAX Research Institute, then CEO of 3-V Biosciences, and finally Head of R & D/Executive Vice President at Portola Pharmaceuticals until semi-retirement in 2019.

2

Introduction

John: It’s so good talking with you again. It has been far too long! Fred: Likewise! I’ve been thinking of our early days when we worked on the oxidase. I’d like to reconstruct that story. Tell me what you’ve been doing. John: Yes, two perspectives are better than one’s own memory. Fred: We will have to bracket the story, roughly to when you started as a student in Bernie Babior’s laboratory and when I stopped working actively on the oxidase by the early ‘80’s. As you know, I did some early physical characterizations and then moved to activation mechanisms, especially with phorbol esters and accompanying biochemical events. I see my oxidase story as largely a second chapter to yours. I entered the field in 1976 and by then you had already made the key observations. John: Let’s reconstruct that part of the tale, roughly from when I was in college and started my oxidase odyssey with

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_2

65

66

J. T. Curnutte and A. I. Tauber

3

Fig. 2.1 Bernard Babior

Bernie (Fig. 2.1). When we met in 1976, Bernie and I had already described the production of superoxide by human neutrophils, which quickly led to the NADPH oxidase, the controversy with a contending alternate, and various other enigmas. Fred: As a historian, I want to have an accurate record of that exciting period. John: So, we begin by setting the context. I know you’ve written a lot about the origins of our field and like most scientists, I little inkling of that story. Fred: I appreciate you using “our,” because, yes, I do feel that the oxidase is at the very least, our shared experience. Bernie’s laboratory offered us the opportunity to do something new. I described hydroxyl radical production in human neutrophils, but that line of research was not pursued [1]. By the time I left his lab in 1978, you had outlined the fate of oxidase research for the next decade, and I followed those leads. John: And then 20 years later, you wrote the history of the origins of phagocyte research (see Appendix). Fred: A lot happened in between! But let’s get to the early days of your story with Bernie and Manfred Karnovsky.

Early Days

John: I first met Bernie in 1970. I recounted that part of the oxidase story in a special 80th anniversary edition of the Journal of Clinical Investigation that had featured our first article on superoxide as one of its most frequently cited articles [2, 3]. Fred: Yes, I know that article, but let’s get to the underbelly of it! John: Ok. I was a freshman at Harvard College, majoring in biochemistry and in search of an adviser. I soon learned of Bernie, a brilliant young professor of medicine and a gastroenterologist at Harvard Medical School’s Thorndike Laboratory at Boston City Hospital (BCH). Since he was also a tutor in the Biochemistry Department at the College, he was able to take me under his wing and for the next 2 years patiently taught me the complex biochemistry of vitamin B12. These included rigorous one-on-one tutorials and, eventually, some direct research at the bench. Fred: Ah yes, I know the Thorndike. I did my first clinical rotations at BCH about the same time you began your work there. Maybe we saw each other in the halls? In any case, I fulfilled an early dream of working at BCH when I left Harvard in 1982 to become the hospital’s Chief of Hematology and Oncology. That was 1982. John: I also eventually became a hematologist, but then Bernie was part of the gastroenterology group. While I was inspired by Bernie’s passion for the study of cobalamindependent enzymes and the sophisticated intricacies of his experiments, I was even more drawn to the hematology research two floors below at the Thorndike in the laboratories of William Castle (then an emeritus professor), James Jandl, H. Franklin Bunn, and Herman Godwin. In fact, Herman introduced me to Bernie. Fred: Yes, I knew them. It was a remarkable group. Castle was world-renown for describing intrinsic factor (missing in pernicious anemia) and the work done on the red cell there became the foundation of contemporary hematology. John: Right, but Bernie moved us to the white cell. And he wasn’t even a hematologist! He was a member of the gastroenterology group. Fred: Yes, and you were the one who moved him! A nice twist, like so many in this tale. John: Utterly unanticipated. In any case, I was disgruntled. Although hesitant, I finally found the courage to tell Bernie that my interests were in hematology and that I wanted to do research on blood cells instead of vitamin B12. Rather than expressing disappointment, he looked at me first with fatherly understanding and then with an intense excitement in his eyes. I saw a signal that I learned would be the beginning of a creative surge.

2

The Phagocyte Oxidase: The Early Years

Fred: Yes, he was an idea man, with a penchant for molecules and mechanisms. Descriptive biology was not of interest to him. And that is why I joined his lab. John: I didn’t know that. Fred: While you were beginning with Bernie, I was in Frank Austen’s lab at Harvard doing early work on prostaglandin effects on histamine release. I was completing medical school and intent on “curing” my mother’s asthma. I had early success with Frank and when I sought to continue work on the biochemistry of inflammation as a hematology fellow, I sought out Bernie. His leukocyte research was the closest parallel to my earlier investigations. John: So, you were a therapeutic idealist! Fred: No more than you, but my early career plans were motivated by quite a narrow interest. It was, in retrospect, quite different from yours. I always regarded you as fascinated by science itself. The way to get to the basics. I think of you as driven by the curiosity. That has proven sustaining for you. I had another agenda. John: Tell me. Fred: Too complicated. It is in a book [4]. However, at least from my current vantage, I see my science philosophically. Beyond the emotional motivation, it was to master a way of thinking, a road towards certainty. John: That’s funny. There is little certainty in science. We only find more questions and in some sense a way of measuring or seeing the scope of uncertainty! Fred. Correct, but naivete has a way of ensnaring the young. John: Of course! And I was just a college kid. What did I know, but I did have my bias. Fred: Bias? John: Yes. I always wanted to be clinically relevant. I wanted a disease problem that was significant. I just knew that B12 was not “it.” So, I pushed against Bernie, which in retrospect might be seen as what you’d call, “chutzpah.” Fred: I wonder if the post-60’s loosened the lines of authority. I have my own stories about challenging mentors, but this one is particularly interesting because you catalyzed a transition for Bernie that proved far more fruitful than anything he had done before. John: You’re right. He had graduated from medical school at University of California, San Francisco in 1959 and later received his Ph.D. in biochemistry from Harvard in 1965 under the tutelage of Konrad Bloch. His research from graduate school until we began the superoxide studies in 1972 was focused on the mechanism of action of enzymes, predominantly ethanolamine deaminase, a vitamin B12-dependent enzyme from Clostridium that catalyzes the conversion of ethanolamine to acetaldehyde and ammonia. Bernie had published a series of ten papers on this, mostly in the Journal of Biological Chemistry, between 1969 and 1972. These were the papers I read with him in my biochemistry tutorials.

67

Fred: He needed to move on. John: And Bernie had that glint in his eye! He told me that he was fascinated by the biology of phagocytosis and had been following several apparently divergent areas of investigation. He had mulled over a hypothesis about a potentially novel mechanism by which leukocytes kill microbes—one involving superoxide (O2-). For the next hour, he explained the evolution of his thinking. He was intrigued by the work of Irwin Fridovich and his colleagues at Duke University, who had shown that biologic systems could generate superoxide [5, 6]. This highly reactive oxygen radical was produced by enzymes such as xanthine oxidase in the course of their normal catalytic function or during the oxidation of hemoglobin to methemoglobin. To counterbalance the potentially harmful effects of this oxygen free radical, cells also contain an enzyme to metabolize superoxide—superoxide dismutase (SOD). Bernie was particularly struck by the observation that obligate anaerobes lacked superoxide dismutase, while those that could survive in oxygen contained this protective enzyme [7]. Fred: I know he read the journals and was aware of a wide range of advances, but I wonder what caught his eye about superoxide? I must admit that I am suspicious of personal reconstructions. When historians have examined origin stories, much is revised in the re-telling. I’ve documented such distortions and there are many others [8]. John: Well, I think this tale is accurate, because the link of the respiratory burst to bactericidal events came from several other reports. Manfred Karnovsky at Harvard, showed that while phagocytic leukocytes could ingest bacteria under anaerobic conditions, oxygen was required for their efficient destruction (Fig. 2.2). In conjunction with phagocytosis, a “respiratory burst” occurred, in which increased amounts of oxygen were consumed by neutrophils, not by mitochondrial oxidative phosphorylation, but by a cyanide-insensitive pathway leading to the production of hydrogen peroxide (H2O2) [9, 10]. Fred: And that finding was built from early studies that first detected a change in respiration during phagocytosis. As you know, the respiratory burst was discovered in 1932, when Baldridge and Gerrard reported that canine neutrophils ingesting Sarcina lutea consumed oxygen at a rate twice that of resting cells by 15 minutes, which persisted for 10–15 min, and then completely subsided by 90 minutes [11].1 Apparently, the experiments were done on the simple supposition that “as other cells increase their energy turnover during activity, it might be anticipated that the respiration of leucocytes would increase during active phagocytic ingestion—unless the entire process is controlled by physical tensions.” Their study was not placed within the immune or 1

Although a paper by A.D. Ado in 1933 is cited as an early study of leukocyte respiration, his data are equivocal [12].

68

J. T. Curnutte and A. I. Tauber

Fig. 2.2 Manfred Karnovsky

inflammatory contexts and were undertaken for unspecified reasons. Gerard was a neurophysiologist interested in the metabolism of neurons and the general effects of respiration. He was not part of the infectious disease investigators and, remarkably, the finding of a phagocyte respiratory burst had no impact on immunologists of the period.2 John: Imagine, it took 30 years to push the ball downfield. Fred: Forever the football player! Harvard took a naïve kid from a working-class family who played sports, had high test scores and grades, and turned you into a scientist. John: How did you remember that?! Fred: I remember, because you are what I call an American Success, rising by merit. John: And you are the son of immigrants. Fred: True, but my culture was a lot closer to Harvard, then yours. Anyway, get on with it.

4

Superoxide

John: Bernie was attracted to superoxide from another angle. Seymour Klebanoff in Seattle had demonstrated that myeloperoxidase (MPO), present in abundant amounts in the azurophil granules of neutrophils, catalyzed a reaction between H2O2 generated in the respiratory burst and halide ions (Cl- or I-) to produce hypochlorous acid (HOCl) or the corresponding iodide acid (HOI) [14, 15]. There was no doubt in Bernie’s mind that HOCl played an important role in phagocytic killing. Fred: The clinical scenario, however, directed the research traffic. If patients lacking MPO were highly susceptible to 2 For example, in the 1939 comprehensive fifth edition of Zinsser’s Immunity, no reference is made to this study and when considering phagocyte functions (in particular, bactericidal mechanisms) the discussion focused on leukocyte enzymes ([13], pp. 284–339).

infection, this mechanism would have been implicated as the major bactericidal pathway. John: Correct, but the finding failed the clinical scenario. Bernie was puzzled by a report from Robert Lehrer and Martin Cline at University of California, Los Angeles (UCLA) that patients with a congenital deficiency in myeloperoxidase had minimal problems clearing infections except in special circumstances such as diabetes [16]. Bernie reasoned that there must be other oxygen-dependent killing mechanisms. This is where I think his work on vitamin B12dependent enzymes played a role since he was keenly interested in the one-electron reduction steps for the three oxidation states of cobalamin—Co3+, Co2+, and its super-reduced form, Co1+—so much so that he convinced me to study the formation of super-reduced Vitamin B12. That became my initial research project for my undergraduate biochemistry thesis! Bernie hypothesized that phagocytic leukocytes might be capable of catalyzing the 1-electron reduction of oxygen to O2- in addition to (or instead of) the 2-electron reduction to H2O2 and that this oxygen radical could be part of the microbicidal system of the cell. Fred: And thus, the search for O2- and the enzyme(s) that produced it commenced. John: Fridovich noted in his review of the origins of his research that another scientist in an adjacent lab to his—an MD-PhD student named Skip Kessler—had hypothesized a link between the phagocyte respiratory burst and O2- but did not pursue his intuition [6]. When Bernie asked for SOD, Fridovich declined, citing the conflicting interest of his young colleague. Bernie then turned to a commercial source for the enzyme and that is how we pursued the characterization of the respiratory burst. Fred: Fascinating! I didn’t know that angle, but it suggests that superoxide was in the scientific ether and others clearly were thinking along the same lines as Bernie. In any case, let’s step back for a moment and look at the big picture: The story of the NADPH oxidase joins two research traditions. The biology of inflammation, more specifically, the challenge of elucidating the mechanisms of host defense, begins with Metchnikoff’s early descriptions of phagocytes engaging bacteria as part of a more general response to injury (see Appendix). That story resumed in the 1950s when the metabolic properties of neutrophils ingesting foreign material were re-examined. The research leading to the NADPH oxidase begins with the seminal report by Anthony Sbarra and Manfred Karnovsky in 1959 [17]. During the uptake of particles under aerobic conditions, they observed increased lactate production, increased oxygen uptake, and active glycolysis. The increase in oxygen consumption was not blocked by inhibitors of mitochondrial oxidative respiration such as cyanide, antimycin C, or dinitrophenol, which then posed the question of why this pronounced increase in oxygen uptake was necessary if it was not being used for

2

The Phagocyte Oxidase: The Early Years

respiration. Soon thereafter, Iyer, Islam, and Quastel showed that formate is converted to CO2 during phagocytosis and concluded that this catalytic oxidation results from the production of H2O2, a hypothesis proven shortly thereafter [18– 20]. At the time, the prominent debate between Karnovsky and Juda Hirsch Quastel concerned differentiating the cyanide sensitive/insensitive enzymatic activities and their respective sub-cellular localization [21]. John: Then Relevance appeared in full costume when Selveraj and Sbarra demonstrated that oxygen was required for bacterial killing in vitro [22], a finding soon extended by Holmes, Page, and Good, who showed that leukocytes from boys with CGD failed to show a respiratory burst and hydrogen peroxide formation upon phagocytosis [23]. Fred: The larger context is interesting, because phagocyte biochemistry grew from the basic characterization of respiration and more specifically the research of electron transfer that had been elucidated in mitochondrial phosphorylation. The link between the oxidation of glucose and ATP formation was first observed in the early 1940s and the theoretical and mechanistic basis was presented by Peter Mitchell’s chemiosmotic theory in 1961 [24]. Such a marriage between basic research to discern the etiology of clinical disorders is the Ur Tale of twentieth century biomedicine and hardly qualifies as a unique chapter in the roster of many achievements.3 John: Precisely! That is my story in a nutshell. Fred: In retrospect, the oxidase story flows as expected, but that’s because we know the arc of events. With the plot ending in place, the preceding course is plain, which naturally omits all the dead ends. What we don’t know, and cannot know, is how in fact Bernie made the connections. From our perspective, the cascade of the thought process seems evident: By the 1970s, oxygen reduction was understood in fine detail, and cases of application seem to have abounded. Because Bernie was conversant with electron movement in biological systems, he could make the key extrapolation. Moving from an intuition to a scientific fact is, of course, the heart of science. It’s the motor of our trolley. So, how was the hypothesis pursued? John: I can offer only scant details about Bernie’s thought process, but I can say that it didn’t take me long to decide that I would study the phagocyte. I must note here that his generosity in allowing me to work in this area is one of the major reasons for my affection and respect for Bernie. He enlisted in an area of research that addressed my interests and his only peripherally, at least in the beginning. I read the stack of 3

Ur was a major Sumerian city-state located in Mesopotamia, founded circa 3800 BCE and gaining prominence around 2000 BCE (https:// www.thoughtco.com/ancient-city-of-ur-mesopotamia-173108). The reference pertains to “source” or “beginning” of a history, most notably Abraham’s biblical city of origin from which he departed to eventually settle in Palestine.

69

papers he gave me and found additional references. In a few days, the initial experiments began. The first challenge was to decide on the kind of phagocytes to study. We agreed neutrophils would be the best, but Bernie, ever the biochemist, thought we should use rat or guinea pig peritoneal cells as they would be plentiful and easily obtained, while I argued for fresh circulating human neutrophils. My reasoning was based on the larger clinical setting. If successful with human cells, we could expand our work to patients with susceptibility to infections or those suffering immunodeficiencies. Bernie supported my recommendation, and after a few weeks, I was able isolate human neutrophils. Fred: During this period, you were still an undergraduate? John: Yes, it was the summer after my third year at Harvard. The first few weeks of experiments were frustrating, because I could see no evidence of superoxide production using tetranitromethane as the detector. Neither Bernie nor I knew much about cell biology, so it did not immediately occur to us that this reagent might be toxic to the cells! Fortunately, we soon realized this problem and switched to a “gentler” cytochrome c detector. The first experiment, although not optimal, revealed low levels of cytochrome c reduction after the addition of latex particles to the neutrophils. Fred: The phagocytosis initiated the respiratory burst, and similar to the original report from 1932 [11], the augmentation was modest. Reminds me of my first experiments . . . vividly recalled. The fresh enthusiasm of seeing something novel is exhilarating. And while perhaps a likely outcome, failure sits at every turn, as we both know. John: Like all research, a web of results sets the question and, if posed correctly, within the question lays the answer. McCord and Fridovich had established the foundation of our experiment by elegantly showing that SOD blocked cytochrome c reduction by superoxide [25]. By using SOD, the increased cytochrome reduction by activated neutrophils could be shown to be O2- because it was blocked by SOD [2]. Fred: Seems simple enough, but then again, you were the first who demonstrated the process. John: But beyond the initial insight or intuition, proof requires experimental rigor. Bernie taught me how to construct experiments with extensive controls. That approach proved critical as we later sorted out confusing artefacts. Every constituent in the experimental reaction was first omitted and then replaced by a boiled inactivated form. Cytochrome c spectra were carefully analyzed to check for authenticity of reduced cytochrome. Kinetics were measured, albeit roughly. Ruby Kipnes, Bernie’s gifted technician, who had already been schooled in his rigorous approach, provided a keen eye to make sure the controls met Bernie’s high standards.

70

Fred: The successful early experiments were fortuitous in several respects. John: Absolutely. In retrospect, it is remarkable that we were able to measure the increase in superoxide under the experimental conditions we used. The rate of superoxide production was approximately 0.1 nmol O2- /min per 107 cells, just two-fold greater than the “resting” rate seen in cells not undergoing phagocytosis. We later learned that serumopsonized bacteria were much more potent stimulators of superoxide production [26] and could result in rates of more than 100 nmol O2-/min per 107 cells—a rate that represents a several-hundred-fold increase over that seen in resting neutrophils [27]. Thanks to the extensive replicate controls we performed in our initial experiments with latex, though a poor activator, we were able to establish that superoxide production increased upon phagocytic stimulation. The anticipated advantage of using human neutrophils proved valuable shortly thereafter. Fred: Once you had the basic assay, the clinical promise beckoned. John: Right-on. Bernie and I attempted to discover whether superoxide production might be defective in neutrophils from patients with CGD. Robert Good’s laboratory had discovered that the neutrophils from these children did not undergo a respiratory burst when stimulated [23], nor did they generate hydrogen peroxide as described by Pincus and Klebanoff [28]. We reasoned that if superoxide was truly a product of the phagocytic respiratory burst, it should not be detectable in the stimulated CGD neutrophils. In collaboration with Dana Whitten, a pediatric hematologist from the New England Medical Center, we studied two youngsters with CGD and found that neither produced detectable levels of superoxide [29]. Thus, our hypothesis about the relation of superoxide to the respiratory burst—and its possible role in host defense—was strengthened by this important study. Fred: The implications of these findings were noteworthy. First, the generation of a superoxide by neutrophils opened the possibility that other reactive oxygen species (ROS) could be generated by these cells. This has been studied extensively in the subsequent 50 years and deepened our understanding of both the mechanisms of microbial killing and the harmful effects of acute and chronic inflammation [30]. Second, the cytochrome c assay—and the specificity for superoxide afforded by the use of SOD—provided a rapid and highly sensitive method for measuring the respiratory burst that has largely replaced the cumbersome and low-throughput assays for oxygen consumption and H2O2 production. Third, the studies highlighted the profound impact of studying cells with human genetic defects, in this case CGD, on elucidating complex biochemical and cellular pathways. And finally, the ongoing debate at the time regarding the enzymatic basis of the respiratory burst could be addressed, and perhaps resolved, by determining which of

J. T. Curnutte and A. I. Tauber

the candidate enzymes generated superoxide in quantities sufficient to account for the large magnitude of reactive oxygen species measured in intact cells.

5

NADH Oxidase vs. NADPH Oxidase

John: Not surprisingly, I was excited. I based my undergraduate biochemistry honors thesis on these findings. One of the two reviewers, Manfred Karnovsky, offered a mixed critique: on the one hand, he acknowledged that the research had been well conducted, but on the other hand, he was still skeptical of the findings. Herein lies the next chapter of my story. I had been accepted into Harvard Medical School, where he resided as a senior professor in biological chemistry. Manfred had done extensive work on the respiratory burst and had characterized an enzyme that he believed was responsible for that activity, which he posited was an NADH oxidase. It was about the same time in late 1973 that Bernie began to lean increasingly toward NADPH as the electron donor. The character of the respiratory burst oxidase would be dependent on resolving this issue. That question became the focus of contention, and I was eventually caught in the middle of their controversy. Fred: Yes, I too became a bit embroiled in their debate, but I had much less at stake than you did. While their argument was about identifying the “right” enzyme, it was also a personal competition: Bernie fell in the generation between you and Manfred; Bernie had moved from Harvard to Tufts, an up-start by Manfred’s patrician airs; and perhaps most importantly, much of their professional stature was at stake. Whose enzyme would triumph? The Km’s were not just about reaction rates. John: I naively went forth. Indeed, I thought having two expert mentors would only serve me well. After the first day of medical school classes in the fall of 1973, I walked into Manfred’s office and introduced myself. Recall, he had reservations about my senior thesis, but I did not fully appreciate his own interests, nor the stakes involved. In any case, we had a cordial and productive meeting at the end of which we agreed that I could work in his lab in parallel with my ongoing research in Bernie’s lab (while attending classes!). Manfred was particularly interested in the enzymatic basis of the respiratory burst and was committed to identifying which of the candidate oxidases could generate superoxide. At that time as I mentioned earlier, he was the lead proponent of the NADH oxidase, a soluble flavoprotein with a MW of ~310,000 that had been identified in his lab 10 years earlier by a graduate student, Bob Cagan, who used elicited guinea pig neutrophils [21]. Later, Manfred and Robert Baehner reported deficiency of the enzyme in CGD patients [31]. The other candidate to account for the respiratory burst was NADPH oxidase, a particulate, Mn++- dependent

2

The Phagocyte Oxidase: The Early Years

enzyme also studied primarily in guinea pig neutrophils, that had been first described by Iyer and Quastel in 1963 and Rossi and Zatti in 1964 [32, 33]. Fred: Thus, well before Bernie appeared on the scene, Manfred was promoting the NADH oxidase while Fillippo Rossi championed an NADPH oxidase [10]. So, with both candidate oxidases having been described a decade before, how did that figure in your thinking? John: It was not clear at that time which of these two pyridine nucleotide oxidases (or perhaps a yet to be discovered third enzyme) was the true respiratory burst oxidase. I viewed it as an advantage that much work had already been done since I wanted to focus initially on which enzymes were capable of generating O2-. I also realized that others in the field would likely follow a similar strategy, but that was fine. Manfred and I agreed that I would systematically study guinea pig NADH and NADPH oxidases, determine whether they generated superoxide, and then compare these findings with work that Bernie and I were doing in his laboratory studying the human respiratory burst oxidase in lysates of human neutrophils. The early experiments in Bernie’s lab performed by Ruby Kipnes using SOD-inhibitable nitro blue tetrazolium (NBT) reduction as a detector for superoxide production pointed toward NADPH as its Km was two orders of magnitude less than that for NADH [34]. The interpretation of these studies was confounded, however, by the finding that the oxidase activity measured with the NBT detector was normal in two CGD patients. Meanwhile, I prepared guinea pig NADH oxidase according to the method used in Manfred’s lab and confirmed its specific activity was comparable to that seen in his lab previously [21]. Instead of using NBT as a detector for superoxide, I utilized cytochrome c for two reasons. First, I had worked with a half dozen different tetrazolium salts in high school as part of my Westinghouse Science Talent Search project to assay for bacteria in urine. I had seen high background signals caused by nonspecific reduction and thus had concerns about an NBT assay. Second, by partially purifying the candidate oxidases, I hoped to diminish the background signal of nonspecific reduction of the cytochrome c. I was excited to find—as was Manfred—that the guinea pig NADH oxidase generated substantial superoxide, albeit at a rate that was only about 20% of the rate of NADH oxidation ([35], pp. 134–139). We didn’t know how to account for the other 80%. Perhaps there was also divalent reduction of oxygen to H2O2 in addition to the univalent reduction to O2-. Pre-activation of the neutrophils with opsonized bacteria prior to disruption did not change the rate of NADH oxidation or superoxide production. I then turned my attention to the human NADH oxidase using the method Manfred had developed with Baehner [31, 36]. The

71

specific activity of NADH oxidation consistently agreed with the results reported earlier but none of the multiple preparations, except one, showed even a trace of superoxide. It thus appeared that that there was an important species difference between the guinea pig and human NADH oxidases [35]. Fred: I got roped into Manfred’s NADH story by helping John Badwey, who was then a post-doc in Manfred’s lab, characterize the human neutrophil NADH-cytochrome b5 reductase, which we later purified [37, 38]. Why, I still don’t know! Regarding the phagocyte respiratory burst, it led nowhere. Perhaps I had been seduced by the low-hanging fruit syndrome: easy to grab and pocket, hoping something interesting would emerge. But unlike the NADPH oxidase, the reductase, although plentiful in neutrophils, had no known function beyond some general biochemical activities like desaturation and elongation of fatty acids, cholesterol biosynthesis, drug metabolism, and, in erythrocytes, methemoglobin reduction. John: I had plenty of dead-ends, but let’s not dwell on those byways. Fred: Right. The meetings you had with Bernie and Manfred to review the data must have been quite interesting. Of course, each of you sought a resolution of sorts that would preserve Manfred’s and Bernie’s respective enzymatic champions. Where did you sit on that seesaw? John: We had frequent meetings that were spirited but without animosity. As an example of that collegial spirit, Manfred wrote an article for the New England Journal of Medicine several years later that John Badwey and I co-authored with him entitled “The enzyme of granulocytes that produces superoxide and peroxide—an elusive pimpernel.” [39]. In that article we wrote, “For about the past 50 years, a great deal of research has focused on attempts to elucidate the enzymologic basis of the respiratory phenomenon. The quest for the truth in this field has become a rather contorted history over the years, characterized by some controversy, but, perhaps fortunately, little acrimony.” As the graduate student caught between Bernie and Manfred, I appreciated his graciousness. So, after seeing that human NADH oxidase did not generate O2-, I confidently moved over to characterizing the NADPH oxidase. The data compelled that decision. Fred. Yes, but specific comparable data were required, and that took longer to generate than expected. When I enrolled in the discussion, Manfred, Badwey, and I published comparisons in an obscure place several years later [40]. By then, I think the jury had already selected Bernie’s baby. John: The earlier data favored NADPH despite Manfred’s reluctance to concede. Fred: Different species have different biochemistries. Humans are not guinea pigs.

72

John: Yes, but egos sometimes get in the way of judgments. In any case, when we studied superoxide production by the guinea pig neutrophil NADPH oxidase prepared by the method used in Rossi’s laboratory, the particulate fraction thus prepared had a specific activity for NADPH oxidation in close agreement with that reported earlier [41]. There was a 36% increase in activity if the particulate fraction was prepared from cells stimulated prior to homogenization with opsonized bacteria. Despite the rapid rate of NADPH oxidation repeatedly observed, multiple attempts to detect superoxide were unsuccessful [35]. Almost by a conditioned reflex, I included controls that followed Bernie’s tutelage ingrained from our first superoxide experiments several years earlier. One of those systematic controls was to add SOD to the reaction mixture to measure its effect on NADPH oxidation. The astonishing result was that SOD inhibited the rate of NADPH oxidation by 90%. Similar results were seen if consumption of the other substate, oxygen, was measured instead. Honestly, the result caught me—and Bernie and Manfred—by surprise. We all met soon thereafter and hypothesized that Mn++ might be the culprit by interacting with O2- as a first step in a free radical chain reaction. We were correct and reported our findings and the preliminary mechanism in an abstract in early 1975 [42]. Rossi’s team had uncovered the same SOD effect in the Mn++-dependent system and reported their findings around the same time [43]. We followed our preliminary report with an extensive series of studies that established that the requirement for Mn+ + was artifactual and resulted from a non-enzymatic free radical chain reaction in which NADPH was oxidized to NADP with O2- serving as both the chain initiator and one of the propagating species (along with Mn++); in this reaction, the neutrophil particles only served as a source of O2and could be replaced by another defined O2- generator such as xanthine oxidase [44]. Fred: That was a critical insight and then another soon followed when you and Bernie studied the particulate human NADPH oxidase—but without manganese—in CGD patients and healthy controls using the cytochrome c assay for superoxide. I think that was the inflection point of the research. John: Agreed. We optimized the system by activating neutrophils with opsonized yeast particles (zymosan) instead of latex or bacteria prior to disruption and then fractionated the cells using the method developed by Bob Lehrer and David Hohn in their human NADPH studies that had included Mn++ [45]. Our results showed that the particulate fraction from three CGD patients failed to produce superoxide [46]. It is important to point out that Hohn and Lehrer were able to see a deficiency of NADPH oxidase activity (as measured by NADPH oxidation and by oxygen consumption) despite the large amplification artefact induced by Mn++ which they still included in their experiments. They were

J. T. Curnutte and A. I. Tauber

aided by the use of a potent respiratory burst stimulus, opsonized zymosan, that caused substantial activation of the neutrophils prior to disruption. Fred: Remember those experiments we did with zymosan and latex? The latex-stimulated cells produced miniscule amounts of superoxide with roughly the same oxygen consumption as zypmosan [47]. I hesitate to call this finding an anomaly, because it might represent an alternate pathway. Perhaps we are witnessing two different electron pathways, one going directly to superoxide and the other to hydrogen peroxide. The latex-induced respiratory burst may have no physiological significance and, perhaps, represents a redundant, secondary pathway? John: I don’t know, but those late-night experiments were fun! And recall how difficult it was to publish that paper? Fred: It was only published in Pediatric Research because Joe Bellanti, the editor, was a friend of mine and he pushed it through the review jungle. How many journals rejected our paper? John: I don’t remember, but more than a couple! Fred: How many findings succumb to the dominance of guiding dogmas! “Anomaly” means “irregular,” “unexpected,” or “unorthodox.” In any case, no one has pursued the basis of the divergent activation profiles. We didn’t. . .or perhaps couldn’t. Oh well. Let’s get back to our story. John: I entered the MD-PhD program at Harvard in 1975 with Bernie and Manfred as co-advisors and focused on further studies on NADPH oxidase [48], its subcellular location [49], its mechanism of activation using fluoride activation as a model system [50], and the role of phosphorylation in oxidase activation. Fred: That was when we met. I was a research fellow in Bernie’s lab and went back to Harvard 2 years later, where I embarked on my own research that built upon Bernie’s early work on the oxidase solubilization. John: Bernie was not pleased that you left after only 1 year. Fred: Correct, but I had a faculty appointment across town that did not exist at Tufts. John: Bernie began his attempt to isolate the oxidase about the time you exited. Fred: Recall, solubilization was a difficult hurdle to solve, but once a cocktail was devised, the basic biochemical characteristics could be measured. While Bernie used Triton-X 100, I employed deoxycholate to solubilize the enzyme [51, 52]. My preparation captured 10–20% of the particulate preparation, which I stabilized in glycerol. The Km was lower for NADPH than NADH, and the sizing procedure yielded two peaks of Mr 150,000 and another >300,000. Later, when the various components of the oxidase were discerned, it seems that I had the intact enzyme and then something else. John: There was a lot of interest in the b-cytochrome.

2

The Phagocyte Oxidase: The Early Years

Fred: I wonder at times why we didn’t pursue the cytochrome more aggressively. Tony Segal had described the bcytochrome558 and that radically changed the research trajectory [53]. I recall Bernie was really flummoxed by Tony’s original paper and then by showing the missing cytochrome in some CGD patients but not others, he opened the door to the multiplicities of the defect [54]. Basically, Bernie was upset that he hadn’t looked for such a common electron transporter. Like so many discoveries, Tony’s was seemingly so simple. John: Tony’s work on cytochrome b was, indeed, a major advance for both the oxidase biochemistry and the elucidation of CGD. What came into focus is how the flavin and cytochrome b functioned as an electron transport chain. Two of the young scientists who worked with Tony and Owen Jones on cytochrome b and phosphorylation—Paul Heyworth and Sandy Cross—later joined my lab when I went to Scripps in 1986. Before that, I finished my clinical training at the Massachusetts General Hospital and then Boston Children’s Hospital/Dana Farber Cancer Institute in pediatric hematology-oncology. In 1983 I joined the faculty at the University of Michigan in Larry Boxer’s division. But in none of those laboratories did I pursue the physicochemical characteristics of the oxidase. Fred: Bernie’s solubilized activity required the addition of FAD to reconstitute superoxide production [55] and he went on with Ted Gabig to further characterize the oxidase [52, 56]. In a brief collaboration with Chris Walsh at MIT, I used the Triton-X solubilized enzyme complex to further examine the flavin requirement and demonstrated that it accepted one electron from NADPH. We also measured the ready oxidation of the cytochrome, with a high k and an oxidation-reduction potential of -235 mV [57]. But at that time, it remained unclear whether the cytochrome was directly involved in the electron transfer scheme. John: Bernie was reluctant for a long time to accept the b-cytochrome as definitively part of the oxygen reduction pathway. Perhaps he was influenced by the old data that showed the respiratory burst was insensitive to cyanide and other heme inhibitors. Fred: Indeed, it was still unsettled. In any case, my paper was full of biophysical data, but it turned out to be my last work on the oxidase itself. I loved the biochemistry, but I drifted from enzymology into cell biology, having become intrigued with activation mechanisms. John: Why did you shift? Fred: I don’t much like “why” questions. Motivations are too complex and too often hidden. All I can say now is that for both scientific and for professional reasons (more honestly, social and political concerns), I shifted my focus of interest.

73

John: I get that! Next time, over a whiskey or two, we can discuss the undercurrents of a life in science. After all, the papers are the tip of the iceberg. Fred: Agreed. John: We both drifted into studies of the respiratory burst activation. You went one way, I another. Fred: Yes. My activation studies began innocently enough. Phorbol myristate acetate (PMA) was widely used as a neutrophil activator, and Peter Blumberg, an expert on phorbol esters, had a lab on the floor below mine at Harvard. We published on the binding characteristics of the PMA receptor and then I showed that its target, protein kinase C (PK-C), could directly activate unstimulated particulate preparations from human neutrophils [58–60]. I thought this short-cut would facilitate tracking the activation cascade, and we found that that arachidonic acid and SDS could bypass the PKC pathway, which made the stimulation pathway a rather complex process [61]. We then studied some of the associated activities in the activation cascade, but the phosphorylation story, the key to activation, was more complicated than I initially thought [62–65]. John: Phosphorylation joined our interests and became another site of our interchange. My early work on phosphorylation in 1976–77 with Manfred and Bernie hit nothing but dead ends. Later, at Scripps and collaborating again with Bernie and Naoki Okamura in his lab, we used the different genetic forms of CGD to help elucidate the multisite phosphorylation of a 48 kDa protein in PMA-activated neutrophils (later determined to be the p47-phox subunit of NADPH oxidase that is deficient in one type of autosomal recessive CGD) [66, 67]. Fred: You had a very specific clinical orientation. John: Absolutely! I wanted to determine if the CGD oxidase defect was one of activation and/or defective enzyme or both. The latter seemed likely because at least two modes of inheritance characterized the disease. I dimly saw a therapeutic horizon in which new pharmacological approaches would be devised to increase or decrease oxidase activity to treat infection or control inflammation. This was at a time of my growing interest in congenital immunodeficiencies that arose from encountering my first CGD patients as a medical student in 1973. By 1979, I was an intern directly involved in the care and treatment of CGD patients. Fred: That was also the time when many labs became interested in the subcellular location of oxidase, which we assumed was membrane bound [68] and which you, Bernie, and Marco Baggiolini had studied a few years earlier [49]. That directed my later work in isolation techniques [69]. You took another course. John: The research about the oxidase fanned out in several directions by the late 1970s. Indeed, the research agenda divided into three segments: First, the purification of the oxidase components and its structure was of primary interest.

74

The cytochrome and flavin parts were evidence of a multiplex of some kind. You, Bernie, Ted Gabig and others worked on the early solubilization of oxidase in its active form and the components were characterized soon thereafter. Second, the mechanisms of activation required the cell-free system, one in which the oxidase could be activated in a disrupted, unstimulated neutrophils and distinguished from the already established method for studying NADPH oxidase from preactivated neutrophils. The various experiments John Badwey and I did between 1981 and 1984 with arachidonic acid and other fatty acids (e.g., oleic acid, linoleic acid) to activate the neutrophils helped set the stage for the development of a cellfree system for activating the dormant NADPH oxidase that in turn helped identify and purify the oxidase components [70–73]. And third, establishing the molecular and genetic basis of CGD became an essential tool in the further elucidation of the oxidase. Two genetic forms were known at the time, but was it activation or structural defects that were the basis of CGD [74, 75]? Fred: Remind me of the approach you took to determine whether the defect was in activation or in the structural components of the oxidase (or both)? John: Beginning in 1975, one of my primary goals was to develop a cell-free activation system. I had seen in one of our first superoxide papers in which we looked at various metabolic inhibitors that the enolase inhibitor, fluoride, caused a potent activation of O2- production in neutrophils. With Bernie’s blessing, I decided to study it in great detail [50]. I was intrigued that this activation occurred with a short lag time and hypothesized it might be shortcutting a portion of the activation pathway, which at that time was a black box. When I added fluoride to unstimulated neutrophil homogenates, however, I failed to achieve superoxide production. When Badwey and I began studying arachidonic acid in neutrophils as a possible membrane-perturbing agent, we were struck by the large magnitude of O2- produced and the nearly instantaneous activation of the burst. As with fluoride, I was unable on several attempts to get arachidonate to activate the oxidase in unstimulated neutrophil sonicates. I made what I thought would be my last attempt in the spring of 1983, this time with Manfred’s neutrophils that I obtained after we had all had a lab lunch at a restaurant—with wine. Incredibly, activation was achieved! I tried several more times in the following week or so to repeat this experiment, but each time activation failed. Again, in what would have been my last attempt a few weeks later, I tried to repeat the earlier successful experiment by having Manfred drink a glass of white wine before I drew his blood. But again, failure. Fred: Why didn’t you give up at that point? John: The rate of superoxide production was so striking and reproducible the one time it worked I had to believe it was real. That summer I moved to the University of Michigan

J. T. Curnutte and A. I. Tauber

and as soon as I had my lab set up in early 1984, I resumed the experiments. This time, I tried different dilutions of the sonicate and saw that lower concentrations unmasked the activity. With reproducible and potent activation achieved with arachidonate, I characterized the system and observed that the activity was neither in the 50,000 g supernatant nor in the pellet. If both were recombined, full activity was observed. I then arranged for some of my CGD patients to visit my clinic. I studied three males, two of whom had clear X-linked inheritance. Their sonicates did not generate superoxide and on cross-mixing experiments, the defect resided in the particulate fraction. Amazingly, at least to me, their cytosolic supernatant fractions were normal when mixed with control particulate fractions. I reported these results at the American Society of Hematology meeting in December of 1984 and then followed with the rapid publication the following May [73]. As is often the case in science, other investigators were on tracks leading to similar findings in cell-free systems around the same time. Edgar Pick and Yael Bromberg achieved cell-free activation with arachidonate in macrophages [76]; Linda McPhail and Ralph Snyderman reported activation of human neutrophils that required a cytosolic component [77, 78]; and Heyneman and Vercauteren saw activation of resting horse neutrophils with linoleic or oleic acid [79]. Fred: So how did you further leverage the cell-free system to understand CGD? John: Between the growing number of CGD patients I followed at Michigan and Scripps as well as those cared for by various colleagues, my lab extended our studies to other types of CGD. I had identified two autosomal recessive/ cytochrome b-negative patients and found that they, like the X-linked cytochrome b-negative patients, had defects in the particulate fraction (that contained cytochrome b) but not the cytosol fraction [80]. In a series of seven patients with autosomal recessive/cytochrome b-positive CGD, we found the exact opposite—the defect resided in the cytosol fraction while the particulate fraction was normal [81]. Fred: With the basic schema established—multicomponent enzyme electron transfer effected by the linkage of particulate and cytosolic proteins with phosphorylation and consequent assembly comprising the activation of the NADPH oxidase—the enzyme responsible for the respiratory burst was basically characterized. Of course, much followed and other related oxidases were found and deciphered. However, I think we can close this chapter of early phagocyte oxidase research. My story, and yours as well, assumed a diversification that hardly seems feasible in today’s research climate. Having moved from the oxidase to activation mechanisms and related activities associated with the respiratory burst, I came back to the clinical setting. With Kevan Hartshorn, we developed a model system for studying influenza A-human neutrophil interactions [82]. That work has

2

The Phagocyte Oxidase: The Early Years

75

proven very useful in the COVID crisis [83]. By 1995, I closed my active laboratory research and became a philosopher of science. And you went on to very ambitious drug development. John: In 1993 I moved from Scripps to Genentech to head the Immunology Department. I established my lab there and continued the work on characterizing the structure of the oxidase, its activation, the molecular genetics of CGD, and the development of potential therapeutic inhibitors of the oxidase leveraging the drug development expertise within Genentech. My focus was increasingly directed, though, toward other major projects initiated in the Department and at the company—the development of Rituxan and Xolair, the mechanism of action of Herceptin, and establishment of platform for the company to develop humanized monoclonal antibodies. In 2000, I moved to DNAX Research Institute in Palo Alto to be President and CEO.

6

Coda and Dedication

Fred: We have been very fortunate to have participated in what appears now as a Golden Age of biomedical research (Fig. 2.3). Not that the glory days are over! Far from it, but at least within the borders of our research community, I see other massive social forces at work outside the laboratory. John: Well, I certainly saw a vastly expanded panorama in the biomedical industry. Fred: Yes, and as your own career testifies, the oxidase story reflects an extraordinary political and economic commitment (primarily) by American and European governments to sponsor basic research. This is not the place to detail the extra-curricular forces that underly the oxidase elucidation but suffice it to mention that the dramatic progression in our understanding reflects far more than the ingenuity of a few dedicated laboratories. Any attempt to chronicle this history must at least acknowledge the parallel technological advances developed for a broader array of research projects. From spectroscopy to molecular engineering, the investigative approaches that were adopted for elucidation of the enzyme and its activation were indebted to a vast technical infrastructure [84]. And beyond chronicling these methodological advances, the investments universities made to provide the institutional support for laboratories to perform their activities can hardly be measured by “overhead costs.” After all, the educational system generated the personnel to conduct the bench research and as we appreciate, sponsored our own training and later independent careers. John: And certain places become “generators.” Boston offered us a superb environment in which to find mentors, collaborators, and rich resources to support the research. Fred: That larger research incubator originated in the period immediately following World War II, when the

Fig. 2.3 John Curnutte (left) and Fred Tauber (right), Crete, 1990

medical-industrial conglomerate began its inexorable rise in Western countries. The educational commitment provided the talent and expertise to pursue the research agenda, which in turn was largely driven by the availability of funding that closed the circle of political commitments driven by the Cold War. We vaguely understand that the process of allocations is governed by a medley of granting agencies vying for budgets and our oxidase story took place during a period of bounty.4 Typically, researchers are opportunistic and follow Willy Sutton’s dictum of why he robbed banks: 4

Using the appropriations budget of the National Institute of Allergy and Infectious Diseases as a very rough indicator of funding for immunerelated research (excluding a vast universe of other immunological disorders covered by other sources), in 1954 (during the great push against polio), $5.7 million was appropriated, but by 1964, a ten-fold increase was recorded and then the Golden Age of immunological research commenced. By 1974, at the dawn of the oxidase, NIAID’s budget was $111 million. Over the next 30 years it more than tripled each decade (1984, $320 million, 1994, $1 billion; 2004, $4.3 billion). Since 2000, the budget for NIAID has grown more modestly (ca. 25%), while other medical sectors have grown disproportionately, but with the Covid pandemic this trend has been shattered. https://www.nih.gov/ about-nih/what-we-do/nih-almanac/appropriations-section-1

76

“That’s where the money is.” The clinical implications of the phagocyte respiratory burst were compelling, for that drama was framed in the on-going posturing of interest groups ranging from patients to mega-corporations representing Big Pharma and other health related industries. Controlling inflammation and enhancing host defense command undeniable prominence in the health sector. The money then flowed, and the neutrophil oxidase was characterized, which in turn quickly led to other related oxidases in various tissues. Join that story to the aging research implicating reactive oxygen species and we see an explosion of discoveries that originated in the early work on the phagocyte NADPH oxidase. Scientists live within an interlocking political web that hardly can be measured by even an exact accounting of financial statements and budgets. The point is simple: The so-called “internal” account of scientific progress as told by a review of laboratory data and interpretation is only part of a far more complex tale [85]. Examination of a few case studies has only scratched the surface of science’s sociology, but a lesson has been learned: The independent investigator is a crucial, yet singular player in a vast economic and political culture, whose dimensions are hardly realized. On reflection, this adage seems self-evident, so what is the relevance of such reflection here? John: So, this essay, like the others collected here, provides a necessary internal account, but it misses the other important elements that contribute to our tale. The reminisce given here is only a segment of the oxidase story, one told by two explorers who were largely oblivious to the larger territory in which we trekked. Fred: Correct, but another confounding matter concerns the memories of our intentions and how the early influences on what we pursued and why are not really discerned. Moreover, our dual recollection is based on an insular view of the scientist-at-work. After all, we were trained to think that empirical data adjudicates hypotheses and while interpretation may be disputed, the investigative process depends on the on-going gathering of factual material. While interpretation may be prejudiced by “extra-curricular” influences (e.g., Eddington’s promotion of Einstein’s relativity theory; tobacco-sponsored research disputing toxic effects of smoking; Pasteur’s tribunal that rejected spontaneous generation), bias is typically expunged by open exchange and debate [86]. From this vantage, pragmatic concerns overwhelmingly dominate discourse with careful boundaries putatively protecting neutrality and objectivity to govern practice. That understanding has been rigorously disputed by historians, sociologists and philosophers of science, a story summarized elsewhere.5 But for us, the practicing scientist, we require For instance, the “immune self” has been shown to be a metaphor borrowed from common psychological and philosophical conceptions of personal identity, whose use has had a deep influence on the development of immune theory and practice [87]. For a general overview of

5

J. T. Curnutte and A. I. Tauber

only the generation of evidence, the data that serves as the medium of discussion. And to be fair, within the confines of the maze in which scientists conduct their business (experiments, papers, lectures), extra-curricular elements influencing their work rarely appear. And more to the point, just as lawyers do not require training in formal psychology to influence a jury, scientists do not need to examine the hidden currents underlying their work. But that is hardly to deny that the matrix in which they function frames and often directs the course of development. John: Recognizing the larger context of our experimental story brings us to the closing acknowledgement: Before we congratulate ourselves too rigorously for particular contributions, the NADPH oxidase is a story of we, not I, and the we is well beyond the individualized accounts collected here. Fred: We might pause and ask, how might one acknowledge the role of competition and cooperation, inspiration and controversy, that have played in the oxidase story? In other words, if the net is cast widely, how might the community of investigators engaged in this arena of research be characterized? And, more particularly, how have each of us fit into that society? What kind of science citizens did we encounter and how did we find our respective places in the commune? Who were the leaders of the “thought collective” and how did they arrive in their position?6 John: We’re not going there! Enough to remember the supporting cast of many forgotten co-authors and technicians who comprised a crucial cadre of crucial colleagues who enabled our work. While we make no attempt to address that multitude, we can at least declare our heartfelt appreciation to Ruby Kipnes, Patricia Scott, Laura Mayo, Julie Rae, Rich Erickson, and Nellie Pavlotsky who “ran” our laboratories. We salute, in particular, these assistants who worked most closely with us during our early years in the laboratory. This paper is dedicated to them as a token of gratitude for facilitating our pursuit of the wonderous questions addressed here.

the constructivist understanding, see [88]; for a comprehensive overview of the field of science studies see [89]. 6 “Thought collective” (or “school of thought,” “style of thought,” “community of thought”) refers to the general way in which a scientific problem is defined and approached. The terminology was coined by Ludwig Fleck in an early sociological study of scientific practice [90].

2

The Phagocyte Oxidase: The Early Years

Appendix: The Phagocyte in Immunology History The story of the NADPH oxidase originates at the end of the nineteenth century. A new formulation of the relationship between host and contagious disease was formally stated in 1883 by Ilya Metchnikoff’s convergence of three disparate and thus far unrelated streams: (1) bacteria as etiologic agents of infection, (2) the nature and role of inflammation, and (3) the place of evolutionary principles as applied to physiology.7 The germ theory of disease was established by Louis Pasteur and Robert Koch by the mid-1870s, but there was no theory akin to our modem notion of immunological defense. Pasteur as late as 1880, while developing vaccines, believed that immunity was conferred by exhaustion of essential nutrients, analogous to the test tube model systems of bacterial growth. Koch was not even interested in the host response, confining himself to the establishment of bacterial etiology. Inflammation was generally viewed as a deleterious process, whose various components were regarded as reactive, not defensive. The white cells, already identified as amoeboid phagocytes, with purposeful movement and containing bacteria, were dismissed as transport vehicles for the pathogens, with no protective function hypothesized. In short, how bacteria might cause disease, and more fundamentally, the relation of host and pathogen from a physiological (organism) or evolutionary (species) perspective was left mute. At this early stage of immunology, Metchnikoff, an embryologist, proposed that mesodermic phagocytes, which in primitive organisms served a nutritive function, in higher animals with a digestive cavity, assumed new functions, devoid of their original digestive purpose. He extended the metaphor of “eat or be eaten” to a dedicated function of these cells, which now wandering beneath epithelial surfaces and various interstices, recognized foreign elements and devoured them. Originally, he viewed the process as a general physiological mechanism, which he called “physiological inflammation,” for the phagocytes in protecting the host, recognized the ‘other’ in every form—from senescent, malignant, damaged, or otherwise diseased cells, to foreign invaders. The latter became his focus only as he was drawn into vociferous debate with “pathologists” (microbiologists) and early immunochemists, who were by then fully engaged in establishing the physics and chemistry of life processes 7 For a review of how inflammation was understood before Metchnikoff, see ([91], pp. 108–20). Metchnikoff was an effective popularizer of his theory and in a celebrated series of lectures, he dramatically reported early observations that included the famous experiment in which thorns placed into the transparent bodies of star fish larvae resulted in phagocytes congregating around the intrusive body and then digesting it. He saw a similar process in the tadpole, where phagocytes literally ate the tail to transform the juvenile into the adult body form [92].

77

[91, 93, 94]. The issue focused on what they saw as Metchnikoff’s portrayal of the phagocyte as an independent agent, possessing primitive independent volition and thereby he was condemned as invoking vitalism to explain immunity.8 And here the oxidase story begins. In their search for chemical mechanisms, the immunochemists of Metchnikoff’s day were preoccupied with exorcising mysterious, unaccounted forces that would compromise their aspirations for establishing a physics of life. Metchnikoff became a focal point of dispute because he described the phagocyte as exhibiting autonomous behavior. The cells seemingly ‘knew’ where to go (chemotaxis) and once at the site of damage or invasion they undertook the ‘responsibility’ of protecting the host organism by eating everything in their target range. The chemists would have none of it and in criticizing the absence of defined mechanisms they sought a pre-arranged chemical basis for host defense. They soon identified antibody and complement as chemical anti-bacterial substances and by 1908, when Metchnikoff shared the Nobel Prize with Paul Ehrlich (the leading immune-chemical theoretician), the course of twentieth century immunology was set on the chemical mechanisms of host defense. In fact, Metchnikoff was forced to follow the chemists’ lead. In his magisterial account, Immunity in Infectious Diseases (1905), he cited the first studies of the biochemical basis of bacterial killing by phagocytes. He noted that following active ingestion, a drop in pH within the digestive vacuoles correlates with bacterial destruction that he thought were enacted by intracellular enzymatic “cytases” ([96], pp. 175–206). And by the time he died a decade later, characterization of “endolysins”—lumped together as unspecified enzymes and bacteriolysins of uncertain origin (i.e., endogenous serum or phagocyte-derived)—were subordinated to the characterizations of soluble serum factors ([97], pp. 296–310). And there matters stood as the focus on acquired immunity—the specificity of the antibody reaction—dominated the first decades of the twentieth century and effectively displaced interest in so-called “natural” immune mechanisms [98, 99]. However, during the closing decades of the twentieth century, following the appreciation of antigen presentation, cytokine regulation, and the various complementary roles of phagocyte-based immunity, the so-called “innate” immune 8

Metchnikoff arrived at his early schema of phagocyte function (chemotaxis, phagocytosis, killing) coupled to projections of human agency, where independent volition was assigned to cells that seemed to possess faculties for their own decision-making analogous to humans engaged in combat. This was the origin of his indictment as a vitalist. Vitalism asserted that life processes possess properties irreducible to physicochemical analysis. Accordingly, “living organisms are fundamentally different from non-living entities because they contain some non-physical element or are governed by different principles than are inanimate things” [95].

78

system again received prominent attention. The history of this inflection has not been critically appraised by historians or philosophers of science despite the importance of this chapter in immunology’s development. In this regard, the phagocyte’s respiratory burst would be found at the crossroads of biochemistry, genetics, cell biology, and immunology to illustrate the mosaic of disciplines required to decipher the mechanisms of host defense and the biochemistry of inflammation [9, 100]. Thus, Metchnikoff’s so-called “vitalism” required elucidation of a dynamic complex system, whose organization and regulation, while still only partially understood, testifies to the prescient insight of his early vision that awaited confirmation by the methodological turn to systems biology and the modern integration of diverse life sciences.

References 1. Tauber AI, Babior BM (1977) Evidence for hydroxyl radical production by human neutrophils. J Clin Invest 60:374–379 2. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. J Clin Invest 52:741–744 3. Curnutte JT (2004) Superoxide production by phagocytic leukocytes: the scientific legacy of Bernard Babior. J Clin Invest 114(8):1054–1057. https://doi.org/10.1172/JCI23377 4. Tauber AI (2022) The triumph of uncertainty. Science and self in the postmodern age. Central European University Press, Budapest 5. Fridovich I (1975) Superoxide dismutases. Annu Rev Biochem 44: 147–159 6. Fridovich I (1998) The trail to superoxide dismutase. Protein Sci 7: 2688–2690 7. McCord JM, Keele BB Jr, Fridovich I (1971) An enzyme-based theory of obligate anaerobiosis: the physiological function of superoxide dismutase. Proc Natl Acad Sci U S A 68:1024–1027 8. Podolsky SH, Tauber AI (1997) The generation of diversity: clonal selection theory and the rise of molecular immunology. Harvard University Press, Cambridge 9. Klebanoff SJ, Clark RJ (1978) The neutrophil: function and clinical disorders. Elsevier, Amsterdam 10. Berton G, Dusi S, Bellavite P (1988) The respiratory burst of phagocytes. In: Sbarra AJ, Strauss RR (eds) The respiratory burst and its physiological significance. Plenum, New York, pp 33–52 11. Baldridge CW, Gerard RW (1932) The extra respiration of phagocytosis. Am J Physiol 103:235–236 12. Ado AD (1933) Über den verlauf der oxidativen und glykolytischen Prozesse in den Leukocyten des entzundeten Gewebes wahrend der Phagocytose. Z Ges Exp Med 87:473–480 13. Zinsser H, Enders JF, Fothergill LD (1939) Immunity. Principle and applications in medicine and public health. Resistance to infectious diseases, 5th edn. Macmillan, New York 14. Klebanoff SJ, Clem WH, Luebke RG (1966) The peroxidasethiocyanate-hydrogen peroxide antimicrobial system. Biochim Biophys Acta 117:63–72 15. Klebanoff SJ (1967) Iodination of bacteria: a bactericidal mechanism. J Exp Med 126:1063–1078 16. Lehrer RI, Cline MJ (1969) Leukocyte myeloperoxidase deficiency and disseminated candidiasis: the role of myeloperoxidase in resistance to Candida infection. J Clin Invest 48:1478–1488

J. T. Curnutte and A. I. Tauber 17. Sbarra AJ, Karnovsky ML (1959) The biochemical basis of phagocytosis. I. Metabolic changes during the ingestion of particles by polymorphonuclear leukocytes. J Biol Chem 234:1355–1362 18. Iyer G, Islam M, Quastel J (1961) Biochemical aspects of phagocytosis. Nature 192:535–541. https://doi.org/10.1038/192535a0 19. Paul B, Sbarra AJ (1968) The role of the phagocyte in host-parasite interactions. 13. The direct quantitative estimation of H2O2 in phagocytizing cells. Biochim Biophys Acta 156:168–178 20. Root RK, Metcalf J, Oshino N, Chance B (1975) H2O2 release from human granulocytes during phagocytosis. I. Documentation, quantitation, and some regulating factors. J Clin Invest 55:945–955 21. Cagan RH, Karnovsky ML (1964) Enzymatic basis of the respiratory stimulation during phagocytosis. Nature 204:255–256 22. Selveraj RJ, Sbarra AJ (1966) Relationship of glycolytic and oxidative metabolism to particle entry and destruction in phagocytosing cells. Nature 211:1272–1276 23. Holmes B, Page AR, Good RA (1967) Studies from the metabolic activity of leukocytes from patients with a genetic abnormality of phagocyte function. J Clin Invest 46:1422–1432 24. Mitchell P (1961) Coupling of phosphorylation to electron and hydrogen transfer by a chemi-osmotic type of mechanism. Nature 191:144–148 25. McCord JM, Fridovich I (1969) Superoxide Dismutase. An enzymatic function for erythrocuprein (hemocuprein). J Biol Chem 244: 6049–6055 26. Curnutte JT, Babior BM (1974) Biological defense mechanisms. The effect of bacteria and serum on superoxide production by granulocytes. J Clin Invest 53:1662–1672 27. Badwey JA, Curnutte JT, Robinson JM, Lazdins JK, Briggs RT, Karnovsky MJ, Karnovsky ML (1980) Comparative aspects of oxidative metabolism of neutrophils from human blood and guinea pig peritonea: magnitude of the respiratory burst, dependence upon stimulating agents, and localization of the oxidases. J Cell Physiol 105:541–551 28. Pincus SH, Klebanoff SJ (1971) Quantitative leukocyte iodination. N Engl J Med 284:744–750 29. Curnutte JT, Whitten DM, Babior BM (1974) Defective superoxide production by granulocytes from patients with chronic granulomatous disease. N Engl J Med 290:593–597 30. Winterbourn CC, Kettle AJ, Hampton MB (2016) Reactive oxygen species and neutrophil function. Annu Rev Biochem 85:765–792 31. Baehner RL, Karnovsky ML (1968) Deficiency of reduced nicotinamide-adenine dinucleotide oxidase in chronic granulomatous disease. Science 162:1277–1279 32. Iyer GYN, Quastel JH (1963) NADPH and NADH oxidation by guinea pig polymorphonuclear leukocytes. Can J Biochem Physiol 41:427–434 33. Rossi F, Zatti M (1964) Biochemical aspects of phagocytosis in polymorphonuclear leukocytes. NADH and NADPH oxidation by the granules of resting and phagocytizing cells. Exp Dermatol 20: 21–23 34. Babior BM, Curnutte JT, Kipnes RS (1975) Pyridine nucleotidedependent superoxide production by a cell-free system from human granulocytes. J Clin Invest 56:1035–1042 35. Curnutte JT (1980) Superoxide production by human neutrophils. Ph.D. Dissertation. Harvard University 36. Baehner RL, Gilman N, Karnovsky ML (1970) Respiration and glucose oxidation in human and guinea pig leukocytes: comparative studies. J Clin Invest 49:692–700 37. Badwey JA, Tauber AI, Karnovsky ML (1983) Properties of NADH-cytochrome b5 reductase from human neutrophils. Blood 62:152–157 38. Tauber AI, Wright J, Higson FK, Edelman SA, Waxman DJ (1985) Purification and characterization of the human neutrophil NADHcytochrome b5 reductase. Blood 66:673–678

2

The Phagocyte Oxidase: The Early Years

39. Badwey JA, Curnutte JT, Karnovsky ML (1979) The enzyme of granulocytes that produces superoxide and peroxide - an elusive pimpernel. N Engl J Med 300:1157–1160 40. Karnovsky ML, Badwey JA, Tauber AI (1981) How, where, and why phagocytic leukocytes produce superoxide and peroxide. In: Bloch K, Bolis L, Tosteson DC (eds) Membranes, molecules, toxins, and cells. John Wright and PSG, Inc, Boston, pp 163–173 41. Patriarca P, Cramer R, Moncalvo S, Rossi F, Romeo D (1971) Enzymatic basis of metabolic stimulation in leucocytes during phagocytosis: the role of activated NADPH oxidase. Arch Biochem Biophys 145:255–262 42. Curnutte JT, Karnovsky ML, Babior BM (1975) Manganesedependent NADPH oxidation by a particulate preparation from guinea pig granulocytes: an alternate explanation. Clin Res 23: 271A 43. Patriarca P, Dri P, Kakinuma K, Tedesco F, Rossi F (1975) Studies on the mechanism of metabolic stimulation in polymorphonuclear leukocytes during phagocytosis. I. Evidence for superoxide anion involvement in the oxidation of NADPH2. Biochim Biophys Acta 385:380–386 44. Curnutte JT, Karnovsky ML, Babior BM (1975) Manganesedependent NADPH oxidation by granulocyte particles. The role of superoxide and the nonphysiological nature of manganese requirement. J Clin Invest 57:1059–1067 45. Hohn DC, Lehrer RI (1975) NADPH-oxidase in X-linked chronic granulomatous disease. J Clin Invest 55:707–713 46. Curnutte JT, Kipnes RS, Babior BM (1975) Defect in pyridine nucleotide dependent superoxide production by a particulate fraction from the granulocytes of patients with chronic granulomatous disease. New Engl J Med 293:628–632 47. Curnutte JT, Tauber AI (1983) Failure to detect superoxide in human neutrophils stimulated with latex particles. Ped Res 17: 281–284 48. Babior BM, Curnutte JT, McMurrich BJ (1976) The particulate superoxide-forming system from human neutrophils. Properties of the system and further evidence supporting its participation in the respiratory burst. J Clin Invest 58:989–996 49. Dewald B, Baggiolini M, Curnutte JT, Babior BM (1979) Subcellular localization of the superoxide-forming enzyme in human neutrophils. J Clin Invest 63:21–29 50. Curnutte JT, Karnovsky ML, Babior BM (1979) Fluoride-mediated activation of the respiratory burst in human neutrophils. A reversible process. J Clin Invest 63:637–647 51. Tauber AI, Goetzl EJ (1979) Structural and catalytic properties of the solubilized superoxide-generating activity of human polymorphonuclear leukocytes. Solubilization, stabilization in solution, and partial characterization. Biochemistry 18:5576–5584 52. Gabig TG, Kipnes RS, Babior BM (1978) Solubilization of the O2—forming activity responsible for the respiratory burst in human neutrophils. Properties of the solubilized enzyme. J Biol Chem 253:6663–6665 53. Segal AW, Jones OTG (1978) Novel cytochrome b system in phagocytic vacuoles of human granulocytes. Nature 276:515–517 54. Segal AW, Jones OTG, Webster D, Allison AC (1978) Absence of a newly described cytochrome b from neutrophils of patients with chronic granulomatous disease. Lancet 2:446–449 55. Babior BM, Kipnes RS (1977) Superoxide-forming enzyme from human neutrophils: evidence for a flavin requirement. Blood 50: 517–524 56. Gabig TG, Babior BM (1979) The O2--forming oxidase responsible for the respiratory burst in human neutrophils. Properties of the solubilized enzyme. J Biol Chem 254:9070–9074 57. Light DR, Walsh C, O'Callaghan AM, Goetzl EJ, Tauber AI (1981) Characteristics of the cofactor requirements for the O2- – generating NADPH-oxidase of human polymorphonuclear leukocytes. Biochemistry 20:1468–1476

79 58. Tauber AI, Brettler DB, Kennington EA, Blumberg PM (1982) Relation of human neutrophil phorbol ester receptor occupancy and NADPH-oxidase activity. Blood 60:333–339 59. Cox JA, Jeng AY, Sharkey NA, Blumberg PM, Tauber AI (1985) Activation of the human neutrophil NADPH-oxidase by protein kinase C. J Clin Invest 76:1932–1938 60. Tauber AI (1987) Protein kinase C and the activation of the human neutrophil NADPH-oxidase. Blood 69:711–720 61. Cox JA, Jeng AY, Blumberg PM, Tauber AI (1987) Comparison of sub-cellular activation of the human neutrophil NADPH-oxidase by arachidonic acid, sodium dodecyl sulfate (SDS), and phorbol myristate acetate (PMA). J Immunol 138:1884–1888 62. Maridonneau-Parini I, Tauber AI (1986) Activation of NADPHoxidase by arachidonic acid involves phospholipase A2 in intact human neutrophils but not in the cell-free system. Biochem Biophys Res Commun 138:1099–1105 63. Maridonneau-Parini I, Tringale SM, Tauber AI (1986) Identification of distinct activation pathways of the human neutrophil NADPH-oxidase. J Immunol 137:2925–2929 64. Wright J, Schwartz JH, Olson R, Kosowsky JM, Tauber AI (1986) Proton secretion by the Na+/H+ antiporter in the human neutrophil. J Clin Invest 77:782–788 65. Wright J, Maridonneau-Parini I, Schwartz JH, Tauber AI (1988) The role of the Na+/H+ antiporter in the human neutrophil respiratory burst. J Leuk Biol 43:183–186 66. Okamura N, Curnutte JT, Roberts RL, Babior BM (1988) Relationship of protein phosphorylation to the activation of the respiratory burst in human neutrophils. Defects in the phosphorylation of a group of closely related 48-kD proteins in two forms of chronic granulomatous disease. J Biol Chem 263:6777–6782 67. Okamura N, Malawista SE, Roberts RL, Rosen H, Ochs HD, Babior BM, Curnutte JT (1988) Phosphorylation of the oxidaserelated 48K phosphoprotein family in the unusual autosomal cytochrome-negative and X-linked cytochrome-positive types of chronic granulomatous disease. Blood 72:811–816 68. Borregaard N, Tauber AI (1984) Subcellular localization of the human neutrophil NADPH-oxidase: b-cytochrome and associated flavoprotein. J Biol Chem 259:47–52 69. Higson FK, Durbin L, Pavlotsky N, Tauber AI (1985) Studies of cytochrome b-245 translocation in the PMA stimulation of the human neutrophil NADPH-oxidase. J Immunol 135:519–524 70. Badwey JA, Curnutte JT, Karnovsky ML (1981) Cispolyunsaturated fatty acids induce high levels of superoxide production by human neutrophils. J Biol Chem 256:12640–12643 71. Curnutte JT, Badwey JA, Robinson JM, Karnovsky MJ, Karnovsky ML (1984) Studies on the mechanism of superoxide release from human neutrophils stimulated with arachidonate. J Biol Chem 259:11851–11857 72. Badwey JA, Curnutte JT, Robinson JM, Berde CB, Karnovsky MJ, Karnovsky ML (1984) Effects of free fatty acids on release of superoxide and on change of shape by human neutrophils. Reversibility by albumin. J Biol Chem 259:7870–7877 73. Curnutte JT (1985) Activation of human neutrophil nicotinamide adenine dinucleotide phosphate, reduced (triphosphopyridine nucleotide, reduce) oxidase by arachidonic acid in a cell-free system. J Clin Invest 75:1740–1743 74. Tauber AI, Borregaard N, Simons ER, Wright J (1983) Chronic granulomatous disease: a syndrome of phagocyte oxidase deficiencies. Medicine 62:286–309 75. Curnutte JT, Babior BM (1987) Chronic granulomatous disease. In: Harris H, Hirshchorn K (eds) Advances in human genetics, vol 16. Plenum Publishing Corporation, pp 229–297 76. Bromberg Y, Pick E (1984) Unsaturated fatty acids stimulate NADPH-dependent superoxide production by cell-free system derived from macrophages. Cell Immunol 88:213–221

80 77. McPhail L, Clayton C, C. C., and Snyderman, R. (1984) Evidence that activation of human neutrophil NADPH oxidase involves association of a cytosolic factor with membrane components. Clin Res 32:315a 78. McPhail LC, Shirley PS, Clayton CC, Snyderman R (1985) Activation of the respiratory burst enzyme from human neutrophils in a cell-free system: evidence for a soluble cofactor. J Clin Invest 75: 1735–1739 79. Heyneman RA, Vercauteren RE (1984) Activation of a NADPH oxidase from horse polymorphonuclear leukocytes in a cell-free system. J Leukoc Biol 36:751–759 80. Curnutte JT, Kuver R, Scott PJ (1987) Activation of neutrophil NADPH oxidase in a cell-free system. Partial purification of components and characterization of the activation process. J Biol Chem 262:5563–5569 81. Curnutte JT, Berkow RL, Roberts RL, Shurin SB, Scott PJ (1988) Chronic granulomatous disease due to a defect in the cytosolic factor required for nicotinamide adenine dinucleotide phosphate oxidase activation. J Clin Invest 81:606–610 82. Hartshorn KL, Tauber AI (1988) The influenza virus-infected phagocyte: a model of deactivation. Hem Onc Clin N Am 2:301– 315 83. Hartshorn KL (2020) Innate immunity and influenza A virus pathogenesis: lessons for COVID-19. Front Cell Infect Microbiol 10: 563850 84. Gaudillière J, Löwy I (eds) (1998) The invisible industrialist: manufacture and the construction of scientific knowledge. Palgrave Macmillan, London 85. Keating P, Cambrosio A (2002) Biomedical platforms: realigning the normal and the pathological in late-twentieth-century medicine. MIT Press, Cambridge 86. Strevens M (2020) The knowledge machine. How irrationality created modern science. Liveright, New York

J. T. Curnutte and A. I. Tauber 87. Tauber AI (2017) Immunity, the evolution of an idea. Oxford University Press, New York 88. Zammito JH (2004) A nice derangement of epistemes. Postpositivism in the study of science from quine to latour. University of Chicago Press, Chicago 89. Felt, U. Fouche, R., Miller, C.A, and Smith-Doerr, L. (eds.) (2016) The handbook of science and technology studies, 4th ed. Cambridge: MIT Press 90. Fleck L ([1935] 1979) Genesis and development of a scientific fact. In Trenn TJ, Merton RK (eds.) Chicago: University of Chicago Press 91. Tauber AI, Chernyak L (1991) Metchnikoff and the origins of immunology. From metaphor to theory. Oxford University Press, New York 92. Metchnikoff E ([1892] 1893) Lectures on the comparative pathology of inflammation. In Starling FA, Starling EH (trans.). London: Kegan, Pauls Trench Trubner; reprinted 1968. New York: Dover 93. Tauber AI (1991) The immunological self: A centenary perspective. Perspect Biol Med 35:74–86 94. Tauber AI (2003) Metchnikoff and the phagocytosis theory. Nat Rev Mol Cell Biol 4:897–901 95. Bechtel W, Williamson RC (1998) Vitalism. In: Craig E (ed) Routledge encyclopedia of philosophy. Routledge, London 96. Metchnikoff E (1905) Immunity in infectious diseases. In Binnie FG (trans.) Cambridge: Cambridge University Press; reprint: 1968. New York: Johnson Reprint Corp, pp 175–206 97. Zinsser H (1914) Infection and resistance. Macmillan, New York 98. Silverstein AM (2009) A history of immunology, 2nd edn. Elsevier, Amsterdam 99. Mazumdar PMH (1995) Species and specificity. Cambridge University Press, Cambridge 100. Sbarra AJ, Strauss RR (eds) (1988) The respiratory burst and its physiological significance. Plenum Press, New York

3

Reflections on My Life in Noxes J. David Lambeth

Abstract

Personal perspectives and reflections of more than three decades in the Nox field are summarized, including influences that shaped my career in science. Keywords

Nox · Duox · NADPH-oxidase · Cytochrome b558 · Flavocytochrome · Reactive oxygen · Superoxide · Hydrogen peroxide

1

Overview and Dedication

I am grateful to the many students, postdocs and visiting faculty whose hard work and intelligence made the discovery of the Nox family possible. This chapter is dedicated to those individuals, and to the students and postdocs in the entire Nox field. It is not meant to be comprehensive or historically all-inclusive, but rather to summarize my lab’s contributions and some earlier studies and personalities who provided inspiration for my research directions and that led to the identification and study of Nox1 and other members of the Nox family. I apologize in advance to the many investigators whose work has advanced the Nox field, but who I have failed to cite. The field has grown and branched so enormously in the past two decades that a comprehensive summary would be impossible. Rather, other chapters in this book document more thoroughly and with extensive citations the pioneering contributions and histories of specific sub-areas. I focus here on my own journey in the Nox field, from graduate school in 1974 to retirement in 2018. My hope is that the interested students and postdocs draw some lessons that are useful as they pursue their own careers.

J. D. Lambeth (✉) Emory University School of Medicine, Atlanta, GA, USA

2

Roots and the Importance of Mentors

Let me say at the outset that I am an advocate of following one’s interests, even when they lead out of one’s comfort zone in terms of training, techniques, and funding. I have always found that upon entering a new area, while support maybe uneven for a few years, persistence pays and if the work is solid, funding from the National Institutes of Health and other agencies is eventually forthcoming. Anyway, as they say “change is good”, and is an antidote to boredom. I had not been trained in the phagocyte respiratory burst (a.k.a. oxidative burst) field as it was then called. Rather, as a graduate student at Duke, my research focus was enzymology, specifically having to do with the enzymatic mechanism by which electrons are passed from NADPH through a flavoprotein to an iron-sulfur protein and finally to the heme of a mitochondrial P-450 (P-450scc) that then metabolized cholesterol into steroid hormones. I was fortunate to have carried out my graduate work in a remarkable environment with a terrific mentor (Henry Kamin of flavoprotein fame, from whom I learned not only redox biochemistry but also a lot about life in general, including good wine and good food) along with very talented fellow students and postdocs. In a stroke of good fortune (I’ve had a lot of that during my career!), my mentor’s laboratory was adjacent to those of Irwin Fridovich and K.V. Rajagopalin (Raj for short). Irwin, of course, was the famed discoverer of superoxide dismutase, while Raj was an expert in the use of spectroscopic methods applied to enzymes. The three labs held a joint weekly meeting we called the “Wing Ding”, where I was immersed in the biochemistry of superoxide as well as in redox enzymology and biophysics, along with a healthy dose of kinetics and thermodynamics. Discussions were wideranging and seldom stuck to the topic, with more questions generated than answers. My interests and environment led me to an overarching interest in enzymes, especially redox enzymes, and the nitty gritty of their molecular mechanisms. At some point in my graduate career, I learned from Raj

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_3

81

82

J. D. Lambeth

about an interesting process in neutrophils called the “respiratory burst”, which involved the generation of superoxide, and its function (along with its metabolites) in killing invading microbes. Very little was known at the time about the process, but it looked to be enzyme-mediated. Nevertheless, my hands were tied with my Ph.D. project, so I filed it for later in the back of my mind.

3

The Prehistoric Era of the Phagocyte NADPH Oxidase

After a brief postdoc, I took my first faculty position in 1980 as an assistant professor at Emory Medical School in the Department of Biochemistry. I continued one branch arising from my postdoctoral work focused on the how P450scc is regulated, but I continued to read as much as I could about the respiratory burst. The enzymatic basis for phagocyte superoxide generation was mysterious at the time. While no one had succeeded in purifying and characterizing the enzyme responsible, elegant work from Segal’s group had shown that the enzyme, then termed the respiratory burst oxidase, contained a membrane-associated heme (soon to be called cytochrome b558) [1], and the same group later implicated FAD as a component of the enzyme [2]. I had read the work of Bromberg and Pick [3] who showed that its activity in a cell-free system could be triggered by anionic amphiphiles such as unsaturated fatty acids, a discovery that eventually proved to be key to identifying the enzyme components. The activation required both membrane and cytosol, implying a minimum of two components. Having had experience in the areas of electron transport, flavoproteins and membraneassociated cytochromes, this caught my attention. It was around this time that I uttered (at least in my head!) the words “Given my training in membrane enzyme purification and redox biology, this should be pretty straightforward”, and one of my postdocs began a “side project” to isolate the enzyme. I couldn’t have been more wrong about the “easy” part! Having quickly found that the enzyme activity was extremely unstable when extracted from membranes, we used the UV/visible spectrum of cytochrome b558 (rather than its enzymatic activity) to track the purification of the cytochrome component. We purified cytochrome b558 50-fold from a resting neutrophil membrane fraction [4] and did some early characterization, but because it lacked FAD, the preparation was inactive and of course as we now know, the cytosolic regulatory proteins were missing. We knew from Pick’s work that the activated enzyme must be comprised of two or more subunits, and posited that it must have been falling apart during attempts at purification. To try to get around these problems, we attempted to stabilize the activated enzyme by chemical cross-linking [5]. Minoru Tamura found that cross-linking the activated system resulted

in remarkable stabilization of the activity. However, the process turned the enzyme into something akin to teflon, and the material proved impossible to purify and sequence. Another heroic but frozen visiting scientist spent several months in a -20° freezer room attempting purification in ethylene glycol and other solvents, also to no avail. So much for “easy”, but as they say, “hindsight is 20-20”. Nevertheless, these early attempts whet our appetites to keep pursuing the problem, and what began as a side project began to consume more and more of our research efforts.

4

The “Only Child” Comes of Age: Identification of the Components of the Phagocyte Oxidase

The identification of the proteins associated with the NADPH oxidase was eventually sorted using molecular genetics as well as classical biochemistry, with important clues coming from several genetic forms of the clinical condition chronic granulomatous disease (CGD) in which one or another of the protein components was mutated leading to an inactive enzyme. Stuart Orkin’s group cloned cytochrome b558 as the missing component of X-linked CGD [6], and its sequence had homology with flavoprotein dehydrogenases, implying that it was a flavocytochrome (see Chap. 5 by Stuart Orkin). The same lab also identified and cloned p22phox as a second membrane-associated component [7]. Edgar Pick’s group [8] used a conventional biochemistry approach to fractionate the cytosol into two required fractions, and this later led to identification of Rac1 as one of the cytosolic factors [9]. Clark and Nauseef had developed an antibody made against the non-Rac cytosolic fraction, and this proved to be key to the cloning of these components by their group and the Malech group at NIH [10, 11]. Meanwhile, we had noted the potential of the human neutrophil to study signal transduction and had been working on various upstream signaling systems that included protein kinase C and phospholipase D that were involved in activating the oxidase e.g [12–14].

5

Sorting Out the Protein-Protein Interactions That Regulate the Phagocyte NADPH Oxidase

After key players were identified, we returned to biochemical/kinetic/thermodynamic approaches to characterize the multiple protein-protein interactions that govern oxidase activation. Several groups had used immunochemical methods to study the translocation of p47phox p67phox and Rac in intact cells, e.g. [15]. Our group sought to study the assembly process in a cell-free system, which would allow a kinetic

3

Reflections on My Life in Noxes

and thermodynamic description of the interaction of proteins and their sub-domains with one another. This required a deep dive into then-emerging technologies using recombinant DNA in non-mammalian cells as factories to express large amounts of proteins. A postdoc, David Uhlinger, figured out how to express p47phox and p67phox in insect cells as well as Rac in bacteria, and we developed what we called the “semirecombinant system” [16, 17], which used the isolated plasma membranes from resting neutrophils in combination with recombinant cytosolic factors. This allowed the study of the oxidase in a “clean” system wherein concentrations of each component were known, permitting the quantification of EC50 values for binding of each component and mutants thereof. During the 1990s, we used this system to dissect details of the protein-protein interactions that were involved in forming the active complex. Among the conclusions that came from these studies were the following: 1. Efficient translocation of p67phox required p47phox but not vice versa. Efficient translocation of p47phox, on the other hand, required the cytochrome [17]. The system could achieve full activity in the absence of p47phox if high concentrations of p67phox were used [18], a finding made independently by Pick’s group [19]. These studies implied that p47phox was functioning as an “adapter protein” to allow efficient binding of p67phox and that p67phox and/or Rac was responsible for activating the cytochrome. 2. Cytosolic components bind to the cytochrome in a highly synergistic manner (i.e., the binding of one increased the binding of the others). This suggested that rather than assembling independently, all three were joined in some manner in the active complex [20]. The stoichometry of the active complex of p47phox, p67phox and cytochrome b558 was 1:1:1 [17, 21]. 3. GTP analogs enhanced the binding of the other components, implicating Rac in the assembly [17, 22]. An “insert region” on Rac was important for activation [23, 24]. Mutational studies showed that an “effector region” that was known to change conformation in response to GTP binding, associated with p67phox but not p47phox [21]. Mutations in the insert region weakened its binding to the complex, but did not affect direct binding to p67phox suggesting that this region might bind to the cytochrome.

6

Discovery of the Activation Domain of p67phox and Its Regulation of Hydride Transfer from NADPH to FAD

Among the important discoveries during this period, we identified an “activation domain” in p67phox as essential for triggering the electron flow within flavocytochrome b558. The activation domain was needed regardless of whether p47phox

83

was present [25]. Mutations in this region did not affect the binding to Rac. However, such mutant versions of p67phox competed with wild-type p67phox in activation and assembly. These studies pointed to an interaction of the activation domain with the flavocytochrome as the key step in activating electron transfer from NADPH to oxygen. The sum of the studies of the protein interactions supported a model in which p47phox and Rac acted as regulated adapter or “chaperone” proteins that responded to signaling pathways (phosphorylation of p47phox and GTP binding to Rac) to help escort p67phox to flavocytochrome b558 where its activation domain triggered electron flow from NADPH to oxygen. We then turned our attention to identifying the mechanism by which the activation domain on p67phox turns on electron transfer within the flavocytochrome. While no one had succeeded in solubilizing and purifying active cytochrome b558 with bound FAD, Segal et al. [26] had developed a method to isolate homogeneous flavocytochrome in an inactive form that lacked FAD. Using a modification of this method, we developed a preparation that, although inactive, could be reconstituted into an artificial phospholipid membrane and could then be stably reconstituted with FAD and FAD analogues in a ratio of one flavin to two hemes [27]. Reconstitution rendered the preparation fully active. Moreover, while FAD itself was non-fluorescent when bound, one of the analogues in its oxidized form was highly fluorescent, and reduction of the FAD rendered it non-fluorescent. This provided a tool to monitor its reduction state and explore the regulated step. Earlier hypotheses proposed variously that NADPH binding or the transfer of electrons from FAD to heme or from heme to oxygen was rate-determining and therefore the regulated step. However, by tracking the oxidation state of the fluorescent FAD analogue, wild-type p67phox but not its inactive or less active versions mutated in the activation domain increased the steady state % reduction of FAD, but the heme remained in the fully oxidized form in all cases [28]. If any downstream electron transfer had been rate-limiting, the opposite pattern would have been seen. These data implied that once an electron gets to the hemes, it is removed by oxygen much more rapidly than it can be replaced by upstream processes. Thus, FAD reduction but not FAD-toheme/oxygen electron transfer was the regulated step. Using deuterated NADPH (i.e., NADPD) the system showed a fairly large deuterium isotope effect, which would have been obscured had other steps such as NADPH binding contributed significantly to the slow step. In addition, there was very little effect of p67phox on the binding constant for NADPH, also eliminating its binding as a regulated step. Thus, the hydride transfer from bound NADPH to FAD was the slow step and was regulated by the p67phox activation domain. We confirmed this by showing that electron transfers to dye by the dehydrogenase domain alone, which contained the FAD and NADPH-binding site but lacked the heme, was

84

J. D. Lambeth

regulated by p67phox, showing that the FAD-containing domain (a.k.a., dehydrogenase domain) of gp91phox is the target of regulation by the p67phox activation domain [29]. For these (and many other) studies over the years, I give a shout-out to my long-term very talented colleague and good friend, Yukio Nisimoto, who visited my lab from Japan on a 2-year sabbatical in the mid-1980s and continued to visit nearly every summer for the next 25 years. To my knowledge, a structural explanation for how p67phox facilitates hydride transfer is still unknown and remains in my mind an important question. Nevertheless, detailed structural information is needed and the flavocytochrome has proven stubbornly resistant to crystallization efforts, of which there have been many attempts both in my lab and in others. Neither crystalized subdomains nor molecular models of the flavocytochrome proved to be helpful in this regard, but perhaps cryo-electron microscopy will provide insights. Suffice it to say that some sort of conformational change induced by p67phox must somehow align the pyridinium ring of NADPH with the N5 of FAD to form the proper geometry to allow hydride transfer. It should be cautioned that such an alignment could be prevented by quite subtle effects (rather than a gross conformationl change) in the active site geometry. For example a single amino acid might move out of the way, allowing a productive complex. The author suggests that detailed structural comparison of Nox2 with the constitutively active Nox4 may be a productive approach.

7

Discovery of the Noxes and Duoxes: The Only Child Becomes a Member of a Family

While the above work on the phagocyte oxidase was underway, I became curious about a number of reports, particularly from the laboratory of O.T.G. Jones [30], that reactive oxygen was being generated by a variety of other cell types, often in response to hormones or growth factors, and that these activities had certain properties in common with the phagocyte oxidase. A little later, fragments of DNA began appearing in online databases as a result of early genome sequencing efforts, and one of these was homologous to gp91phox. We used this sequences as a starting point for the cloning of Nox1 (first called Mox1) [31]. Guangie (Jeff) Cheng, a whirlwind molecular biologist, rapidly cloned and reported sequences of Nox3, Nox4, and Nox5 [32, 33]. Other labs [34, 35] also reported some of these as well as Duox1 and Duox2 [36] at around the same time, which completed the family of mammalian Nox and Duox enzymes. Nox enzymes were also seen throughout the plant and animal

kingdoms, e.g., see [37], pointing to the universal biological importance of the enzyme family. It is worth a digression here to mention the origin of the Nox and Duox naming system. At the time these new genes were being cloned, a “Tower of Babel” of various nomenclatures had begun to emerge from various labs, including Thox, NOH-1, Mox1, and Renox, named for tissue locations, presumed functions or other criteria. Karl-Heinz Krause and I discussed the problem and felt that given the widespread tissue localizations and uncertainties regarding biological functions, none of these names made a lot of sense. Karl-Heinz contacted HUGO, the Human Genome Organization, and we worked with them to come up with the Nox/Duox terminology. gp91phox was designated with the name Nox2, not because of its temporal discovery, but since we and HUGO assumed that it would continue to be known mainly by its historic name. Nevertheless, Nox2 stuck and is now routinely used.

8

What in the World Are All These Nox and Duox Enzymes Doing?

The widespread distribution of Nox/Duox isoforms in a variety of mammalian tissues and organisms along with the fact that many were regulated by hormones or growth factors raised the question: what are their biological roles? Prior to this, reactive oxygen species (ROS) had been mostly described as “toxic” causing chain chemical reactions that wreaked havoc on biomolecules including proteins, nucleic acids, and lipids. The phagocyte oxidase was a prime example of nature harnessing the chemical reactivity of ROS to help kill invading microbes, and had been implicated in inflammatory conditions in which host “bystander” biomolecules were accidentally damaged. It was tempting, therefore, to try to cast the new Noxes in the same mold. However, the new enzymes were typically expressed at much lower levels than the phagocyte oxidase resulting in much lower levels of ROS. There were examples in the literature suggesting a role for ROS in signal transduction, but these were largely indirect. To approach this question, we began to develop a Nox1 knockout mouse. At that time the knockout technology was in its infancy and progress was painfully slow. Meanwhile, we began genetic studies to knock out Nox/Duox in simple organisms, where knockout (or “knock down”) technologies were simple and rapid. These areas lay well outside the expertise of my lab, but fortunately, a couple of expert collaborators at Emory made this work possible. Our first effort in this regard used C. elegans as a model organism, a collaboration between Lisa Sharling and Bill Edens in my lab with Guy Benian [38]. In addition to the ease of suppressing specific gene expression in these animals

3

Reflections on My Life in Noxes

by RNA interference (RNAi), worms had only two genes in the Nox family, both Duox homologues, and one of these appeared to be a pseudogene (i.e., not expressed). Because in mammals, the Noxes might share overlapping or redundant functions, this made C. elegans a far simpler model system than mouse. After cloning the cDNA, expressing the protein and making antibodies, localization studies showed that the enzyme resided in the outermost cell layer immediately below the cuticle, the tough collagenous protective layer surrounding the worm. RNAi suppressed the expression of Ce-Duox1 and resulted in a a dramatic phenotype in which the cuticle fragmented easily and showed large “blisters”. Biochemically, the cuticle is stabilized by cross-linking of tyrosines to form di- and tri-tyrosine, and this linkage was eliminated in RNAi worms. Using the expressed peroxidase domain of Ce-Duox, this region, like other peroxidases, catalyzed H2O2-dependent tyrosine cross-linking. We suggested a model in which the transmembrane Ce-Duoxderived H2O2 supported cross-linking of tyrosines by both its own ecto-facing peroxidase domain as well as other secreted peroxidases. The importance of these studies was that it demonstrated a role of Duox-type enzymes in catalyzing chemical modifications of extracellular matrices. The iodination of thyroid hormone, supported by Duox-generated H2O2 seems to be a related example, although the peroxidase-like domains of mammalian Duox enzymes do not appear to have true peroxidase activity, making iodination reactions likely dependent on other secreted peroxidases. No doubt there are other examples awaiting investigation. Another genetic approach in a model organism was carried out in Drosophila by Darren Ritsick in collaboration with Vicky Finnerty, a colleague and fly expert [39]. Nox-mediated generation of superoxide or H2O2 in response to agonists had been seen in cell lines including in Angiotensin 2-stimulated vascular smooth muscle [40] and in insulin-stimulated fat cells [41], and had been implicated in signaling. To our knowledge, proof of the participation of a Nox signaling pathway linked to a specific biological function in a whole animal did not previously exist. The Drosophila genome encodes two Noxes, d-Nox (a Nox5 homologue) as well as d-Duox, and RNAi had been developed to knock down expression of specific genes in this organism. Upon interfering with the expression of d-Nox, female flies showed sterility and massive distention of the abdomen attributable to retention of eggs. Eggs were mature and fertile, but failed to be expelled, which turned out to be a defect in smooth muscle function. The muscle developed normally but failed to contract. Smooth muscle contraction and ovulation in flies involves the hormone proctolin, which triggers ROS generation and caused an elevation in muscle

85

calcium. When d-Nox expression was knocked out, the ROS generation was prevented and the calcium elevation and resulting muscle contraction were likewise blocked. The calcium elevation was restored when H2O2 was added along with proctolin. These studies established a role for a Nox in a signaling pathway in an intact animal and provided a paradigm for investigations of Nox roles in mammalian smooth muscle.

9

Nox4, the Odd-Ball of the Family: An Oxygen-Sensing Enzyme?

I’ve always been intrigued in science by things that don’t fit the established paradigm, and feel that these exceptions sometimes augur exciting discoveries. Nox4 was one such exception to the rules. For years I had been puzzled by its differences from other Nox and Duox enzymes: its constitutive activity, its independence from regulatory subunits or domains, and the lack of its activation by any known agonists. This made it unique among the Nox/Duox family, all of which seemed to be regulated by specific proteins or domains. In a fairly recent study [42], we proposed an unusual signaling function for mammalian Nox4: that it functions as an oxygen sensor. This was based on the following: because of technical challenges to do otherwise, nearly all of us carry out experiments under room oxygen conditions (21%). However, Nox/Duox enzymes do not live in this environment in tissues. For example, in the peripheral vasculature where Nox4 is expressed, oxygen tensions can measure a few %, or even lower, e.g. in exercising muscle. We found that while Nox2 has a low Km, around 2–3% O2, Nox4 has an anomalously high Km, nearly ten-fold higher. This meant that while Nox2 is able to function at nearly full speed in relatively hypoxic environments, Nox4 activity is severely limited by oxygen availability in a range within physiological span of oxygen tensions. This study also confirmed an earlier study [43] that Nox4 generates H2O2 and not superoxide, and showed that free superoxide is not released by the enzyme. We proposed [42] that a specific histidine in Nox4, which when mutated allows superoxide rather than H2O2 to be formed [43], stabilizes enzyme-bound superoxide until a second electron transfer from heme converts it into H2O2. H2O2 had previously been shown to participate in signaling pathways, e.g. in muscle, and this study established that Nox4 possesses all the properties needed to permit it to function as an “oxygen sensor”, i.e., high Km for oxygen and direct generation of a signaling molecule. In my opinion, this hypothesis remains inadequately explored or tested in in vivo systems, and provides a promising opportunity for further investigations.

86

10

J. D. Lambeth

Similarities and Differences Among Nox/Duox Enzymes

During the 2000s and 2010s, my lab also described other functional and structural features of Nox family members, especially aspects that differed from the Nox2 system which had already been characterized extensively. Some of our conclusions are listed briefly: 1. Unlike p47phox its homologue NOXO1 (which lacks regulatory domains of the former, e.g. phosphorylation sites), co-localizes with Nox1 in membranes in the absence of cell activation, and does so in part via lipid binding by its PX domain [44]. Alternative mRNA spliced forms of NOXO1 differing in the PX domain were also identified and characterized [45]. 2. Surprisingly, human (but not rodent) NOX3 could be activated by NOXO1 in the absence of NOXA1, supporting a model in which in Nox3, multiple interactions with regulatory subunits stabilize an active conformation of the catalytic subunit [46]. Thus, the “activation domain” paradigm developed around Nox2 regulation does not seem to be a sine qua non for regulating some Nox isoforms. 3. Inactivating point mutations in p22phox function as dominant inhibitors of Nox1 and Nox2, providing a tool for investigating e.g. the source of ROS in some cells [47]. 4. Rac activation of Nox1 was characterized including binding of Rac to both NOXA1 and Nox1. Since NOXO1 and NOXA1 but not Rac are already constitutively bound to Nox1, this suggested a model in which activation of Rac may serve as the primary trigger for Nox1 activation [48]. 5. Phosphatidylinositol 4,5 bisphosphate binding to an N-terminal polybasic region of Nox5 mediates its cellular localization to the plasma membrane [49]. Mutation of this region resulted in mis-localization to internal membranes. 6. The Nox4 B-loop (located in the heme-containing transmembrane domain) provides an interface with the FAD domain [50]. The Nox4 dehydrogenase domain alone transferred electrons from NADPH to artificial dyes, but did not produce H2O2 without the heme domain [51]. This showed that the constitutive activity of Nox4 is due to absence of Nox2-like “brakes” within its dehydrogenase domain: the Nox2 dehydrogenase domain also transferred electrons to artificial acceptors, but required regulatory subunits (see above). 7. Nox5, which does not require subunits and is independent of p22phox, forms a complex in the plane of the membrane with itself to form an oligomer of a size consistent with a tetramer [52]. The association is mediated by the dehydrogenase domain. Such a complex provides an explanation for why mutant versions of Nox5 that are catalytically

inactive inhibit wild-type Nox5. Inactive versions of other Nox isoforms also inhibit their wild-type counterparts, suggesting that in-membrane oligomerization is a general feature of the Nox family. Collectively, these studies elucidated some of the functional and regulatory similarities and differences among the different Nox isoforms.

11

Nox Enzymes and Human Diseases

In addition to investigating the enzymatic and regulatory properties of then new Nox isoforms, my lab was inevitably drawn into a variety of more clinically-related studies, many in collaboration with labs that were expert in these areas. Reactive oxygen had been previously implicated in a variety of diseases, and the hypothesis was that over activity or over expression of specific Nox isoforms participated in their pathobiology. My lab’s biomedical focus during the early 2000s was in the area of cancer, as we had originally cloned Nox1 from a colon cancer cell line. I briefly summarize cancer-relevant work, and then list for the interested reader disease-related studies in other areas carried out in collaborator labs. In a majority of human colon tumors compared with paired normal tissue controls from the same anatomical region in the same patient, we saw up to five-fold overexpression of Nox1 at fairly early tumor stages. Overexpression correlated with activating mutations in K-Ras [53]. A similar induction of Nox1 was seen in transgenic mice over-expressing mutationally activated K-Ras in the colon [53], implying Nox1 expression was regulated by K-Ras. Kamata’s lab had found that transformation of cells by V12-Ras was dependent on Nox1 and that the phenotype was blocked by RNAi against Nox1 [54]. Nox1 was also implicated in normal proliferation and repair of intestinal cells [55] including stimulation of repair by commensal bacteria [56], and it seems likely that some forms of colon cancer might represent this normal mechanism “gone rogue”. Whether Nox participates in the pathology of human cancers has not been definitively established to my knowledge, but it is intriguing that Nox isoforms have been implicated in collaborator labs in growth and angiogenesis in prostate cells/cancers [57–59], growth patterns of melanoma cells [60], cyclin D1 expression and cell proliferation in lung epithelial cells [61], esophageal cancers and precancerous lesions [62, 63], and genome instability [64]. Nox isoforms were studied by collaborators in a variety of other disease states. These include hypertension [65], atherosclerosis [66, 67], restenosis injury [68], allergic responses [69], glomerulonephritis [70], and responses to viral infections [71, 72].

3

12

Reflections on My Life in Noxes

Adventures (Misadventures?) in Drug Discovery

While work was ongoing that implicated various Nox isoforms in various human diseases, I became interested in the idea of developing small molecule inhibitors of Noxes that might prove useful as drugs [73–75]. I suppose this had something to do with the hope among many (especially older) scientists to see their basic research become useful beyond the scientific community. We got into this business naively, and in the beginning I did not appreciate the many difficulties and enormous expense associated with drug development. Aided by an opportunely timed NIH-funded robotic screening facility at Emory, we embarked on a drug discovery program, which involved development of miniaturized screening assays for Nox isoform-dependent reactive oxygen, secondary assays of initial “hits” to eliminate false positives, screening out of compounds with undesirable chemical properties, re-synthesizing and re-testing larger quantities of bona fide inhibitors with a panel of cells expressing specific Nox isoforms to determine inhibitor selectivity, and finally optimizing structures by synthesizing and testing chemically modified versions of promising hits. Starting with chemical libraries of hundreds of thousands of compounds, we ended up with somewhere around ten validated molecules. After years of work by several very talented organic chemistry-proficient postdocs and faculty, only a few showed potential as drugs. During the course of this work, we also characterized the mechanisms of some known inhibitors, including ebselen [76]. One of the challenges we (and others) faced was artifacts built into assay systems used for robotic screening. For example, many screening assays required a peroxidase such as myeloperoxidase to visualize the reactive oxygen signal, meaning that there were many false positives that ended up being inhibitors of peroxidases rather than Nox enzymes, e.g., [77]. In a sow’s-ear-to-silk purse story, one of these proved useful in understanding inflammatory responses to influenza [72, 78]; perhaps such molecules may one day prove helpful in treating other viral diseases including possibly Covid-19. Nox drug development was eventually taken over by GenKyoTex, a company I co-founded with KarlHeinz Krause, Bob Clark and Chihiro Yabe-Nishimura, which has focused on treating fibrotic diseases (e.g., liver, lung and kidney fibrosis). Although no Nox-targeted drugs have yet been approved for use, I am optimistic that Nox isoforms represent an important target for a variety of diseases including some “orphan” diseases for which useful therapies are not available. I sincerely hope that Nox drug efforts will continue as I pursue “second careers” and pass times which included music, painting and enjoying the beauty of the mountains of northwest Colorado. I also look

87

forward to following much exciting research being carried out in the Nox field.

References 1. Segal AW, Jones OTG (1978) Novel cytochrome b system in phagocytic vacuoles of human granulocytes. Nature 276:515–517 2. Cross AR, Jones OTG, Garcia R, Segal AW (1982) The association of FAD with the cytochrome b-245 of human neutrophils. Biochem J 208:759–763 3. Bromberg Y, Pick E (1984) Unsaturated fatty acids atimulate NADPH-dependent superoxide production by cell-free system derived from macrophages. Cellular Immunol 88:213–221 4. Pember SO, Heyl BH, Kinkade JM, Lambeth JD (1984) Cytochrome b558 from bovine granulocytes: partial purification from Triton X-114 extracts, and properties of the isolated cytochrome. J Biol Chem 259:10590–10595 5. Tamura M, Tamura T, Burnham DN et al (1989) Stabilization of the superoxide-generating respiratory burst oxidase of human neutrophil plasma membrane by crosslinking with 1-ethyl-3-(3-dimetylaminopropyl) carbodiimide. Arch Biochem Biophys 275:23–32 6. Royer-Pokora B, Kunkel LM, Monaco AP et al (1986) Cloning the gene for an inherited human disorder – chronic granulomatous disease – on the basis of its chromosomal location. Nature 322:32– 38 7. Parkos CA, Dinauer MC, Walker LE, Allen RA et al (1988) Primary structure and unique expression of the 22-kilodalton light chain of human neutrophil cytochrome b. Proc Natl Acad Sci USA 85:3319– 3323 8. Abo A, Pick E (1991) Purification and characterization of a third cytosolic component of the superoxide-generating NADPH oxidase of macrophages. J Biol Chem 266:23577–23585 9. Abo A, Pick E, Hall A, Totty N et al (1992) Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1. Nature 353:668–670 10. Leto TL, Lomax KJ, Volpp BD et al (1990) Cloning of a 67-kD neutrophil oxidase factor with similarity to a noncatalytic region of p60c-src. Science 248:727–730 11. Lomax KJ, Leto TL, Nunoi H et al (1989) Recombinant 47-kilodalton cytosol factor restores NADPH oxidase in chronic granulomatous disease. Science 245:409–412 12. Lambeth JD (1989) Activation of the respiratory burst oxidase in neutrophils: on the role of membrane-derived second messengers, Ca++ and protein kinase C. J Bioenerg Biomembr 20:709–733 13. Perry DK, Hand WL, Edmondson DE, Lambeth JD (1992) On the role of phospholipase D-derived diradylglycerol in the activation of the human neutrophil respiratory burst oxidase: inhibition by phosphatidic acid phosphohydrolase inhibitors. J Immunol 149: 2749–2758 14. Lopez I, Arnold RS, Lambeth JD (1998) Cloning and initial characterization of a human phospholipase D2 (hPLD2). J Biol Chem 273: 12846–12852 15. Heyworth PG, Bohl BP, Bokoch GM, Curnutte JT (1994) Rac translocates independently of the neutrophil NADPH oxidase components p47phox and p67phox. Evidence for its interaction with flavocytochrome b558. J Biol Chem 269:30749–30752 16. Uhlinger DJ, Inge KL, Kreck ML et al (1992) Reconstitution and characterization of the human neutrophil respiratory burst oxidase using recombinant p47-phox, p67-phox and plasma membrane. Biochem Biophys Res Commun 186:509–516 17. Uhlinger DJ, Tyagi SR, Inge KL, Lambeth JD (1992) Guanine nucleotides regulate the assembly of the human neutrophil

88 respiratory burst oxidase: evidence for G protein regulation of the binding of p67-phox and p47-phox. J Biol Chem 268:8624–8631 18. Freeman JL, Lambeth JD (1996) NADPH oxidase activity is independent of p47-phox in vitro. J Biol Chem 271:22578–22585 19. Koshkin V, Lotan O, Pick E (1996) The cytosolic component p47phox is not a sine qua non participant in the activation of NADPH oxidase but is required for optimal superoxide production. J Biol Chem 271:30326–30329 20. Uhlinger DJ, Taylor KL, Lambeth JD (1994) p67-phox enhances the binding of p47-phox to the human neutrophil respiratory burst oxidase complex. J Biol Chem 269:22095–22098 21. Nisimoto Y, Freeman JL, Motalebi SA et al (1997) Rac binding to p67phox: structural basis for interactions of the Rac1 effector region and insert region with components of the respiratory burst oxidase. J Biol Chem 271:18834–18841 22. Kreck ML, Uhlinger DJ, Tyagi SR et al (1993) Participation of the small molecular weight GTP binding protein Rac1 in cell-free activation and assembly of the respiratory burst oxidase: inhibition by a C-terminal Rac peptide. J Biol Chem 269:4161–4168 23. Freeman JL, Kreck ML, Uhlinger DJ, Lambeth JD (1994) Ras effector-homologue region on Rac regulates protein associations in the neutrophil respiratory burst oxidase complex. Biochemistry 33: 13431–13435 24. Freeman JL, Abo A, Lambeth JD (1996) Rac “insert region” is a novel effector region that is implicated in the activation of NADPH oxidase, but not PAK65. J Biol Chem 271:19794–19801 25. Han C-H, Freeman JL, Lee T et al (1998) Regulation of the neutrophil respiratory burst oxidase: identification of an activation domain in p67phox. J Biol Chem 273:16663–16668 26. Segal AW, West I, Wientjes F (1992) Cytochrome b-245 is a flavocytochrome containing FAD and the NADPH-binding site of the microbicidal oxidase of phagocytes. Biochem J 284:781–788 27. Nisimoto Y, Otsuka-Murakami H, Lambeth JD (1995) Reconstitution of flavin-depleted neutrophil flavocytochrome b558 with 8-mercapto-FAD and characterization of the flavin-reconstituted enzyme. J Biol Chem 270:16428–16434 28. Nisimoto Y, Motalebi S, Han C-H, Lambeth JD (1999) The p67phox activation domain regulates electron flow from NADPH to flavin in flavocytochrome b558. J Biol Chem 274:22999–23005 29. Han C-H, Nisimoto Y, Lee S-H et al (2001) Characterization of the flavoprotein domain of gp91phox which has NADPH diaphorase activity. J Biochem 129:513–520 30. Cross AR, Jones OTG (1991) Enzymatic mechanisms of superoxide production. Biochim Biophys Acta 1057:281–298 31. Suh Y-A, Arnold RS, Lassegue B et al (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401:79–82 32. Lambeth JD, Cheng G, Arnold RS, Edens WA (2000) Novel homologs of gp91phox. Trends Biochem Sci 25:459–461 33. Cheng G, Cao Z, Xu X et al (2001) Homologs of gp91phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269:131–140 34. Banfi B, Maturana A, Jaconi S et al (2000) A mammalian H+ channel generated through alternative splicing of the NADPH oxidase homolog NOH-1. Science 287:138–142 35. Geiszt M, Kopp JB, Várnai P, Leto T (2000) Identification of Renox, an NAD(P)H oxidase in kidney. Proc Nat Acad Sci USA 97:8010– 8014 36. De Deken X, Wang D, Many M et al (2000) Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275:23227–23233 37. Kawahara T, Quinn MT, Lambeth JD (2007) Molecular evolution of the NADPH oxidase (Nox, Duox) family of enzymes. BMC Evol Biol 7:109–115 38. Edens WA, Sharling L, Cheng G et al (2001) Tyrosine cross-linking of extracellular matrix is catalyzed by Duox, a multidomain oxidase/ peroxidase with homology to the phagocyte oxidase subunit gp91phox. J Cell Biol 154:879–891

J. D. Lambeth 39. Ritsick DR, Edens WA, Finnerty V, Lambeth JD (2007) Nox regulation of smooth muscle contraction. Free Radic Biol Med 43:31–38 40. Lassegue B, Sorescu D, Szöcs K et al (2001) Novel gp91phox homologues in vascular smooth muscle cells: Nox1 mediates angiotensin II-induced xuperoxide formation and redox-sensitive signaling pathways. Circ Res 88:888–894 41. Mahadev K, Motoshima H, Wu X et al (2004) The NADPH oxidsase homolog Nox4 modulates insulin-stimulated generation of H2O2 and plays an integral role in insulin signal transduction. Mol Cell Biol 24:1844–1845 42. Nisimoto Y, Diebold B, Constentino-Gomez D, Lambeth JD (2014) Nox4: a hydrogen peroxide generating oxygen sensor. Biochemistry 53:5111–5120 43. Takac I, Schroder K, Zhang L et al (2011) The E-loop is involved in hydrogen peroxide formation by the NADPH oxidase Nox4. J Biol Chem 286:13304–13313 44. Cheng G, Lambeth JD (2004) NOXO1: regulation of lipid binding, localization and activation of Nox1 by the phox homology (PX) domain. J Biol Chem 279:4737–4742 45. Cheng G, Lambeth JD (2005) Alternative mRNA splice forms of NOXO1: differential tissue expression and regulation of Nox1 and Nox3. Gene 356:118–126 46. Cheng G, Ritsick DR, Lambeth JD (2004) Nox3 regulation by NOXO1, p47phox and p67phox. J Biol Chem 279:35250–34255 47. Kawahara TD, Ritsick DR, Cheng G, Lambeth JD (2005) Point mutations in the proline rich region of p22phox are dominant inhibitors of Nox1- and Nox2-dependent reactive oxygen generation. J Biol Chem 280:31859–31869 48. Cheng G, Diebold B, Hughes Y, Lambeth JD (2006) Nox1dependent reactive oxygen generation is regulated by Rac1. J Biol Chem 281:17718–17726 49. Kawahara T, Lambeth JD (2008) Phosphatidylinositol (4,5)bisphosphate modulates Nox5 localiztion via an N-terminal polybasic region. Mol Cell Biol 19:4020–4031 50. Jackson HM, Kawahara T, Nisimoto Y et al (2010) Nox4 B-loop creates an interface between the transmembrane and dehydrogenase domains. J Biol Chem 285:10281–10290 51. Nisimoto Y, Jackson HM, Ogawa H et al (2010) Constitutive NADPH electron transferase of the Nox4 dehydrogenase domain. Biochemistry 49:2433–2442 52. Kawahara TD, Jackson HM, Smith SM et al (2011) Nox5 forms a functional oligomer mediates by self-association of its dehydrogenase domain. Biochemistry 50:2013–2025 53. Laurent E, McCoy JW, Macina RA et al (2008) Nox1 is overexpressed in human colon cncers and correlates with activating mutations in k-Ras. Int J Cancer 123:100–107 54. Mitsushita J, Lambeth JD, Kamata T (2004) The superoxidegenerating oxidase Nox1 is functionally required for Ras oncogene transformation. Cancer Res 64:3580–3585 55. Leoni G, Alam A, Neumann AA et al (2013) Anexin A1, formyl peptide receptor, and NOX1 orchestrate epithelial repair. J Clin Invest 123:443–454 56. Jones RM, Luo L, Ardita CS (2013) Symbiotic lactobacilli stimulate gut epithelial proliferation via Nox-mediated generation of reactive oxygen species. EMBO J 32:3017–3028 57. Brar SS, Corbin TP, Kennedy R et al (2003) NOX5 NADPH oxidase regulates growth and apoptosis in DU145 prostate cancer cells. Am J Physiol Cell Physiol 285:C353–C369 58. Arbiser JL, Petros J, Klafter R et al (2002) Reactive oxygen generated by Nox1 triggers the angiogenic switch. Proc Natl Acad Sci USA 99:715–720 59. Lim SD, Sun C, Lambeth JD et al (2005) Increased Nox1 and hydrogen peroxide in prostate cancer. Prostate 2005:200–207 60. Govindarajan B, Sligh JE, Vincent BJ et al (2007) Overexpression of Akt converts radial growh melanoma to vertical growth melanoma. J Clin Invest 117:719–729

3

Reflections on My Life in Noxes

61. Ranjan P, Anathy V, Burch PM et al (2006) Redox-dependent expression of cyclin D1 and cell proliferation by Nox1 in mouse lung epithelial cells. Antioxid Redox Signal 8:1447–1459 62. Fu X, Beer DG, Behar J et al (2006) cAMP-response elementbinding protein mediates acid-induced NADPH oxidase NOX5-S expression in Barret esophageal adenocarcinoma cells. J Biol Chem 281:20368–20382 63. Hong J, Behar J, Wands J et al (2010) Bile acid reflux contributes to developent of adenocarcinoma via activation of phosphatidylinositol-specific phospholipase Cgamma2 and NADPH oxidase NOX5-S. Cancer Res 70:1247–1255 64. Chiera F, Meccia E, Degan G et al (2008) Overexpression of human NOX1 complex induces genome instability in mammalian cells. Free Radic Biol Med 44:332–342 65. Dikalova A, Clempus R, Lassegue B et al (2005) Nox1 overexpression potentiates angiogensin II-induced hypertension and vascular smooth muscle hypertrophy in transgenic mice. Circulation 112:2668–2676 66. Sorescu D, Weiss D, Lassegue B et al (2002) Superoxide production and expression of nox family proteins in human atherosclerosis. Circulation 105:1429–1143 67. Lee MY, San Marti A, Mehta PK et al (2009) Mechanisms of vascular smooth muscle NADPH oxidase (Nox1) contribution to injury-induced neointimal formation. Arterioscler Thromb Vasc Biol 29:480–487 68. Szöcs K, Lasségue B, Sorescu D et al (2002) Upregulation of Nox-based NAD(P)H oxidases in restenosis after carotid injury. Arterioscler Thromb Vasc Biol 22:21–27 69. Ritsick DR, Lambeth JD (2005) Spring breezes, wheezes and pollen oxidases. J Clin Invest 115:2067–2069

89 70. Kondo S, Shimizu M, Urushihara M et al (2006) Addition of the antioxidant probucol to angiotensin II type I receptor antagonists arrests progressive mesangioproliferative glomerulonephritis in the rat. J Am Soc Nephrol 17:783–794 71. De Mochel NS, Seronello S, Wang SH et al (2010) Hepatocyge NADPH oxidases as an endogenous sourcxe of reactive oxygen during hepatitis C virus infection. Hepatology 52:47–59 72. Hofstetter AR, De La Cruz JA, Cao W et al (2016) NADPH oxidase1 is associated with altered host survival and T cell phenotypes after influenza A virus infection in mice. PLoS One 11:e0149864 73. Lambeth JD, Krause K-H, Clark RA (2008) Nox enzymes as novel targets for drug developmen.t. Semin Immunopathol 30:339–363 74. Jaquet V, Scapozza L, Clark RA, Krause KJ, Lambeth JD (2009) Small molecule nox inhibitors: ROS-generating NADPH oxidases as therapeutic targets. Antioxid Redox Signal 11:2535–2552 75. Diebold BA, Smith SM, Li Y, Lambeth JD (2014) NOX2 as a target for drug development: indications, possible complications, and progress. Antioxid Redox Signal 23:375–405 76. Smith SM, Min J, Ganesh T et al (2012) Ebselen and congeners inhibit NADPH oxidase 2-dependent superoxide generation by interrupting the binding of regulatory subunits. Chem Biol 19: 752–763 77. Li Y, Ganesh T, Diebold BA, Zhu Y et al (2015) Thioxodihydroquinazolin-one compounds as novel inhibitors of myeloperoxidase ACS. ACS Med Chem Lett 6:1047–1052 78. De La Cruz JA, Ganesh T, Diebold BA et al (2021) Quinazolinderived myeloperoxidase inhibitor suppresses influenza A virusinduced reactive oxygen species, pro-inflammatory mediators and improves cell survival. PLoS One 16:e0254632

4

The Discovery and Characterisation of Nox2, a Personal Journey Anthony W. Segal

Abstract

This chapter describes the discovery of the nature of the neutrophil NADPH oxidase and the causes of Chronic Granulomatous disease (CGD), the purification and characterisation of the NOX2 flavocytochrome, the identification of p22phox and accessory activating molecules. It also describes the mechanisms by which the oxidase promotes microbial killing within the phagocytic vacuole by activating neutral proteases through an elevation of pH and a potassium flux. A general mechanism of action is proposed in which NOXs act as electrochemical generators of electromotive force to produce ion fluxes across membranes and alter pH and tonicity. Keywords

Neutrophil · NOX · NADPH · Oxidase · Cytochrome · Flavin

1

Introduction

I have decided that rather than simply writing a conventional scientific manuscript, I would use this opportunity to describe the personal journey that I travelled, both literally and metaphorically, in my discovery of NOX2, and the subsequent investigations of this electron transport chain. My involvement with the neutrophil started in the early 1970s when, as a junior doctor, I started to perform the nitroblue tetrazolium (NBT) test [1] on patients. This test had been promoted in The Lancet as one that could distinguish between bacterial infection and other causes of inflammation such as viral infections and immunological and rheumatological diseases. For a clinician such a test would be invaluable, particularly because it is often difficult to A. W. Segal (✉) University College London, London, UK e-mail: [email protected]

differentiate such conditions, and because bacterial infections generally respond to antibiotic therapy. I was particularly interested in this test because of my experiences during 6 months spent working at a cardiothoracic hospital in Zululand in 1969. We had an outbreak of endotoxin shock in our patients because the heart-lung machine was not being properly cleaned. It was very difficult to determine the cause of shock in these postoperative open-heart surgery patients, and to distinguish, for example, failure of an implanted valve from infection. The basis of the NBT test was that the dye, nitroblue tetrazolium, was added to, and incubated with, anticoagulated blood, and then a slide was made and counter stained. Upon inspection it was found that neutrophils had blue deposits associated with them and, in patients with pyogenic infections, the proportion of neutrophils containing blue deposits was greater, with an arbitrary cut off of about 10%. This test was gaining considerable popularity and the numbers of papers extoling its virtues proliferated. I started to personally perform the test and encouraged my clinical colleagues to use it. To my dismay I found the results in my hands to be most disappointing. We then conducted a proper clinical trial and demonstrated the test to be inaccurate and of only very limited clinical value [2]. I then went on to work out the underlying mechanisms of the test [3]. When added to blood, the NBT dye precipitated fibrinogen and heparin out of solution to form small particles which were then phagocytosed by neutrophils. NADPH oxidase produced by the neutrophils reduced the dye from yellow to dark blue insoluble formazan that was visible on the slides. The uptake of the dye-enriched precipitates was enhanced by acute phase proteins acting as opsonins [4]. Because these acute phase proteins become elevated in a variety of pathological conditions, the test could never be specific for only one group of diseases. Following our publications, the NBT test in this form was abandoned for the diagnosis of pyogenic infections.

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_4

91

92

2

A. W. Segal

Discovery of the Nature of the NADPH Oxidase

My interest in the NADPH oxidase followed from my work on the NBT test because I was interested in biochemical mechanisms and wanted to understand how the neutrophils reduced the dye. It had been reported that neutrophils from patients with Chronic Granulomatous Disease (CGD) [5] failed to demonstrate a respiratory burst upon phagocytosis [6], and were also incapable of reducing NBT to formazan [7]. So, the reductase was clearly defective in this condition, which provided a window through which to find the system responsible. The first studies we undertook were the examination of NADH and NADPH dependent reduction of NBT by subcellular fractions of neutrophils separated on isopycnic sucrose gradients. We erroneously thought that NADH was the preferred substrate but observed that the majority of the reductase activity was inhibited by superoxide dismutase (SOD), that it distributed predominantly with the plasma membrane, and that this activity was absent from the plasma membrane of four patients with CGD [8, 9]. I then set out to attempt to purify the enzyme that had by now been established to be a NADPH oxidase. I used buffy coat residues from human blood, purified and stimulated the neutrophils which I then fractionated, and I attempted purification of the oxidase from the solubilised active fractions by a variety of methods. After 18 months of solid work, I was getting nowhere and became convinced that we were not dealing with a simple enzyme. Rather than gaining activity, the further along the purification process I proceeded, the more activity seemed to be lost. At this time, I was undertaking an evening Master’s course in Biochemistry at Chelsea College of Science and Technology. I discussed the problem with one of the course professors, Harold Baum, who suggested that I should consult Brian Chappell, an expert on electron transport at Bristol University. I went down to Bristol to discuss the matter with him, and he postulated that it could involve a b type cytochrome. He suggested that I should confer with his colleague Owen Jones who was away that day. I asked Brian how one demonstrated such cytochromes and was told that I should run a reduced minus oxidised difference spectrum. I ran a rather rudimentary such spectrum on neutrophils with my spectrophotometer in London and faxed the result to Owen. He said that it looked like haemoglobin, but that I should bring some neutrophils down to Bristol University to be run on his advanced, custom-made, spectrophotometer. His statement on obtaining the first spectrum was that “it looks interesting” (which for an Englishman meant that he was quite excited!), and this started off a long and rewarding collaboration. Every week or two I would go to my laboratory in

London, prepare cells or fractions thereof, and then take the train for the 2 h journey to Bristol, where I used to walk up the hill to the Biochemistry Department. Invariably I would arrive just before lunch whereupon the whole group decanted to the nearest pub for beers and sandwiches. I was itching for the results of the experiments but had to wait patiently until the lunch hour was over and the experiments could commence. When these had been completed, I would walk down to the station and spend the time on the train writing up the results or the paper. Almost all the experiments were initially performed by Owen and me, and we were subsequently joined by Andrew (Sandy) Cross. We discovered that neutrophils contained a very low potential cytochrome b which was located primarily in the plasma membrane and specific granules and translocated into the phagocytic vacuoles [10]. Several years after these early discoveries I was reading a book called “the Cytochromes” (editor unknown) when I saw reference made to the discovery of a cytochrome b spectrum in animal neutrophils, published in some esoteric Japanese journals [11, 12]. Unbeknown to us this cytochrome b had been seen more than a decade earlier by Japanese researchers in horse and rabbit neutrophils who considered it to account for the NADI oxidase in these cells, later shown to be caused by cytochrome oxidase [13],—no link had been made to the neutrophil NADPH oxidase.

3

NOX2 and Chronic Granulomatous Disease

In the 1970s, the nature of the NADPH oxidase was the focus of attention of many laboratories internationally. The difficulties of establishing its identity resulted in its being likened to the Scarlet Pimpernel, a fictional hero that rescued individuals from the revolutionaries during the French revolution, and who the authorities were unable to apprehend: “They seek him here, they seek him there, those Frenchies seek him everywhere, is he in heaven or is he in hell? That damned elusive Pimpernel” Baroness Emmuska Orczy, 1905 in “The Scarlet Pimpertnel”

We considered the cytochrome b that we had identified in human neutrophils to be a strong candidate for the NADPH oxidase and were convinced of this fact when we found it to be missing from the cells of patients with CGD [14]. The cytochrome b was also present in other myeloid cells that demonstrated a respiratory burst [15] and developed in parallel with the acquisition of oxidase activity in maturing HL-60 cells [16]. A human promyelocytic leukaemia cell line that could be induced to mature in culture with dimethyl sulphoxide.

4

The Discovery and Characterisation of Nox2, a Personal Journey

93

I was thus very surprised to receive a telephone call from Niels Borregaard, who was then a PhD student at the university of Aarhus, to tell me that he had identified some patients with CGD in whom the spectrum of the cytochrome b was normal. My first reaction was that he was measuring contaminating haemoglobin, but when his spectra were shown to be accurate, I flew to Aarhus to meet him. Clearly, he had discovered some patients, two with an autosomal recessive inheritance and another male who was X-linked, in whose neutrophils the cytochrome b was unequivocally present [17]. They concluded that “These results indicate that the postulated importance of cytochrome b in the oxygen burst during phagocytosis is questionable”. I resolved this matter by showing that although neutrophils from patients with an autosomal inheritance did contain normal amounts of the cytochrome, it was non-functional [18]. This was demonstrated by activating neutrophils under anaerobic conditions. Normally, when the oxidase is activated, electrons are passed onto the cytochrome, which becomes reduced, and it then passes its electrons onto oxygen. In the absence of oxygen, it remains in the reduced state and can be clearly seen in reduced minus oxidised spectra, when compared with the oxidised cytochrome in unstimulated cells. In contrast with normal cells, neutrophils from CGD patients with an autosomal recessive inheritance were completely unable to pass electrons onto the cytochrome in their cells. This finding indicated that the electron transport system in these patients was missing an activating mechanism, or a proximal electron carrier. The patient described by Borregaard with X-linked CGD whose cells contained a seemingly normal cytochrome [19], was the first case of “Variant CGD” documented. Despite normal amounts of the cytochrome b, which demonstrated a normal mid-point potential and carbon monoxide (CO) binding, the respiratory burst was all but absent in the patient and about half normal in his carrier mother and sister, confirming an X-linked inheritance. It is likely that this patient (J.L. in [20]) had a mutation in a regulatory region of the cytochrome that prevented an activation event, such as phosphorylation, or the attachment of an activating molecule. The central role of the cytochrome in the oxidase was established beyond question by two further studies. The first was a multicentre European evaluation [21] in which we measured this cytochrome in granulocytes from 27 patients with CGD and from 64 members of their families. It was undetectable in all 19 of the men in whom the defect appeared to be located on the X chromosome. Female relatives who were heterozygous carriers had reduced concentrations of the cytochrome. In contrast, in all eight patients (seven women) with a probable autosomal recessive inheritance, the cytochrome was present but non-functional. The properties tested, including midpoint potential, CO binding, and organelle distribution, were normal. However, in all

cases with an autosomal recessive pattern of inheritance the cytochrome failed to become reduced when anaerobic cells were stimulated. The second confirmatory study was a molecular genetic investigation conducted by Stuart Orkin [22]. He and colleagues had cloned the gene for X-linked CGD by relying on its chromosomal position. The transcript of the gene was found to be expressed in the phagocytic lineage of haematopoietic cells and was absent, or structurally abnormal, in four patients with X-CGD. The nucleotide sequence of complementary DNA clones predicted a polypeptide of at least 468 amino acids with no homology to proteins described previously. In fact, as will be described later, we had purified the cytochrome b and obtained the sequence of 43 amino acids from the N terminus, and this sequence aligned perfectly with that predicted from the DNA sequence of the cloned gene after we established the correct reading frame [23].

4

Characterisation and Purification of the Cytochrome b

The next task was to determine the redox characteristics of the cytochrome, work that was largely done with Andrew (Sandy) Cross, a lecturer in the Biochemistry Department in Bristol University. One of the most important properties of a redox molecule is its midpoint potential, at which it is balanced at the midpoint between oxidation and reduction, because that determines where it will function in the hierarchy of electron transporting molecules. We found the midpoint potential to be -245 mV, substantially lower than that of any other b-type cytochrome described in mammalian cells [24]. This will be discussed in further detail later. This low midpoint potential was important because it allowed for the passage of electrons onto oxygen to produce superoxide (O2-) which has a more positive midpoint potential of -160 mV. Because haem proteins that react with oxygen are likely to bind CO, we tested this and observed that reduced cytochrome formed a complex with CO, and that reduced cytochrome was rapidly reoxidised by oxygen [24] (Fig. 4.1).

5

The Search for Other Electron Transport Carriers

5.1

Flavin and NADPH Binding Components

At the time that we attempted to identify the nature of these components, considerable confusion existed as to the nature of the electron transporting molecules. It was highly likely

94

A. W. Segal

O2-

H2O2

O2 Outer Fe Fe

Inner

e-

FAD e-

NADPH

NADP+ H+

Fig. 4.1 Schematic representation of NOX2. The NADPH and FAD binding sites and the location of the two haems are shown. FAD accepts a single electron from NADPH and it then passes it sequentially to the two haems before it is accepted by oxygen to produce superoxide

that a flavin would be involved in a pathway between NADPH and a cytochrome b, a pathway linking a two-electron donor (NADPH) to a one-electron acceptor (a haem in the cytochrome). The NADPH oxidase had been considered by Babior to be a flavoprotein because the O2forming activity of zymosan stimulated neutrophils was lost in the presence of detergent but could be stabilised by the addition of flavin adenine dinucleotide (FAD) [25]. A similar conclusion was arrived at by Light and colleagues who found a requirement for FAD in detergent solubilised O2- forming systems from neutrophils and observed their sensitivity to inhibition by FAD analogues [26]. In addition, diphenylene iodonium (DPI), a potent inhibitor of the oxidase, was thought to act at the site of a flavin [27]. Most convincing, Kakinuma had demonstrated a flavin semiquinone by electron paramagnetic resonance when NADPH was added to membranes from activated but not resting neutrophil membranes [28]. There was some evidence to suggest the presence of the flavin in combination with a cytochrome in a flavocytochrome. Some partially purified preparations of the oxidase contained both FAD and haem [29] which were reduced by NADPH [30], but other preparations contained FAD but no haem [31], and yet others contained neither FAD nor haem [32]! In addition, other cytochrome reductases were present in neutrophils that were shown to be capable of reducing the cytochrome b-245 [32].

5.2

FAD

We first measured the FAD and cytochrome concentrations in resting and activated membranes and found that there was little change upon activation and that the ratio of flavin to cytochrome remained the same, indicating that the FAD binding protein was not recruited from the cytosol [33]. We examined the possibility that FAD might be intimately linked to the cytochrome. The absorption spectra of flavins is much more diffuse and less well defined than that of cytochromes, making spectroscopic identification and quantitation more difficult. However, we did find that the plasma membranes of neutrophils had a fluorescence spectrum resembling that of a flavoprotein, with a roughly equimolar concentration of FAD and cytochrome b, both of which had a similar subcellular distribution in the specific granules and plasma membrane. Deficiencies of FAD had been described in neutrophils from X-CGD patients [34, 35]. We repeated the measurements of FAD in the membranes of patients with X-linked and autosomal recessive CGD, using a sensitive D-amino acid oxidase assay, and found that the levels in the X-linked patients were about 30% of those in the normal subjects and patients with the autosomal recessive CGD. Most importantly, we found that plasma membrane fractions prepared from neutrophils from CGD patients with X-linked CGD lacking the cytochrome had about one half of the normal amount of FAD whereas those from patients with autosomal recessive CGD had normal amounts, suggesting an association between the haem and FAD, possibly within a flavocytochrome [36], with a stoichiometry of 1:2. In addition, highly enriched preparations of the cytochrome were also highly enriched in FAD. Once the amino acid sequence of the cytochrome had been determined, a region of strong homology was found that corresponded with the FAD binding pocket of the family of ferredoxin-NADP+ reductases (FNR) [33].

5.3

Other Electron Transporting Molecules

To determine whether redox molecules other than the flavocytochrome b might be involved in the oxidase we enlisted the help of colleagues. Trevor Griffiths (Biochemistry Department, Bristol University) was unable to find ubiquinone in extracts of neutrophil membranes, and Mike Evans (Botany, UCL) did not detect the signal for ironsulphur proteins (unpublished). Several groups had suggested that a quinone, later described as ubiquinone-50, was involved in the superoxide

4

The Discovery and Characterisation of Nox2, a Personal Journey

95

generating system [37]. We undertook an analysis of the subcellular distribution of quinones in neutrophils by subcellular fractionation of these cells and measurements of ubiquinones by spectroscopy and high-performance liquid chromatography (HPLC) and found that almost all the quinones were associated with the mitochondrial fraction, and that little, if any, partitioned into phagocytic vacuoles [38]. We were left with the conclusion that the only redox transporting molecules were the FAD and the haem of the flavocytochrome b.

of appropriate protease inhibitors, partial fractionation of the cells, extraction of the preparation with detergents that extracted many proteins, but not the cytochrome, followed by detergents that did solubilise the cytochrome. The extract was then passed through an affinity column to which many proteins but not the cytochrome bound, followed by a column of heparin-agarose to which it did bind as a consequence of its heavy glycosylation [41]. The purified material was enriched about 200-fold and gave a single broad band with an apparent molecular weight of 68–78 kDa on SDS gels. It did not contain FAD or flavin mononucleotide (FMN) and had no independent NADPH oxidase activity. The purification of this cytochrome allowed us to obtain the sequence of the 43 amino acids from the N-terminus which permitted the definitive identification of the product of the gene that had been cloned by “reverse genetics” and defined as that coding for the oxidase that was defective in X-linked CGD [23]. We found almost complete homology between the sequence that we obtained and that of the complementary nucleotides 19–147 of the sequence of the X-CGD gene originally designated as non-coding because of an incorrect assignation of the ATG start codon.

5.4

The Location of the NADPH Binding Site

There was considerable confusion as to the identity of the NADPH binding molecule. Attempts were made to identify the NADPH-binding protein by affinity labelling with NADPH dialdehyde. A 66 kDa membrane bound protein was identified [39], but that was later thought to be cytosolic [40] and of a lower molecular weight [40]. I was keen to undertake an NADPH binding study and discovered that Ashok Chavan and Boyd Haley, at the Markey Cancer Centre in Lexington, Kentucky, were conducting photoaffinity labelling experiments with 2-Azido-NADP+ labelled with 32P. I called them, packed my bags, and went to see and work with them. The travel was certainly worthwhile. We showed that a band at about 93 kDa was radiolabelled in neutrophil membranes from normal subjects, but not in membranes from four patients with X-linked CGD [33]. We also conducted sequence homology searches against libraries of nucleotide binding sites. These were generally unproductive until we searched specifically with the conserved glycine-rich region in the proposed NADPHbinding site of the ferredoxin-NADP+ reductase flavoenzyme family, with which close homology was discovered [33]. We were thus confident that the NADPH bound directly to the flavocytochrome b located in the membrane.

6

Purification and Characterisation of the Flavocytochrome b

The next task was to purify the cytochrome. In this day and age of molecular biology and DNA sequencing it is hard to imagine how difficult it was in those days to purify and obtain amino acid sequence of proteins, in particular of hydrophobic membrane proteins. It took us roughly 3 years to achieve our objective [41]. We started with human neutrophils removed by leukapheresis from patients with chronic myeloid leukaemia. The keys steps in the subsequent purification involved the use

7

Identification of the Second Flavocytochrome (23 kDa) Subunit

During the initial purification steps, we were aware of the variable presence of a 23 kDa protein in our preparations. As we became more expert and rapid in our purification, we found that this smaller subunit co-purified with the haem containing glycoprotein through a wide variety of purification steps. Proof of the association of these two proteins was once again obtained by recourse to the examination of neutrophils from patients with CGD. By refining our techniques, we were able to purify the cytochrome from as little as 108 cells, which could be obtained from samples of peripheral blood. I demonstrated that both the 76–92 kDa and the 23 kDa subunits were present in roughly equimolar concentrations in normal subjects, and patients with autosomal recessive CGD with a normal cytochrome spectrum in their neutrophils, whereas both proteins were missing from X-CGD neutrophils. These results were confirmed by Western blotting using an antibody we had raised to the 23 kDa protein [42]. Similar results were reported at about the same time by the Parkos group [43]. Subsequently, using a variety of complementary biochemical techniques, we confirmed the stoichiometry of the two subunits to be 1:1 [44]. We thus concluded that the cytochrome was comprised of two subunits, present in equimolar concentrations and in accordance with conventional terminology we called the

96

A. W. Segal

23 and 76–92 kDa proteins the α and β subunits respectively. These have subsequently been designated p22 phox and gp91phox and this notation will be used henceforth.

With Willie Taylor and David Jones at the National Institute for Medical Research at Mill Hill and UCL, we constructed a model of the flavocytochrome [45]. The FAD and NADPH binding sites had been located on this molecule, the C-terminal half of which showed weak sequence similarity to other reductases, including the FNR, of known structure [46]. This enabled us to build a model of the nucleotide binding domains of the flavocytochrome using this structure as a template. The resulting model rationalised much of the observed sequence conservation and identified a large insertion as a potential regulatory domain. It confirmed the inclusion of the neutrophil flavocytochrome as a member of the FNR family of reductases. To add to this model, we located the glycosylation sites which would, of necessity, be on the extracellular face of the molecule. It is N-glycosylated, and site-directed mutagenesis was used to eliminate the five potential N-linked glycosylation consensus sites. Mutated cDNAs were expressed in vitro and provided evidence for glycosylation of residues Asn131, Asn148 and Asn239, but not of Asn96 and Asn429 [47].

This insight was important because it explained the conundrum of the previously observed ratio of FAD to haem of 1:2 in the flavocytochrome [29]. Flavins transfer a single electron, so how would this pass onto two haems of equal potential? The answer supplied by this study is that the electron is transferred to the haems sequentially in series rather than in parallel. In order to identify the haem binding histidines in gp91phox we made use of the strong homology between the flavocytochrome and yeast iron reductase, both of which have very low mid-point potentials [49]. In collaboration with the Dancis group at the National Institutes of Health in Bethesda, USA, we identified the haem binding histidines in the flavocytochrome by determining the haem binding sites in the FRE1 (yeast iron) reductase of Saccharomyces cerevisiae [50]. We found two sets of conserved histidines in alpha helical hydrophobic domains. Both pairs were 13 amino acids apart in FRE1 and FRE2, gp91 phox and several other cytochromes, and the second pair were also separated by roughly the same number of amino acids. Mutation of these histidines abolished haem binding but not protein expression. The coordination of these haems between the two pairs of histidines would locate them one above the other perpendicular to the plane of the membrane and located at opposite sides of the lipid bilayer, ideally positioned to pass electrons across the membrane. The possible relevance of the strong homology between the flavocytochrome of the oxidase and the iron reductases of yeast will be discussed later.

8.2

9

The Discovery of Accessory Activating Molecules

9.1

p47 phox and p67 phox

8

Further Studies into the Structure of gp91phox

8.1

Building a Model

Location of the Haems

Once again, a patient with CGD provided evidence that profoundly influenced our understanding of the oxidase. Sandy Cross and John Curnutte discovered a patient with X-linked CGD whose neutrophils contained a non-functional cytochrome. They detected a very minor shift in the difference spectrum of the cytochrome, and on sequencing the DNA there was a single base substitution of a cytosine for guanine 173 in exon 2 of gp91phox (AGG → ACG) leading to a prediction of the nonconservative replacement of arginine 54 with a serine residue. They then went on to analyse whether this mutation affected the mid-point potential of the cytochrome. In their patient they found the presence of two nonidentical haemes, with midpoint potentials of Em7 = -220 mV and Em7 = -300 mV. Reanalysis of redox titrations of wild-type cytochrome b-245 revealed the presence of two haemes, with closely spaced midpoint potentials of Em7 = -225 mV and Em7 = -265 mV, which merged to give the previously described, incorrect, Em7 of 245 mV [48].

The demonstration that neutrophils from patients with autosomal recessive CGD were unable to transfer electrons onto the cytochrome indicated that either there were additional components required for electron transport that were missing in these patients, or that an activation mechanism was defective. Phosphorylation was gaining prominence as an activating mechanism for many cellular processes, and accordingly we examined phosphorylation of neutrophil proteins after activation of the cells with phorbol myristate acetate (PMA). In all four of the autosomal patients studied there was a selective lack of the enhanced phosphorylation of a 44 kDa protein in contrast to the normal phosphorylation that was observed in normal subjects and in two CGD patients with an X-linked inheritance. We concluded that this molecule could be an important functional component of the oxidase [51].

4

The Discovery and Characterisation of Nox2, a Personal Journey

We were lucky in this endeavour because it was subsequently shown that of the seven proteins associated with the oxidase, most autosomal recessive patients have a defect in this protein, now called p47 phox. We showed that the relative frequency of this defect can be accounted for by a dinucleotide deletion at a GTGT tandem repeat, corresponding to the acceptor site of the first intron-exon junction, and that slippage of the DNA duplex at this site may contribute to the high frequency of defects in this gene [52]. At about the same time Bob Clark, Bill Nauseef and Bryan Volpp took the approach of making antibodies to neutrophil cytosolic proteins that bound to a guanosine 5′-triphosphate– agarose affinity column. They produced antibodies to two predominant proteins of 47 and 67 kDa that were restricted to the cytosol fraction of neutrophils and related myeloid cells. The 47 kDa protein was missing in most patients with autosomal recessive CGD, with the 67 kDa being absent in the others they studied [53]. We purified the phosphoprotein and made antibodies to it, and these antibodies recognised the protein in X-linked but not AR-CGD. We obtained the amino acid sequence for this protein which confirmed it as being the same protein as the 47 kDa protein identified by the Clark group [54]. We showed that activation of the neutrophil oxidase system appeared to be dependent upon phosphorylation of the cytosolic 47 kDa protein and its association with cytochrome b-245 in the membranes. We proposed it as the cytosolic factor required for reconstitution of the active oxidase in cellfree systems [55].

9.2

p21Rac and GDI

Pioneering work by Edgar Pick had demonstrated that in addition to the flavocytochrome b, at least three cytosolic factors (p67 phox and p47 phox, and a third component, sigma 1) were required for oxidase activity in a cell-free system [56]. I contacted Edgar, and his graduate student Arie Abo at Tel Aviv University in Israel and asked them if we could collaborate to attempt to identify sigma 1. Purified sigma 1 contained two proteins from which we obtained amino acid sequence, identifying them as the small GTP-binding protein p21rac1 and the GDP-dissociation inhibitor rhoGDI [57]. Sigma 1 factor could be replaced by p21rac in the presence of added GTP. We determined the structure of rhoGDI [58] and demonstrated that it had a pocket that housed the C-terminal isoprenyl group of rac, keeping it inactive in solution in the cytosol. We showed that upon dissociation, p21rac translocates to the membrane [59] where it binds to the cytochrome complex through an attachment to p67 phox [60].

97

9.3

p40 phox

I set out with Frans Wientjes, in my laboratory, to identify additional proteins associated with the cytosolic oxidase factors. We immunoprecipitated p67 phox from neutrophil cytosol and discovered that it had two other proteins complexed to it, p47 phox and a protein with an apparent molecular mass of about 40 kDa, subsequently referred to as p40 phox [61]. p40 phox and p67 phox seemed to be present in the complex in roughly equimolar quantities. When cytosol was fractionated on a gel filtration column, we found that p67 phox , p47 phox and p40 phox separated in a complex with an apparent molecular weight of about 250 kDa. The primary association appeared to be between p67 phox and p40 phox because the latter was grossly diminished in cytosol from autosomal recessive-CGD cells lacking p67 phox but not those lacking p47 phox. We obtained amino acid sequence from p40 phox and then cloned and sequenced the gene which revealed a previously undescribed protein. Interestingly, the greatest similarities were with p47 phox and p67 phox [61]. That with p47 phox extended from the N-termini of both proteins over a large part of their amino acid sequence, with 22% identity over 245 amino acids. p40 phox also contained a region having similarity with SH3 domains of several proteins and the strongest of these similarities was with the C-terminal SH3 domain of p67 phox. Northern blot analysis of cell lines and tissues to examine the pattern of expression of p40 phox indicated that, like p47 phox and p67 phox, its tissue expression is predominantly in bone marrow and neutrophils. p40 phox is required for activity of the oxidase when this is activated by a phagocytic stimulus, but not after stimulation with PMA. Inherited p40 phox deficiency underlies a distinctive condition, resembling a mild, atypical form of CGD. The patients suffer from hyperinflammation and peripheral infections, but they do not have any of the invasive bacterial or fungal infections seen in CGD [62].

10

The Function of NOX2 and the Probable Function of Other NOXs

My other major interest at the time was to understand the way in which the oxidase promoted microbial killing. In trying to determine whether the reactive oxygen produced by neutrophils might become incorporated into engulfed bacteria I conducted a little reported study employing 15oxygen [63]. The demanding nature of this experiment was that the half-life of 15O2 is just over 2 min. I determined that the isotope did not exchange with the incubation medium or cells to an appreciable extent and unmetabolised oxygen was readily eluted by gassing the

98

A. W. Segal

cell suspension. The polarographic measurements of oxygen consumption closely paralleled the recovery of metabolised 15 O2. Almost all the metabolized 15O2 was converted into water, both in the presence and absence of KCN (added to inhibit mitochondrial metabolism), supporting the concept that the oxygen consumed by neutrophils is converted into H2O2. It was unlikely that an appreciable proportion of this oxygen is incorporated into the organic composition of the cell or of the ingested micro-organism.

10.1

Oxygen Free Radicals and the Role of Myeloperoxidase

The description that the oxidase generated superoxide [64] produced enormous excitement. It led to the idea that oxygen free radicals, which were considered to be very toxic, were responsible for directly killing microbes, as well as causing tissue damage in inflammatory diseases in which neutrophils were found to predominate. Suffice it to say that this is clearly not the case as these reduced oxygen species are not microbicidal under the conditions pertaining within the phagocytic vacuole [65]. The pros and cons of whether myeloperoxidase has a microbicidal role through the production of hypochlorous acid has been debated by the proponents of this theory [66], and by us as opponents [67]. The rival arguments have been clearly set out in these two papers. Time and future work will determine which of these views is correct.

10.2

The Influence of NOX2 on Conditions Within the Phagocytic Vacuole

At the time it was believed that the pH in the phagocytic vacuole was very acidic, at about pH 4–5, that this low pH was important for bacterial killing, and that the enzymes released into the vacuole were similar to lysosomal enzymes with acid pH optima [68]. I had been fascinated by Mitchell’s chemiosmotic theory in which protons were pumped across the inner mitochondrial membrane as electrons went through the mitochondrial electron transport chain. I surmised that the oxidase might acidify the vacuole through a similar mechanism. One way to establish this would be to see whether there was a difference in vacuolar pH between healthy and CGD neutrophils. I approached Mike Geisow at the National Institute for Medical Research, at Mill Hill, who was investigating the pH in macrophage vacuoles using the fluorescent spectrum of fluorescein as an indicator. We coupled fluorescein to opsonised Staphylococcus aureus and measured the fluorescence spectrum after phagocytosis by neutrophils.

To our surprise we obtained completely the opposite results to those predicted. The pH in normal neutrophils was elevated from that of the extracellular medium at about 7.4 to 7.8–8.0, whereas the pH in CGD neutrophils rapidly fell to about 6.0, so that there was a ~2.0 pH difference between that in the normal and CGD vacuole [69]. Subsequently we have used SNARF, a much better indicator, with a more intense signal and wider dynamic range, to measure changes in pH in the phagocytic vacuole and cytosol of neutrophils. In human cells, the vacuolar pH rose to ~9, and the cytosol acidified slightly [70, 71]. The oxidase is electrogenic. This means that the passage of electrons across the vacuolar membrane causes it to depolarise, a change that opposes the further transfer of electrons. For electron transport to continue, the charge across the membrane must be compensated, either by the movement of negatively charged ions such as Cl- from the vacuole into the cytoplasm, or positively charged ions such as H+ or K+ into the vacuole, or a mixture of these fluxes (Fig. 4.2). We went on to investigate the effect of pH upon microbial killing and made some interesting observations. We made knock-out mice lacking cathepsin G and/or neutrophil elastase [72, 73] and discovered that these enzymes were essential to kill bacteria and fungi, and that their absence had different effects on different organisms. Cathepsin G is required to kill S. aureus whereas elastase is important in the killing of C. albicans. These knock-out mice had normal oxygen radical production and normal complements of myeloperoxidase, and yet killing was defective, further strong evidence against an important microbicidal action of these systems. So why are these organisms not killed efficiently in CGD vacuoles into which the cytoplasmic granules release their enzymes normally [74]? The reason is that both these enzymes are neutral proteases with pH optima of about 8.5–9.5, conditions produced by the respiratory burst in normal vacuoles, but they are relatively inactive at the acid pHs pertaining in CGD vacuoles. This failure to kill and digest bacteria in CGD results in bowel inflammation indistinguishable from Crohn’s disease, another area of research in which I elucidated the underlying mechanisms of the disease [75]. How then does the oxidase achieve these optimal pH conditions within the vacuole? The pH of the cytoplasmic granules is maintained at about 5.5 [76], at which the contained enzymes are inactive. Upon phagocytosis these acidic materials are released into the vacuole and at the same time the oxidase is activated, transporting electrons into the vacuole where they attach to oxygen forming superoxide. Superoxide dismutates to form peroxide which becomes protonated to form hydrogen peroxide, consuming protons and elevating the pH. We showed that the optimal pH

4

The Discovery and Characterisation of Nox2, a Personal Journey

99

Fig. 4.2 NOX2 generates an electrochemical gradient that provides the energy to power ion transport. The energy incorporated in NADPH, the substrate for the oxidase, is originally produced by photosynthesis. The NADPH oxidase is electrogenic. The passage of electrons across the membrane depolarises the membrane and electron transport would cease

unless this charge is dissipated by charge compensation. The passage of negatively charged electrons is compensated by the passage of positively charged cations in the same direction as the electrons, or of anions in the opposite direction

is achieved by balancing charge compensation by protons, through the HCVN1 proton channel, and potassium ions. In knock out mice lacking HCVN1, the pH in neutrophil vacuoles rises to about 11! [70]. In addition to optimising the vacuolar pH, the potassium flux has another role [73]. The positively charged cytoplasmic cationic granule proteins are not simply in solution within the granules but are tightly attached to the granule matrix where they are bound to the negatively charged sulphated proteoglycans, in much the way that they would bind to an anion exchange column. The K+ ions compete with the charges on the matrix, releasing the enzymes to undertake their digestive duties. The influx of K+ also has an osmotic effect, attracting water and inducing swelling of the vacuoles. Anecdotally, one of the reviewers of our manuscript submitted to Nature required us to demonstrate that the vacuole membrane was permeable to water. I found that a possible way to demonstrate this was by using Stokes Raman scattering laserscanning microscopy to show the replacement of water by deuterium. The only practical person that I could find to do this was Eric Potma at Groningen University, so I called him and proposed a collaboration. He said that he was unable to do so because he was taking up a postdoc position in the USA in a couple of weeks and had packed his laser away in preparation. I asked him to reconsider after reading the draft of the paper, which he did, and a day or two later I was on my way to Groningen to do the experiment, which fortunately led to a happy outcome.

10.3

The K+ Channel

Our next task was to identify the channel through which the K+ was entering the vacuole. This search led to one of the most shocking experiences of my life. I employed Jatinder Ahluwalia to do this work. He was a postdoc who had obtained his PhD from Imperial College. I first checked with his PhD supervisor, and with his Head of Department at Imperial, to ensure that he was reliable and competent, and both assured me that he was. Our plan to identify the channel was to use a series of specific potassium channel blockers with vacuolar pH as the readout, on the assumption that if the influx of K+ into the vacuole was prevented, these ions would be replaced by H+ and the pH would fall. After obtaining many negative results Ahluwalia showed me results that indicated that the predicted changes were obtained in the presence of iberiotoxin and paxillin, selective inhibitors of the HVCN1 potassium channel. We submitted these findings to Nature who required us to demonstrate the channel electrophysiologically. I approached independent colleagues at UCL, with international reputations as electrophysiologists, who agreed to collaborate with us on this project. Ahluwalia provided them with the cells in which they demonstrated typical HVCN1 channels, and the paper was duly published [77]. About 1–2 years later I received reports that others were having difficulty reproducing our results. I spent the next 2 years attempting to understand the discrepancy between our results and those of others. Eventually we discovered that

100

A. W. Segal

Ahluwalia had been fraudulent. An investigation was conducted by UCL and the findings are available at: https://www.ucl.ac.uk/governance-compliance/sites/gov ernance_compliance/files/research-misconduct-panel.pdf We then retracted the paper from Nature [78]. It transpired that whilst a graduate student at Cambridge university Ahluwalia had been expelled for fraud, and that the major publication of his PhD studies at Imperial College had to be retracted [79], despite which Imperial refused to rescind his PhD. https://ktwop.com/2013/07/31/ahluwalias-phd-clearedof-fraud-by-imperial-college/ The most important requirement is that the scientific literature should be as accurate as possible, and I was pleased that we could correct the inaccuracy that we had placed in it. I felt particularly responsible and very sorry for the adverse impact that these matters had on other investigators, but one cannot predict such happenings and can only address them as they emerge. One of my main concerns was that this misplaced identification of the potassium channel would distract from the importance of the potassium flux to the function of the oxidase. I therefore made a concerted effort to identify the correct potassium channel. We approached this matter by looking at the expression patterns of mRNA in myeloid cells and concentrated on the most highly expressed potassium channels. Rather than relying on inhibitors of uncertain specificity we examined cells from knock-out mice, and from humans with defined genetic lesions. We tested 18 different knock-out mice lines and 4 human conditions with different potassium chanellopathies [80]. Although we did observe some minor influences on vacuolar pH and volume, we did not find that the loss of any single channel in the cells we examined caused the expected changes, either because we had failed to identify the correct channel, or because of redundancy of the K+ or cation channels.

11

NOXs as Electrochemical Generators of Ion Fluxes

I would like to address here what I consider to be the central role of the NOXs. It is generally understood that NOXs exist to produce oxygen free radicals, and that these have an important role in cell signalling. I have yet to find convincing data to support this concept. I think that this idea can be likened to a belief that the function of internal combustion engines is to produce the exhaust fumes that they emit! One of the most important requirements for life by a cellular organism is the regulation of its intracellular ionic composition and pH. This requires an electromotive force and the appropriate ion channels. The electrogenic passage of

electrons across a membrane produces the energy that can drive these ion fluxes. Rather than considering NOXs as generators of oxygen free radicals, I consider them as the simplest possible electrogenic mechanism for passing electrons across membranes, and therefore powering the passage of other ions. There is no simpler mechanism to transport electrons across a membrane than a single FAD molecule coupled to two haems, one at either side of the membrane. Given their strong homology to the yeast ferric reductases, it is possible that the NOXs evolved from such molecules where the electron acceptor was iron rather than oxygen. Such molecules might have powered ion transport across the membranes of primitive organisms even before the accumulation of oxygen on our planet. What is the evidence for the role of NOXs in the passage of ions across membrane?

11.1

NOX3 and the Head-Tilt Mouse

Defective function of NOX3 in mice results in a vestibular defect as a consequence of which the mouse lacks spatial awareness [81]. The reason for that is that the otoconia, calcified proteinaceous bodies that couple mechanic forces to the sensory hair cells in the utricle and saccule of the vestibular system, fail to become calcified in these mice. We have shown that NOX2 is required to alkalinise the phagocytic vacuole. Otoconia are normally calcified by calcium carbonate, the solubility of which is very pH dependent, requiring an alkaline environment for calcification and being dissolved by acidity, so it is a strong possibility that the defective calcification of otoconia in NOX3 deficient mice is caused by the failure of this molecule to alkalinise the endolymph.

11.2

NOXs in the Generation of Osmotic Pressure (Considered in Greater Depth in Reference [82])

NOXs are almost ubiquitous across the microbial, plant and animal kingdoms. It is instructive to observe the consequences of their deletion in fungi and plants. Fungi invade plants and insects and to do this they require a mechanism with which to penetrate the tough cuticle of leaves, or the carapace of insects. They employ structures known as appressoria [83] for this purpose. When fungal hyphae find a suitable focus for invasion, they segment off the end of the hyphum to form a spherical structure. This appressorium becomes firmly attached to the surface by a cross-linking reaction using H2O2 generated by a NOX. The appressorium then drives a penetration peg through the

4

The Discovery and Characterisation of Nox2, a Personal Journey

101

adjacent surface. The penetration peg develops a pressure of about 80 atmospheres, which is 40 times that in a car tyre. Such pressure across the surface area of a hand could lift a London bus. Penetration by the peg is driven by an osmotic process requiring a second NOX [84]. Plants contain NOXs called respiratory burst oxidase homologs (RBOHs). These RBOHs are required for several important functions including the growth of root hairs and pollen tubes and the closing of stomata on leaves. All these functions require the generation of substantial osmotic pressure which could be developed by NOX driven ion fluxes followed by water molecules. It is also important to consider the effect of ions fluxes, followed by water, in the function of NOXs in humans. For example, NOXs are present in endothelial cells and have been linked to the regulation of blood pressure. There are no known regulators of perfusion and pressure operating at the capillary level. NOX activity, passing electrons into the capillary lumen, would allow for charge compensation by the passage of Cl- from the lumen into the endothelial cell which would cause osmotic swelling and a reduction of the lumen. According to Poiseuille’s Law, the resistance to laminar flow in a long cylindrical pipe is related to the fourth power of the radius, so minor changes in endothelial cell volume would have marked effects on flow and pressure. In summary, I consider one, if not the, key role of NOXs is to act as electrochemical generators of the forces to drive ion fluxes across the plasma membrane, thereby regulating such essential functions as pH and osmolality [82].

indebted to the many patients and their relatives whose contributions were so invaluable.

12

Conclusion

This chapter is a description of my major scientific journeys from the exploration of neutrophil biology through the discovery of the first NOX, and elucidation of its characteristics, to an hypothesis regarding the overall function of these molecules. I feel incredibly lucky to have had the opportunities that allowed me to make this journey, and to have been accompanied in it by so many wonderful colleagues, collaborators and competitors that made the whole enterprise such an exciting and enriching experience. Acknowledgements I would like to thank all those many colleagues that worked in my laboratory on these projects over the years and who made this journey possible. They are: Arie Abo, Philippe Behe, MarieChristine Bohler, Ann Boyhan, Colin Casimir, Margaret Chetty, Steven Coade, Jan Davidson-Moncada, Ludo Dekker, Jenny Dunne, Juliet Foote, Louisa Forbes, Rodolpho Garcia, Ann Grogan, Angela Harper, Penny Harrison, Paul Heyworth, Nick Keep, Aroon Lal, Adam Levine, Guillermo Lopez-Lluch, Tova Meshulam, Carlo Messina, Emma Murphy, Markus Nagl, Eddie Odell, Sabrina Pachero, Janne Plugge, Daphne Putman, Pam Roberts, Jurgen Roes, Peter Rowe, Dongmin Shao, Olivia Sheppard, Caroline Shrimpton, Carmel Teahan, Jane Tempero, Adrian Thrasher, Tim Wallach, Frans Wientjies and Jodi Young. I am also

References 1. Park BH, Fikrig SM, Smithwick EM (1968) Infection and nitrobluetetrazolium reduction by neutrophils: a diagnostic acid. Lancet 2: 532–534. https://doi.org/10.1016/s0140-6736(68)92406-9 2. Segal AW, Trustey SF, Levi AJ (1973) Re-evaluation of nitrobluetetrazolium test. Lancet 2:879–883 3. Segal AW, Levi AJ (1973) The mechanism of the entry of dye into neutrophils in the nitroblue tetrazolium (NBT) test. Clin Sci Mol Med 45:817–826 4. Segal AW, Levi AJ (1975) Factors influencing the entry of dye into neutrophil leucocytes in the Nitroblue tetrazolium test. Clin Sci 48: 201–212. https://doi.org/10.1042/cs0480201 5. Berendes H, Bridges RA, Good RA (1957) A fatal granulomatosus of childhood: the clinical study of a new syndrome. Minn Med 40: 309–312 6. Holmes B, Page AR, Good RA (1967) Studies of the metabolic activity of leukocytes from patients with a genetic abnormality of phagocytic function. J Clin Invest 46:1422–1432. https://doi.org/10. 1172/JCI105634 7. Baehner RL, Nathan DG (1967) Leukocyte oxidase: defective activity in chronic granulomatous disease. Science 155:835–836 8. Segal AW, Peters TJ (1977) Analytical subcellular fractionation of human granulocytes with special reference to the localization of enzymes involved in microbicidal mechanisms. Clin Sci Mol Med 52:429–442. https://doi.org/10.1042/cs0520429 9. Segal AW, Peters TJ (1978) Analytical subcellular fractionation of neutrophils from patients with chronic granulomatous disease: demonstration of the enzyme defect in four cases. Q J Med 47:213–220 10. Segal AW, Jones OT (1978) Novel cytochrome b system in phagocytic vacuoles of human granulocytes. Nature 276:515–517. https:// doi.org/10.1038/276515a0 11. Hattori H (1961) Studies on the labile, stable Nadi oxidase and peroxidase staining reactions in the isolated particles of horse granulocyte. Nagoya J Med Sci 23:362–378 12. Shinagawa Y, Tanaka C, Teraoka A, Shinagawa Y (1966) A new cytochrome in neurophilic granules of rabbit leucocyte. J Biochem 59:622–624. https://doi.org/10.1093/oxfordjournals.jbchem. a128352 13. Seligman AM, Plapinger RE, Wasserkrug HL et al (1967) Ultrastructural demonstration of cytochrome oxidase activity by the Nadi reaction with osmiophilic reagents. J Cell Biol 34:787–800. https:// doi.org/10.1083/jcb.34.3.787 14. Segal AW, Jones OT, Webster D, Allison AC (1978) Absence of a newly described cytochrome b from neutrophils of patients with chronic granulomatous disease. Lancet 2:446–449. https://doi.org/ 10.1016/s0140-6736(78)91445-9 15. Segal AW, Garcia R, Goldstone H et al (1981) Cytochrome b-245 of neutrophils is also present in human monocytes, macrophages and eosinophils. Biochem J 196:363–367. https://doi.org/10.1042/ bj1960363 16. Roberts PJ, Cross AR, Jones OT, Segal AW (1982) Development of cytochrome b and an active oxidase system in association with maturation of a human promyelocytic (HL-60) cell line. J Cell Biol 95:720–726. https://doi.org/10.1083/jcb.95.3.720 17. Borregaard N, Johansen KS, Taudorff E, Wandall JH (1979) Cytochrome b is present in neutrophils from patients with chronic granulomatous disease. Lancet 1:949–951 18. Segal AW, Jones OT (1980) Absence of cytochrome b reduction in stimulated neutrophils from both female and male patients with

102 chronic granulomatous disease. FEBS Lett 110:111–114. https://doi. org/10.1016/0014-5793(80)80035-4 19. Borregaard N, Cross AR, Herlin T et al (1983) A variant form of X-linked chronic granulomatous disease with normal nitroblue tetrazolium slide test and cytochrome b. Eur J Clin Investig 13:243– 248. https://doi.org/10.1111/j.1365-2362.1983.tb00095.x 20. Bolscher BG, de Boer M, de Klein A et al (1991) Point mutations in the beta-subunit of cytochrome b558 leading to X-linked chronic granulomatous disease. Blood 77:2482–2487 21. Segal AW, Cross AR, Garcia RC et al (1983) Absence of cytochrome b- 245 in chronic granulomatous disease: a multicenter European evaluation of its incidence and relevance. N Engl J Med 308(5):245–251 22. Royer-Pokora B, Kunkel LM, Monaco AP et al (1986) Cloning the gene for an inherited human disorder–chronic granulomatous disease–on the basis of its chromosomal location. Cold Spring Harb Symp Quant Biol 322:32–38. https://doi.org/10.1038/ 322032a0 23. Teahan C, Rowe P, Parker P et al (1987) The X-linked chronic granulomatous disease gene codes for the beta-chain of cytochrome b-245. Nature 327:720–721. https://doi.org/10.1038/327720a0 24. Cross AR, Jones OT, Harper AM, Segal AW (1981) Oxidationreduction properties of the cytochrome b found in the plasmamembrane fraction of human neutrophils. A possible oxidase in the respiratory burst. Biochem J 194:599–606. https://doi.org/10. 1042/bj1940599 25. Babior BM, Kipnes RS (1977) Superoxide-forming enzyme from human neutrophils: evidence for a flavin requirement. Blood 50: 517–524 26. Light DR, Walsh C, O’Callaghan AM et al (1981) Characteristics of the cofactor requirements for the superoxide-generating NADPH oxidase of human polymorphonuclear leukocytes. Biochemistry 20:1468–1476. https://doi.org/10.1021/bi00509a010 27. Cross AR (1987) The inhibitory effects of some iodonium compounds on the superoxide generating system of neutrophils and their failure to inhibit diaphorase activity. Biochem Pharmacol 36:489–493. https://doi.org/10.1016/0006-2952(87)90356-x 28. Kakinuma K, Kaneda M, Chiba T, Ohnishi T (1986) Electron spin resonance studies on a flavoprotein in neutrophil plasma membranes. Redox potentials of the flavin and its participation in NADPH oxidase. J Biol Chem 261:9426–9432 29. Bellavite P, Cross AR, Serra MC et al (1983) The cytochrome b and flavin content and properties of the O2--forming NADPH oxidase solubilized from activated neutrophils. Biochim Biophys Acta 746: 40–47. https://doi.org/10.1016/0167-4838(83)90008-0 30. Cross AR, Parkinson JF, Jones OT (1984) The superoxidegenerating oxidase of leucocytes. NADPH-dependent reduction of flavin and cytochrome b in solubilized preparations. Biochem J 223: 337–344. https://doi.org/10.1042/bj2230337 31. Kakinuma K, Fukuhara Y, Kaneda M (1987) The respiratory burst oxidase of neutrophils. Separation of an FAD enzyme and its characterization. J Biol Chem 262:12316–12322 32. Doussiere J, Vignais PV (1985) Purification and properties of an O2--generating oxidase from bovine polymorphonuclear neutrophils. Biochemistry 24:7231–7239 33. Segal AW, West I, Wientjes F et al (1992) Cytochrome b-245 is a flavocytochrome containing FAD and the NADPH-binding site of the microbicidal oxidase of phagocytes. Biochem J 284:781–788 34. Gabig TG, Lefker BA (1984) Deficient flavoprotein component of the NADPH-dependent O2-.-generating oxidase in the neutrophils from three male patients with chronic granulomatous disease. J Clin Invest 73:701–705. https://doi.org/10.1172/JCI111262 35. Bohler MC, Seger RA, Mouy R et al (1986) A study of 25 patients with chronic granulomatous disease: a new classification by correlating respiratory burst, cytochrome b, and flavoprotein. J Clin Immunol 6:136–145. https://doi.org/10.1007/BF00918746

A. W. Segal 36. Cross AR, Jones OT, Garcia R, Segal AW (1982) The association of FAD with the cytochrome b-245 of human neutrophils. Biochem J 208:759–763. https://doi.org/10.1042/bj2080759 37. Crawford DR, Schneider DL (1982) Identification of ubiquinone-50 in human neutrophils and its role in microbicidal events. J Biol Chem 257:6662–6668. https://doi.org/10.1016/S0021-9258(18) 34480-6 38. Cross AR, Jones OT, Garcia R, Segal AW (1983) The subcellular localization of ubiquinone in human neutrophils. Biochem J 216: 765–768. https://doi.org/10.1042/bj2160765 39. Umei T, Takeshige K, Minakami S (1986) NADPH binding component of neutrophil superoxide-generating oxidase. J Biol Chem 261: 5229–5232 40. Umei T, Babior BM, Curnutte JT, Smith RM (1991) Identification of the NADPH-binding subunit of the respiratory burst oxidase. J Biol Chem 266:6019–6022 41. Harper AM, Chaplin MF, Segal AW (1985) Cytochrome b-245 from human neutrophils is a glycoprotein. Biochem J 227:783–788. https://doi.org/10.1042/bj2270783 42. Segal AW (1981) Absence of both cytochrome b-245 subunits from neutrophils in X-linked chronic granulomatous disease. Nature 326: 88–91. https://doi.org/10.1038/326088a0 43. Parkos CA, Allen RA, Cochrane CG, Jesaitis AJ (1987) Purified cytochrome b from human granulocyte plasma membrane is comprised of two polypeptides with relative molecular weights of 91,000 and 22,000. J Clin Invest 80:732–742. https://doi.org/10. 1172/JCI113128 44. Wallach TM, Segal AW (1996) Stoichiometry of the subunits of flavocytochrome b558 of the NADPH oxidase of phagocytes. Biochem J 320(Pt 1):33–38. https://doi.org/10.1042/bj3200033 45. Taylor WR, Jones DT, Segal AW (1993) A structural model for the nucleotide binding domains of the flavocytochrome b-245 betachain. Protein Sci 2:1675–1685. https://doi.org/10.1002/pro. 5560021013 46. Karplus PA, Daniels MJ, Herriott JR (1991) Atomic structure of ferredoxin-NADP+ reductase: prototype for a structurally novel flavoenzyme family. Science 251:60–66 47. Wallach TM, Segal AW (1997) Analysis of glycosylation sites on gp91phox, the flavocytochrome of the NADPH oxidase, by sitedirected mutagenesis and translation in vitro. Biochem J 321(Pt 3): 583–585. https://doi.org/10.1042/bj3210583 48. Cross AR, Rae J, Curnutte JT (1995) Cytochrome b-245 of the neutrophil superoxide-generating system contains two nonidentical hemes. Potentiometric studies of a mutant form of gp91phox. J Biol Chem 270:17075–17077. https://doi.org/10.1074/jbc.270.29.17075 49. Shatwell KP, Dancis A, Cross AR et al (1996) The FRE1 ferric reductase of Saccharomyces cerevisiae is a cytochrome b similar to that of NADPH oxidase. J Biol Chem 271:14240–14244 50. Finegold AA, Shatwell KP, Segal AW et al (1996) Intramembrane bis-heme motif for transmembrane electron transport conserved in a yeast iron reductase and the human NADPH oxidase. J Biol Chem 271:31021–31024 51. Segal AW, Heyworth PG, Cockcroft S, Barrowman MM (1985) Stimulated neutrophils from patients with autosomal recessive chronic granulomatous disease fail to phosphorylate a Mr-44,000 protein. Nature 316:547–549. https://doi.org/10.1038/316547a0 52. Casimir CM, Bu-Ghanim HN, Rodaway AR et al (1991) Autosomal recessive chronic granulomatous disease caused by deletion at a dinucleotide repeat. Proc Natl Acad Sci USA 88:2753–2757. https://doi.org/10.1073/pnas.88.7.2753 53. Clark RA, Volpp BD, Leidal KG, Nauseef WM (1990) Two cytosolic components of the human neutrophil respiratory burst oxidase translocate to the plasma membrane during cell activation. J Clin Invest 85:714–721. https://doi.org/10.1172/JCI114496 54. Teahan CG, Totty N, Casimir CM, Segal AW (1990) Purification of the 47 kDa phosphoprotein associated with the NADPH oxidase of

4

The Discovery and Characterisation of Nox2, a Personal Journey

103

human neutrophils. Biochem J 267:485–489. https://doi.org/10. 1042/bj2670485 55. Heyworth PG, Shrimpton CF, Segal AW (1989) Localization of the 47 kDa phosphoprotein involved in the respiratory-burst NADPH oxidase of phagocytic cells. Biochem J 260:243–248. https://doi. org/10.1042/bj2600243 56. Abo A, Pick E (1991) Purification and characterization of a third cytosolic component of the superoxide-generating NADPH oxidase of macrophages. J Biol Chem 266:23577–23585 57. Abo A, Pick E, Hall A et al (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1. Nature 353:668–670. https://doi.org/10.1038/353668a0 58. Keep NH, Barnes M, Barsukov I et al (1997) A modulator of rho family G proteins, rhoGDI, binds these G proteins via an immunoglobulin-like domain and a flexible N-terminal arm. Structure 5:623–633. https://doi.org/10.1016/s0969-2126(97)00218-9 59. Abo A, Webb MR, Grogan A, Segal AW (1994) Activation of NADPH oxidase involves the dissociation of p21rac from its inhibitory GDP/GTP exchange protein (rhoGDI) followed by its translocation to the plasma membrane. Biochem J 298:585–591. https:// doi.org/10.1042/bj2980585 60. Diekmann D, Abo A, Johnston C et al (1994) Interaction of Rac with p67phox and regulation of phagocytic NADPH oxidase activity. Science 265:531–533. https://doi.org/10.1126/science.8036496 61. Wientjes FB, Hsuan JJ, Totty NF, Segal AW (1993) p40phox, a third cytosolic component of the activation complex of the NADPH oxidase to contain src homology 3 domains. Biochem J 296:557– 561. https://doi.org/10.1042/bj2960557 62. van de Geer A, Nieto-Patlán A, Kuhns DB et al (2018) Inherited p40phox deficiency differs from classic chronic granulomatous disease. J Clin Invest 128:3957–3975. https://doi.org/10.1172/ JCI97116 63. Segal AW, Clark J, Allison AC (1978) Tracing the fate of oxygen consumed during phagocytosis by human neutrophils with 15O2. Clin Sci Mol Med 55:413–415. https://doi.org/10.1042/cs0550413 64. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanisms: the production by leukocytes of superoxide, a potential bactericidal agent. J Clin Invest 52:741–744. https://doi.org/10. 1172/JCI107236 65. Reeves EP, Nagl M, Godovac-Zimmermann J, Segal AW (2003) Reassessment of the microbicidal activity of reactive oxygen species and hypochlorous acid with reference to the phagocytic vacuole of the neutrophil granulocyte. J Med Microbiol 52:643–651. https:// doi.org/10.1099/jmm.0.05181-0 66. Klebanoff SJ, Kettle AJ, Rosen H et al (2013) Myeloperoxidase: a front-line defender against phagocytosed microorganisms. J Leukoc Biol 93:185–198. https://doi.org/10.1189/jlb.0712349 67. Levine AP, Segal AW (2016) The NADPH oxidase and microbial killing by neutrophils, with a particular emphasis on the proposed antimicrobial role of myeloperoxidase within the phagocytic vacuole. Microbiol Spectr 4:599–613. https://doi.org/10.1128/ microbiolspec.MCHD-0018-2015 68. Jacques YV, Bainton DF (1978) Changes in pH within the phagocytic vacuoles of human neutrophils and monocytes. Lab Invest 39: 179–185

69. Segal AW, Geisow M, Garcia R, Harper A, Miller R (1981) The respiratory burst of phagocytic cells is associated with a rise in vacuolar pH. Nature 290(5805):406–409. https://doi.org/10.1038/ 290406a0. PMID: 7219526 70. Levine AP, Duchen MR, de VS et al (2015) Alkalinity of neutrophil phagocytic vacuoles is modulated by HVCN1 and has consequences for myeloperoxidase activity. PLoS One 10:e0125906. https://doi. org/10.1371/journal.pone.0125906 71. Foote JR, Levine AP, Behe P et al (2017) Imaging the neutrophil phagosome and cytoplasm using a ratiometric pH indicator. J Vis Exp 122:55107. https://doi.org/10.3791/55107 72. Tkalcevic J, Novelli M, Phylactides M et al (2000) Impaired immunity and enhanced resistance to endotoxin in the absence of neutrophil elastase and cathepsin G. Immunity 12:201–210 73. Reeves EP, Lu H, Jacobs HL et al (2002) Killing activity of neutrophils is mediated through activation of proteases by K+ flux. Nature 416:291–297. https://doi.org/10.1038/416291a 74. Baehner RL, Karnovsky MJ, Karnovsky ML (1969) Degranulation of leukocytes in chronic granulomatous disease. J Clin Invest 48: 187–192. https://doi.org/10.1172/JCI105967 75. Segal AW (2019) Studies on patients establish Crohn’s disease as a manifestation of impaired innate immunity. J Intern Med 286(4): 373–388. https://doi.org/10.1111/joim.12945 76. Styrt B, Klempner MS (1982) Internal pH of human neutrophil lysosomes. FEBS Lett 149:113–116. https://doi.org/10.1016/00145793(82)81083-1 77. Ahluwalia J, Tinker A, Clapp LH et al (2004) The large-conductance Ca2+-activated K+ channel is essential for innate immunity. Nature 427:853–858. https://doi.org/10.1038/nature02356 78. Retraction (2010) The large-conductance Ca(2+)-activated K(+) channel is essential for innate immunity. Nature 468(7320):122. https://europepmc.org/article/MED/21048767. Accessed 14 Oct 2021 79. Retraction, Ahluwalia J, Yaqoob M, Urban L et al (2003) Activation of capsaicin-sensitive primary sensory neurones induces anandamide production and release. J Neurochem 84:585–591. https:// doi.org/10.1046/j.1471-4159.2003.01550.x 80. Foote JR, Behe P, Frampton M et al (2017) An exploration of charge compensating ion channels across the phagocytic vacuole of neutrophils. Front Pharmacol 8:94. https://doi.org/10.3389/fphar. 2017.00094 81. Paffenholz R, Bergstrom RA, Pasutto F et al (2004) Vestibular defects in head-tilt mice result from mutations in Nox3, encoding an NADPH oxidase. Genes Dev 18:486–491. https://doi.org/10. 1101/gad.1172504 82. Segal AW (2016) NADPH oxidases as electrochemical generators to produce ion fluxes and turgor in fungi, plants and humans. Open Biol 6(5):160028. https://doi.org/10.1098/rsob.160028 83. Talbot NJ (2019) Appressoria. Curr Biol 29:R144–R146. https://doi. org/10.1016/j.cub.2018.12.050 84. Egan MJ, Wang Z-Y, Jones MA et al (2007) Generation of reactive oxygen species by fungal NADPH oxidases is required for rice blast disease. Proc Natl Acad Sci USA 104:11772–11777. https://doi.org/ 10.1073/pnas.0700574104

5

Reminiscences on Positional Cloning of X-CGD Gene (Aka CYBB, gp91phox, Nox2) Stuart H. Orkin

Abstract

In 1986 my laboratory reported the cloning of the gene that encodes the protein mutated in the X-linked form of chronic granulomatous disease (X-CGD), which represented the first example of “positional cloning” of a human disease gene. At the time, molecular tools were primitive compared with those available to investigators today. Successful cloning of the gene relied on pivotal collaborations, intuition, and good fortune. Here I provide my recollections of how I came to study X-CGD and how the experiments unfolded. Keywords

X-CGD · Positional cloning · Gene deletion · HL60 cells · DMD · Subtractive hybridization · Phagocyte oxidase

1

Background

In the early 1980s, I established my research laboratory in the Division of Hematology/Oncology at Boston Children’s Hospital (BCH), an affiliate of the Harvard Medical School (HMS). The theme of my work was to deploy the emerging tools of molecular biology and gene cloning to understand diseases of the blood with the ultimate goal of developing new therapeutic approaches for better treatment, or cure. As an alternative to military service during the Vietnam War era, I chose a Research Associate position in the US Public Health Service at the National Institutes of Health (NIH), where I learned the rudiments of molecular biology at the time (1973–1975) in the laboratory of the late Philip Leder, a pioneer in molecular genetics and then gene cloning. At the NIH I studied how messenger RNA for globin was expressed S. H. Orkin (✉) Harvard Medical School, Howard Hughes Medical Institute, Boston, MA, USA e-mail: [email protected]

in developing erythroid precursors, a topic that has consumed my thinking for nearly my entire career. Later, as a newly appointed junior faculty member at BCH and HMS in the early 1980s, I applied genomic cloning to define the mutations in the β-thalassemias, a collection of anemias due to underproduction of β-globin. In collaboration with Haig Kazazian at Johns Hopkins, we elucidated the myriad of mutations in the β-globin gene leading to β-thalassemia, culminating in the first comprehensive analysis of a human inherited disorder with recombinant DNA methods [1]. By ~1984, I considered this work complete and began to think of other areas where molecular approaches might shed light on disease pathogenesis. Having been trained in pediatrics and hematology at BCH and recruited to my faculty position by David G. Nathan, at the time (and still) a “giant” in hematology, I had learned about the disorder known locally as Fatal Granulomatous Disease, or Chronic Granulomatous Disease (CGD), as it was intimately connected to our hospital. At the annual meeting of the Society for Pediatric Research in 1954, Charles Janeway, the Chief of the Department of Pediatrics when I was an intern, described several children with elevated serum gamma globulin and repeated severe infections [2, 3]. Pediatric trainees became familiar with many of the CGD patients, as they were often in the hospital for extended periods of treatment with antibiotics. After its initial description, Good and colleagues dubbed the disorder fatal granulomatous disease and noted the preponderance of boys, indicating its X-linked inheritance [4]. With Robert Baehner, a fellow in the hematology program, Nathan developed a simple staining protocol, which could identify those phagocytic cells unable to reduce nitroblue tetrazolium (NBT, the “NBT Test”) and also distinguish X-linked from recessive forms of CGD [5]. Despite considerable interest in this disorder of host defense, the pathogenesis was a puzzle. What proteins comprised the phagocyte oxidase?

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_5

105

106

2

S. H. Orkin

Searching for the X-CGD Gene

At the time the work on characterization of β-thalassemia mutations was winding down, two paths intersected to draw me to the CGD problem. Two floors below my laboratory in the Enders Research Building at BCH, Lou Kunkel was intensively hunting for the X-linked muscular dystrophy gene. I interacted closely with Lou from the time of his recruitment to BCH by the late Sam Latt, a geneticist with whom my laboratory collaborated on generating chromosome specific genomic clones [6]. Lou was using gene linkage and bacteriophage cloning to identify the gene for X-linked muscular dystrophy (DMD), a topic that became the thesis of a talented MD/PhD student, Tony Monaco. A breakthrough in the DMD field at the time was discovery of a rare patient (called BB) who had three genetic diseases— DMD, retinitis pigmentosa, and CGD—, a constellation of phenotypes resulting from a small, but cytologically detectable, deletion on the short arm of the X-chromosome (Xp21). Kunkel and Monaco were actively engaged in identifying lambda bacteriophage clones of X-chromosome DNA that resided with the BB deletion region, as some of these would be anticipated to encompass a portion of the DMD gene [7]. As part of their pursuit of such clones, they employed a subtractive hybridization approach to enrich for sequences within the deletion. I knew of their work and had many discussions with both Kunkel and Monaco at the time. So, I began to think about whether we might be able to clone the CGD gene using similar logic. My interest was spurred further when I received a letter from Baehner (Fig. 5.1), Nathan’s former trainee, who was completing a term as Chief of Pediatrics at LA Children’s Hospital, and sought to do a sabbatical in the laboratory. Baehner wanted to use material from a patient (NF) he had encountered that was similar to BB. He arrived in 1984 with

DNA samples of X-CGD patients and their families to perform linkage analysis, in collaboration with Lou and Tony, in an effort to pinpoint more precisely the location of the CGD gene. Although he was a neophyte at the bench, this work went smoothly and we published a paper supporting the position of the CGD gene within the BB deletion region [8] (Fig. 5.1). However, if we were to clone the unknown gene, someone with more experience in molecular biology needed to get involved. I was fortunate that such an individual joined my group, Brigitte Royer-Pokora, a trained junior scientist whose husband I knew quite independently from other interactions around the Harvard Medical area. We hypothesized that phagocytic cells would contain abundant mRNA from the unknown CGD gene. To enrich for such RNA sequences (or cDNA) we employed the subtractive hybridization protocol used by Kunkel and Monaco in their DMD studies to subtract widely enriched, expressed sequences from RNA prepared from differentiated HL60 cells, a model of phagocytic cells. To guard against subtracting away the desired RNA, we used RNA from immortalized B-cells of patient NF. Using this approach, we obtained cDNA sequences presumably enriched for those from the CGD gene (Fig. 5.2). But what might this be used for? Royer-Pokora and I discussed various options. She also met frequently with Monaco as well. In the Kunkel laboratory, numerous bacteriophages clones carrying small portions ( R, where R showed some, not zero, cytochrome c reducing activity;

6

On Katsuko Kakinuma: Spectroscopic Studies of Redox Centers in NADPH Oxidase – “Identifying and O. . .

117

2. NADPH ! NADPH oxidase (flavin redox center) ! cytochrome b558: activity of S > R, where R showed some, not zero, cytochrome b558 reducing activity. Thus, the S preparations showed higher activities than R for both electron-transfer reactions, but the activity of the R preparations was never zero. These results indicated that, while the electron flows from NADPH to cytochrome b558 or cytochrome c were occurring, these activities might not be related to the superoxide-generating activity under aerobic condition, since the R preparation still showed some activity. On the other hand, an interesting phenomenon was noted: the photoreduction of cytochrome b558 in membrane preparations takes place under conditions of photo-irradiation at 450 nm, a selective wavelength known to cause excitation of flavoproteins. When the photo-irradiation at 450 nm was applied to the above reaction (1), the photoreduction of cytochrome c was remarkably enhanced only in the S preparation, and not in the R sample. This reducing activity in S preparations was postulated to reflect the superoxideproducing activity of stimulated neutrophils. To clarify the involvement of the spin state of the heme in cytochrome b558 in the superoxide-generating activity, Dr. Kakinuma and colleagues employed a cell-free system consisting of purified cytochrome b558 with varying amounts of low-spin and high-spin heme, under conditions of various pHs and different fatty acid treatments. Based on biochemical and spectroscopic studies of purified cytochrome b558 in this cell-free system [22], the superoxide-generating activity was shown to increase with increasing percentage of the low-spin heme in cytochrome b558 (Fig. 6.6). We concluded that the low-spin ferric form of the heme is essential for the superoxide generation by the neutrophil NADPH oxidase system. Furthermore, the transient high-spin ferric heme that was induced during activation by arachidonic acid in a cell-free system did not correlate with the common activation mechanism in the oxidase system [22].

5

Epilogue

For over 40 years, Dr. Kakinuma had been engaged in studying the electron-transport system of NADPH oxidase in neutrophils, but her work extended beyond that topic. She supported the work of physicians at the nearby the Tokyo Metropolitan Institute of Medical Sciences, who were treating patients with severe infectious diseases. Most of these patients were young children or babies with high fevers who showed the severe symptoms characteristic of recurrent bacterial and fungal infections. In these cases, Dr. Kakinuma helped physicians examine several biological functions of patient’s neutrophils, especially quantitative analysis of

Fig. 6.6 Effect of the percentage of the low-spin heme in cytochrome b558 on the O2--generating activity of the NADPH oxidase system. Samples with different ratios of low-spin to high-spin heme were prepared by adjusting the pH (●) or by heat-treatment (▲) at 40 °C. The O2--generating activities of these preparations were measured in the cell-free system at pH 7.0. Adapted with permission from Fujii et al. [22]

superoxide-generating activity of cells. For newborn babies, it was quite difficult to collect the required 10–20 mL blood. To meet this clinical challenge, Dr. Kakinuma contributed to the development of a new method for measuring oxygen consumption of neutrophils by using whole blood, obviating the need to isolate neutrophils from whole blood. In 1990, Dr. Kakinuma invited Dr. Robert Clark and Dr. William Nauseef, from the University of Iowa to her laboratory in Tokyo. They came to Japan with a polyclonal antiserum (B-1) that recognizes both p47phox and p67phox in the cytosol of neutrophils [23]. Using their polyclonal antiserum, they initiated an investigation of cytosolic proteins obtained from samples from CGD patients, permitting the identification of the first case in Japan of a CGD patient whose cytosol lacks p67phox. In subsequent work, Dr. Kakinuma had continued biochemical and genetic analyses of several types of CGD patients using modern gene analysis methods combined with spectroscopic techniques [24]. Dr. Kakinuma ranks as one of the pioneers of the study of neutrophil NADPH oxidase, although with her usual modesty, she would insist that she had only opened a small door

118

to the science of neutrophils. Nonetheless, as has been shown in this chapter, Dr. Kakinuma has made major contributions on this topic, beginning with her studies of neutrophil functions by stimulating superoxide generation with fatty acids, such as myristic acid. In later years, she expressed her delight in learning that myristic acid stimulates superoxide-generating activity not only in intact neutrophils but also in the cell-free system prepared from neutrophils, a path initiated by another pioneer of the field, Dr. Edgar Pick. Acknowledgments All members of the Department of Inflammation Research at the Tokyo Metropolitan Institute of Medical Sciences, who were active at the time that Dr. Kakinuma was the director, thank all the researchers who visited the laboratory, for their valuable and enthusiastic discussions and collaborations. The authors thank Dr. Akio Tomoda, Department of Biochemistry, Tokyo Medical University, for reading this manuscript and for his helpful suggestions.

References 1. Sbarra AJ, Karnovsky ML (1959) The biochemical basis of phagocytosis. I. Metabolic changes during the ingestion of particles by polymorphonuclear leukocytes. J Biol Chem 234:1355–1362 2. Rossi F, Zatti M (1964) Changes in the metabolic pattern of polymorphonuclear leukocytes during phagocytosis. Br J Exp Pathol 45: 548–559 3. Romeo D, Cramer R, Rossi F (1970) Use of 1-anilino-8-naphtalene sulfonate to study structural transitions in cell membrane of PMN leucocytes. Biochem Biophys Res Commun 41:582–588 4. Kakinuma K (1974) Effects of fatty acids on the oxidative metabolism of leukocytes. Biochim Biophys Acta 348:76–85 5. Kakinuma K, Minakami S (1978) Effects of fatty acids on superoxide radical generation in leukocytes. Biochim Biophys Acta 538:50– 59 6. Kakinuma K, Kaneda M (1980) Kinetic studies on the H2O2 (O2-)forming enzyme in Guinea pig leukocytes. FEBS Lett 111:90–94 7. Gabig TG, Babior BM (1979) The O2--forming oxidase responsible for the respiratory burst in human neutrophils. Properties of the solubilized enzyme. J Biol Chem 254:9070–9074 8. Light DR, Walsh C, O’Callaghan AM et al (1981) Characteristics of the cofactor requirements for the superoxide-generating NADPH oxidase of human polymorphonuclear leukocytes. Biochemistry 20:1468–1476 9. Kakinuma K, Fukuhara Y, Kaneda M (1987) The respiratory burst oxidase of neutrophils. Separation of an FAD enzyme and its characterization. J Biol Chem 262:12316–12322 10. Kakinuma K, Kaneda M, Chiba T, Ohnishi T (1986) Electron spin resonance studies on a flavoprotein in neutrophil plasma

H. G. Fujii and L. S. Yoshida membranes: redox potentials of the flavin and its participation in NADPH oxidase. J Biol Chem 261:9426–9432 11. Vignais PV (2002) The superoxide-generating NADPH oxidase: structural aspects and activation mechanism. Cell Mol Life Sci 59: 1428–1459 12. Segal AW, Jones OT (1978) Novel cytochrome b system in phagocytic acuoles of human granulocytes. Nature 276:515–517 13. Cross AR, Jones OT, Harper AM, Segal AW (1981) Oxidationreduction properties of the cytochrome b found in the plasmamembrane fraction of human neutrophils. A possible oxidase in the respiratory burst. Biochem J 194:599–606 14. Hurst JK, Loehr TM, Curnutte JT, Rosen H (1991) Resonance Raman and electron paramagnetic resonance structural investigations of neutrophil cytochrome b558. J Biol Chem 266: 1627–1634 15. Ueno I, Fujii S, Ohya-Nishiguchi H et al (1991) Characterization of neutrophil b-type cytochrome in situ by electron paramagnetic resonance spectroscopy. FEBS Lett 281:130–132 16. Miki T, Fujii H, Kakinuma K (1992) EPR signals of cytochrome b558 purified from porcine neutrophils. J Biol Chem 267:19673– 19675 17. Fujii H, Johnson MK, Finnegan MG et al (1995) Electron spin resonance studies on neutrophil cytochrome b558: evidence that low-spin heme iron is essential for O2--generating activity. J Biol Chem 270:12685–12689 18. Fujii H, Kakinuma K (1992) Electron paramagnetic resonance studies on cytochrome b-558 and peroxidases of pig blood granulocytes. Biochim Biophys Acta 1136:239–246 19. Yamaguchi T, Hayakawa T, Kaneda M et al (1989) Purification and some properties of the small subuni of cytochrome b558 from human neutrophils. J Biol Chem 264:112–118 20. Fujii H, Finnegan MG, Miki T et al (1995) Spectroscopic identification of the heme axial ligation of cytochrome b558 in the NADPH oxidase of porcine neutrophils. FEBS Lett 377:345–348 21. Fujii H, Kakinuma K (1991) Electron transfer reactions in the NADPH oxidase system of neutrophils – involvement of an NADPH-cytochrome c reductase in the oxidase system. Biochim Biophys Acta 1095:201–209 22. Fujii H, Finnegan MG, Johnson MK (1999) The active form of the ferric heme in neutrophil cytochrome b558 is low-spin in the reconstituted cell-free system in the presence of amphophil. J Biochem 126:708–714 23. Volpp BD, Nauseef WM, Donelson JE et al (1989) Cloning of the cDNA and functional expression of the 47-kilodalton cytosolic component of human neutrophil respiratory burst oxidase. Proc Natl Acad Sci USA 86:7195–7199 24. Kaneda M, Sakuraba H, Ohtake A et al (1999) Missense mutations in the gp91-phox gene encoding cytochrome b558 in patients with cytochrome b positive and negative X-linked chronic granulomatous disease. Blood 93:2098–2104

7

Pierre Vignais, from One Respiratory Chain to Another. . . Marie-Claire Dagher

Abstract

Professor Pierre V. Vignais played a major role in developing biochemistry in the French Alps in the sixties. He was very influential to the scientific bio-medical community in Grenoble. Although his early interest was in mitochondria, he became thrilled by the mystery of the respiratory burst of neutrophils in the middle of his career and transmitted his enthusiasm to his students and coworkers. Besides his research articles, he is the author of two books on the scientific method in biology. Keywords

Biochemistry · Mitochondria · Respiratory burst · Neutrophils · Nox2

Pierre Vignais (1926–2006) played a major role in the development of biochemistry in Grenoble, a city in the French Alps that until then had focused on physics and mathematics. He was the author of important work in the field of bioenergetics, before his passion led him to the respiratory burst of neutrophils, in the 1980s. Just as he himself had embarked on a course of study that combined medical and scientific studies, he encouraged his students to do the same. He used to offer his second year medical students an internship in his laboratory. He then encouraged his students to pursue scientific studies in parallel with medical studies, or conversely, he tried to encourage scientists to undertake medical studies. This double curriculum is quite rare in France: it requires a total commitment to the “hard” sciences. Indeed, he liked the students to stay late in the laboratory and he urged them to come on Saturdays and Sundays as well, and to save the reading of literature for their evenings. In fact, his only concern was to get everyone to do M.-C. Dagher (✉) University Grenoble Alpes, CNRS, CEA, IBS, Grenoble, France e-mail: [email protected]

their best, and he considered research a priesthood, to which one must sacrifice everything. He was walking his talk and truly dedicated his life to research. Tall and serious, always wearing a blue sweater, he impressed his interlocutors and used his presence to intimidate them when they asked him to sign authorizations for absence: “Vacation? You want to take a vacation? Am I taking a vacation?” he would say with a mischievous wink to his secretary in front of the disconcerted student. What I admired most about him was that he was a master of time. He was never overwhelmed, he always had time to answer a question or solve an urgent problem. We frequently went to his office, which he shared with his wife Paulette, co-director of the laboratory. The shelves were lined with large black binders, with the topics marked on them, and which covered the entire bibliography available. For our young readers, until the 1990s, it was possible to know (and print) all the literature on one’s subject. Another of his characteristics was a complete lack of interest in politics or power, in anything that did not directly serve science, i.e. he was concerned only by the life of his laboratory, permanent positions for his researchers, scholarships for his students, and in this area, he never gave up. He used his influence in the commissions of the National Committee of the Centre National de la Recherche Scientifique and in the sessions of the Conseil National des Universités, to find positions for his researchers according to the opportunities. He also provided them with a lot of training; as an example, for those who prepared for the title of Professor and had to prepare a lecture as part of the examination (the French “Agrégation”). There was never a lack of success and funding in his laboratory! He had suffered from hunger during the Second World War and he told that 1 day, invited to the home of a lady who had nothing edible to offer him, he fell back on her mustard pot. The war did not prevent him from beginning his medical studies at the University of Nantes, before moving to Paris where he started an internship at the Pasteur Institute. There

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_7

119

120

he also met his wife Paulette. He did the work for his science thesis on the biochemistry of RNA. He completed his double MD PhD degree at the age of 26. In 1955, at the age of 29, he became a professor at the University of Algiers where he pursued work on the oxidative phosphorylation of isolated mitochondria. In Algeria, he became interested in the properties of atractylate, a poison derived from a thistle, which later proved to be a tool of choice for the study of the adenine nucleotide transporter. When Algeria became independent in 1962, he had to return to the metropole, and he kept an affection for his compatriots who had to leave Algeria. He was offered a position at the University of Grenoble, and set up a clinical biology laboratory attached to the hospital, which performed enzyme assays for patients (Fig. 7.1). At the same time, he was invited to set up a biochemistry laboratory at the Commissariat à l’Energie Atomique in Grenoble, which he co-directed for some 40 years, with his wife. In fact, as the « goodfather » of biochemistry in Grenoble, he was the mentor of many prominent researchers and hospital professors. His views influenced the way the Grenoble bio-medical scientific community thought and worked. He was bold and self-confident, and also quite competitive. In the field of bioenergetics, his work, initiated in 1953 at Oxford in the laboratory of Hans Krebs, the Nobel Prize winner, first concerned the phosphorylation associated with the oxidation of isocitrate in isolated mitochondria. He took an early stand in favor of Mitchell’s chemosmotic theory in the controversy that raged in the 1970s. He succeeded in “uncoupling” ADP/ATP transport from the respiratory chain and ATP synthesis in mitochondria [1]. The sustained efforts of his coworkers led to the structure of the ADP/ATP carrier in 2003 [2]. His team made clever use of chemical probes to study ATP synthesis and translocation [3], but he was to follow a new love, the NADPH oxidase . . . In the early 1980s, he became interested in the cyanideresistant respiration of neutrophils. Neutrophils exposed to opsonized particles that they were phagocytosing started to consume oxygen and oxidize NADPH. This phenomenon was absent in chronic granulomatous disease, and sometimes a type b cytochrome was also missing. Pierre Vignais launched two researchers from his team in search of cytochrome b on the one hand, and of the protein that oxidized NADPH on the other. It was quite a shock when it turned out that both components were identical [4]. It was the time of going to the slaughterhouse, to get cans of beef blood to purify the NADPH oxidase components from bovine neutrophils. Pierre Vignais was very enthusiastic about the discovery of cell-free systems [5, 6]: “they have opened the black box”, he said. Now biochemistry was becoming possible thanks to « a reductionist system ». He was convinced that a technical

M.-C. Dagher

Fig. 7.1 Pierre Vignais in 1993, photographed at the occasion of the 30th anniversary of his clinical biology laboratory

mistake in the concentration of arachidonic acid had led to this discovery: upon adding ten times more arachidonic acid than usual, the cells were dead but the oxidase was active. In the Grenoble variation of the cell-free system, a pre-incubation step between the components allowed to obtain an active oxidase, which was then diluted in the spectrophotometer cell for the determination of the activity, in the presence of cytochrome c (and blocked by superoxide dismutase [7]). It was enough to mix cytosol (the supernatant of a centrifugation at 100,000 g, he kept precising), a membrane fraction, non-hydrolysable GTP, and arachidonic acid. There are differences between species and SDS, as an activator, worked very rarely on bovine neutrophil membranes. The cell-free system allowed the purification of the cytosolic factors to be undertaken but we arrived too late to clone the cDNAs [8, 9]. Pierre Vignais was very enthusiastic for all that was related to science. This was an advantage but also a drawback: « Protein kinase C is involved [10], let’s purify protein kinase C! ». The search for the G protein deduced from the promoting effect of GTP γ-S [11] strayed to the substrate of the Clostridium botulinum C3 exoenzyme which was RhoA and not Rac [12] (in complex with Rho-GDI).

7

Pierre Vignais, from One Respiratory Chain to Another. . .

His laboratory was first in identifying the sequence of the formyl peptide receptor [13] and the C5a receptor [14] by expression cloning. The role of Rac and the importance of its prenylation to interact with guanine nucleotide exchange factors was then shown in a semi-recombinant cell-free system in which the cytosolic factors were overexpressed in the baculovirus system [15]. This recombinant expression opened the way to the determination of crystallographic structures: the Rac1/Rho-GDI complex [16], the active N-terminus of p67phox [17], and the SH3 domain of p40phox [18]. The Grenoble group was recognized for its early contribution to the study of protein-protein interactions in the p40phox/p67phox/p47phox triad, using the yeast two-hybrid system [19]. Pierre Vignais’s last paper was on the cryptic NADPH oxidase of dendritic cells [20], after having been involved in the study of that of EBV-B lymphocytes [21]. Several groups are pursuing the work on the NADPH oxidase in Grenoble. Pierre Vignais was not only enthusiastic; he was a visionary. He had foreseen—for example—the potential of the yeast two-hybrid system to study protein-protein interactions, and the importance of structural biology at a time when crystallography was about to explode. He accepted to serve as the interim director of the Institute of Structural Biology after its creation in 1992. After he retired, he wrote two books on the history of life sciences «Science expérimentale et conception du vivant: la méthode et les concepts» [22] and «La biologie des origines à nos jours: une histoire des idées et des hommes» [23], and had other books in preparation at the time of his death after a short illness [24].

References 1. Stubbs M, Vignais PV, Krebs HA (1978) Is the adenine nucleotide translocator rate-limiting for oxidative phosphorylation? Biochem J 172:333–342. https://doi.org/10.1042/bj1720333 2. Pebay-Peyroula E, Brandolin G (2004) Nucleotide exchange in mitochondria: insight at a molecular level. Curr Opin Struct Biol 14:420–425. https://doi.org/10.1016/j.sbi.2004.06.009 3. Vignais PV, Lunardi J (1985) Chemical probes of the mitochondrial ATP synthesis and translocation. Annu Rev Biochem 54:977–1014. https://doi.org/10.1146/annurev.bi.54.070185.004553 4. Knoller S, Shpungin S, Pick E (1991) The membrane-associated component of the amphiphile-activated, cytosol-dependent superoxide-forming NADPH oxidase of macrophages is identical to cytochrome b559. J Biol Chem 266:2795–2804. https://doi.org/ 10.1016/S0021-9258(18)49917-6 5. Bromberg Y, Pick E (1985) Activation of NADPH-dependent superoxide production in a cell-free system by sodium dodecyl sulfate. J Biol Chem 260:13539–13545. https://doi.org/10.1016/S0021-9258 (17)38756-2 6. Heyneman RA, Vercauteren RE (1984) Activation of a NADPH oxidase from horse polymorphonuclear leukocytes in a cell-free system. J Leukoc Biol 36:751–759. https://doi.org/10.1002/jlb.36. 6.751

121 7. Ligeti E, Doussiere J, Vignais PV (1988) Activation of the O2-generating oxidase in plasma membrane from bovine polymorphonuclear neutrophils by arachidonic acid, a cytosolic factor of protein nature, and nonhydrolyzable analogues of GTP. Biochemistry 27:193–200. https://doi.org/10.1021/bi00401a029 8. Volpp BD, Nauseef WM, Donelson JE et al (1989) Cloning of the cDNA and functional expression of the 47-kilodalton cytosolic component of human neutrophil respiratory burst oxidase. Proc Natl Acad Sci USA 86:7195–7199. https://doi.org/10.1073/pnas. 86.18.7195 9. Leto TL, Lomax KJ, Volpp BD et al (1990) Cloning of a 67-kD neutrophil oxidase factor with similarity to a noncatalytic region of p60c-src. Science 248:727–730. https://doi.org/10.1126/science. 1692159 10. Tauber AI (1987) Protein kinase C and the activation of the human neutrophil NADPH-oxidase. Blood 69:711–720. https://doi.org/10. 1182/blood.V69.3.711.711 11. Seifert R, Schultz G (1987) Fatty-acid-induced activation of NADPH oxidase in plasma membranes of human neutrophils depends on neutrophil cytosol and is potentiated by stable guanine nucleotides. Eur J Biochem 162:563–569. https://doi.org/10.1111/j. 1432-1033.1987.tb10676.x 12. Bourmeyster N, Stasia MJ, Garin J et al (1992) Copurification of rho protein and the rho-GDP dissociation inhibitor from bovine neutrophil cytosol. Effect of phosphoinositides on rho ADP-ribosylation by the C3 exoenzyme of clostridium botulinum. Biochemistry 31: 12863–12869. https://doi.org/10.1021/bi00166a022 13. Boulay F, Tardif M, Brouchon L, Vignais P (1990) Synthesis and use of a novel N-formyl peptide derivative to isolate a human N-formyl peptide receptor cDNA. Biochem Biophys Res Commun 168:1103–1109. https://doi.org/10.1016/0006-291x(90)91143-g 14. Boulay F, Mery L, Tardif M et al (1991) Expression cloning of a receptor for C5a anaphylatoxin on differentiated HL-60 cells. Biochemistry 30:2993–2999. https://doi.org/10.1021/bi00226a002 15. Fuchs A, Dagher MC, Jouan A, Vignais PV (1994) Activation of the O2--generating NADPH oxidase in a semi-recombinant cell-free system. Assessment of the function of Rac in the activation process. Eur J Biochem 226:587–595. https://doi.org/10.1111/j.1432-1033. 1994.tb20084.x 16. Grizot S, Fauré J, Fieschi F et al (2001) Crystal structure of the Rac1-RhoGDI complex involved in nadph oxidase activation. Biochemistry 40:10007–10013. https://doi.org/10.1021/bi010288k 17. Grizot S, Fieschi F, Dagher MC, Pebay-Peyroula E (2001) The active N-terminal region of p67phox. Structure at 1.8 A resolution and biochemical characterizations of the A128V mutant implicated in chronic granulomatous disease. J Biol Chem 276:21627–21631. https://doi.org/10.1074/jbc.M100893200 18. Massenet C, Chenavas S, Cohen-Addad C et al (2005) Effects of p47phox C terminus phosphorylations on binding interactions with p40phox and p67phox. Structural and functional comparison of p40phox and p67phox SH3 domains. J Biol Chem 280:13752– 13761. https://doi.org/10.1074/jbc.M412897200 19. Fuchs A, Dagher MC, Vignais PV (1995) Mapping the domains of interaction of p40phox with both p47phox and p67phox of the neutrophil oxidase complex using the two-hybrid system. J Biol Chem 270:5695–5697. https://doi.org/10.1074/jbc.270.11.5695 20. Elsen S, Doussière J, Villiers CL et al (2004) Cryptic O2-generating NADPH oxidase in dendritic cells. J Cell Sci 117: 2215–2226. https://doi.org/10.1242/jcs.01085 21. Cohen-Tanugi L, Morel F, Pilloud-Dagher MC et al (1991) Activation of O2--generating oxidase in an heterologous cell-free system derived from Epstein-Barr-virus-transformed human B lymphocytes and bovine neutrophils. Application to the study of defects in cytosolic factors in chronic granulomatous disease. Eur J Biochem 202:649–655. https://doi.org/10.1111/j.1432-1033.1991.tb16419.x

122 22. Vignais PV, Vignais PM (2005) Science expérimentale et conception du vivant : la méthode et les concepts. EDP Sciences, Les Ulis 23. Vignais PV (2001) La biologie, des origines à nos jours : une histoire des idées et des hommes. EDP Sciences, Les Ulis

M.-C. Dagher 24. Klein G, Satre M (2007) Professor Pierre Vignais, biochemist (1926–2006). Biochimie 89:1039–1041. https://doi.org/10.1016/j. biochi.2007.06.003

8

Gary M. Bokoch, the Rac-n-Rho Man: His Fascination with Rho-GTPases Becky A. Diebold

Abstract

Dr. Gary Bokoch’s legacy (Fig. 8.1) continues in each of us who learned about RhoGTPases through his training, publications, collaborations, and presentations. His expertise on RhoGTPases provided the root for the branches of his studies into fields such as chemotaxis, the actin cytoskeleton, p-21 activated kinases, LIM kinase, and NADPH oxidases. A survey of his 217 publications from his graduate work to his final days as professor clearly showed that the topic of neutrophils was special to him since that is how his successful scientific career began. From working on arachidonic metabolism in guinea pig neutrophils during graduate school, to identifying the Gialpha (Giα)-subunit while a post-doc in the laboratory of Nobel Laureate, Dr. Alfred Gilman, to discovering functions for RhoGTPases in NOX2 in neutrophils, Dr. Bokoch’s dedication to this field has taught us how to approach studies on signaling mechanisms applicable to many areas of science. I hope my presentation of Dr. Bokoch (Gary as I shall address him hereon) will paint a vivid picture of his journeys that led him to the path of NADPH Oxidases. As for the many other scientific paths on which he traveled, they would each require a separate chapter. Keywords

Gary Bokoch · RhoGTPases · Neutrophils · NADPH oxidases

B. A. Diebold (✉) Developmental Therapeutics Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA e-mail: [email protected]

1

Introduction

Even after he passed away in 2010 Gary still speaks to me from time to time. I guess it’s because we spent so much time together in the lab and became good friends (Fig. 8.1). I sense him here now. He is pulling up a chair next to mine, picking up discarded drafts from the floor and crumpling them into paper balls. “So, what’re you working on?” he asks as he shoots the paper balls into my wastebasket. “I’m writing my memoirs of you for Dr. Edgar Pick’s NOX book,” I state matter-of-factly. “You’re writing your memoirs about me? Do you even remember me?” he interrogates in a surprised, familiar tone. I feel him glaring at me, waiting for an answer. I continue typing, ignoring him as I’ve done so many times in the past. “Gary, of course I do!” I say with full affirmation even though I know that my memories of him have been fading. “It all better be good!” he humorously challenges. “So how far did you get?” he asks as he takes control of my computer’s mouse and scrolls through my almost completed draft. “This is rather boring. Make it interesting, something that people will read after the first page,” he directs. “OK, so where do I start? What do you want me to tell them?” I ask in frustration since I’ve worked on this for almost a year. There is a long, silent pause while he thinks. He swivels his chair, staring out the window imagining that he sees the view of the Torrey Pines Golf Course that he had at his office at The Scripps Research Institute in San Diego, California (Fig. 8.2). After a prolonged moment he turns back to me, “Let them know that I want to be remembered. . .remembered as the Rac-n-Rho Man.” With that, a year’s worth of my work went into the recycle bin, and I started afresh. Gary, himself was a GTPase (Fig. 8.3). Figure 8.3 introduces you to the Rac-n-Rho Man. Since Gary’s name begins with the letter “G”, I came up with this idea to make a GTPase-cycle about him. The state of being bound by the GDP dissociation inhibitor (GDI) was very transient for Gary because he had excellent grant-writing skills that could quickly release him from GDI (Grant Decisions of

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_8

123

124

B. A. Diebold

Insufficient funding). In fact, guanine nucleotide exchange factors (GEFs, Generous Extramural Funding) were abundantly associated with Gary allowing him to bind GTP (Graduate students, Technicians, and Post-docs) and convert GTP to GDP (Grants, Dissertations, and Publications). In the GTP-bound active state, Gary found himself docked onto GTPase binding motifs of effectors (collaborators), thereby tremendously increasing his and their productivity. Eventually, as with all GTPases, the GAPs (Graduations and Postdoc departures) exhausted his energy supply, converting him back temporarily to the resting state. Although, it is possible that he possessed a point mutation like Q61L that made him constitutively active.

2

Fig. 8.1 Dr. Gary M. Bokoch, 1954–2010, the Rac-n-Rho Man

Gary’s Early Career

Gary grew up in Erie, Pennsylvania and obtained his bachelor’s degree with high honors from Penn State University in 1976. He was accepted into the pharmacology graduate program at Vanderbilt University and was mentored by Dr. Peter W. Reed. Dr. Reed had discovered the calcium ionophore, A23187, in 1972 and was studying its effects on mitochondrial function, prostaglandin biosynthesis, and thromboxane biosynthesis in platelets [1–4]. Gary’s first publication showed that a certain cyclooxygenase and lipoxygenase inhibitor, 5,8,11,14-eicosatetraynoic acid (ETYA), inhibited the respiratory burst, degranulation, and release of arachidonate metabolites elicited by the bacterial

Fig. 8.2 Dr. Bokoch in his office at The Scripps Research Institute. Gary had a beautiful view of the Torrey Pines Golf Course in San Diego from his window. The sliding door (behind the closed vertical blinds in this photo) led to a patio where we all could enjoy the fresh air of the Pacific Ocean

8

Gary M. Bokoch, the Rac-n-Rho Man: His Fascination with Rho-GTPases

Fig. 8.3 The Gary Cycle. Gary was like an active GTPase, who produced more than 200 publications and launched many trainees into successful careers

125

GDP:

GEF

GTP

Grants Dissertations Publications

Generous, Extramural Funding

Graduate students Technicians Post-docs

Effectors RhoGDI

Grant Decisions Insuffient funding

GDP

Gary, the GTPase

GAP Graduations And Post-doc departures

peptide, N-formylmethionyl-leucyl-phenylalanine (fMLF) in guinea pig peritoneal leukocytes. This pointed to the importance of arachidonic acid metabolism in generating superoxide production. Gary used a competitive peptide blocker of the N-formyl peptide receptor and demonstrated that arachidonic acid metabolism was a formyl peptide receptormediated event [5, 6]. After receiving a National Science Foundation pre-doctoral fellowship, Gary continued to work on the isolation and identification of several lipoxygenase metabolites released by stimulated leukocytes. He tested each metabolite for its relative ability to induce degranulation, concluding that leukotriene B4 played a significant role in degranulation [7]. In his third Journal of Biological Chemistry publication he continued to study the effects of ETYA on inhibition of 5-lipoxygenase showing that ETYA can inhibit the enzymatic step whereby 5-hydroperoxy-6,8,11,14-eicosatetraenoic acid (5-HETE) is converted to leukotriene A4 in guinea pig polymorphonuclear leukocytes [8]. The biochemical and pharmacological training that he received during this time on the use of enzyme inhibitors and receptor antagonists proved to be very useful tools that he applied throughout his career. Gary’s work on receptor-mediated activation of arachidonic acid metabolism in neutrophils segued to a postdoctoral position in the laboratory of Nobel laureate, Alfred G. Gilman. In 1982, Katada and Ui in Japan had identified a 41-kD membrane protein in rat C6 glioma cell membranes that was ADP-ribosylated by pertussis toxin. The ribosylation was associated with an increase in adenylate cyclase activity in the presence of GTP suggesting that the 41-kD protein might be a G protein subunit [9, 10]. The following year Gary, Dr. Katada, and their colleagues worked

GTP

Publications 217: Leukocytes Arachidonic acid metabolism Trimeric G-proteins NOX 2 Chemotaxis NOX1 PAK (p21-activated kinase) LIM Kinase GEFH1 RhoGDI

Cofilin Chronophin

on purifying the 41-kD membrane protein from rabbit liver membranes. Gary demonstrated that this protein had a specific binding site for guanine nucleotides suggesting that this protein, now known as Giα, was a subunit of the trimeric G-proteins. A flurry of four Journal of Biological Chemistry publications appeared in 1984 further characterizing the effects of Giα activity on adenylate cyclase activity [11– 14]. What is even more impressive is that in the same year Gary had a publication in Cell, resulting from his application of his knowledge of pertussis toxin to his beloved topic of neutrophils. In that work, he found that pertussis toxin abrogated fMLF-stimulated release of arachidonic acid and attenuated granular enzyme release in a time- and concentration-dependent manner in guinea pig neutrophils. Similar to his work with Giα from rabbit liver, he showed that the substrate of pertussis toxin was a single membrane protein in neutrophils with a molecular weight similar to Giα in rabbit liver [15]. In 1985, Gary was recruited by the Department of Immunology at The Scripps Clinic and Research Foundation, now known as The Scripps Research Institute in La Jolla, California. At that time there were many notable researchers working on different aspects of neutrophil or macrophage function. Within the Department of Immunology, Dr. Charles Cochrane, Dr. Algirdas Jesaitis, Dr. Charles Parkos, and Dr. Larry Sklar were working on topics such as N-formyl peptide receptor occupancy dynamics, potentiation of fMLF-induced superoxide production by cytochalasin B, and localization of cytochrome b558 (cytb558) in human neutrophils [16–18]. In addition, Dr. Richard Ulevitch was working on phospholipases and prostaglandin biosynthesis in macrophages [19]. In the Department of Experimental

126

B. A. Diebold

Medicine, Dr. Bernard Babior, Dr. John Curnutte, Dr. Paul Heyworth, and Dr. Andrew Cross were working on the respiratory burst of neutrophils. Thus, Gary was very welcomed at Scripps. Immediately, Gary began to work with human neutrophils comparing adenylate cyclase activity in those cells to that of human platelets [20]. He described the involvement of a trimeric G protein in fMLF-receptor-mediated human neutrophil activation [21]. This led to his authorship on one of the first comprehensive chapters on signal transduction and cytoskeletal activation in the neutrophil [22]. Between 1985 and 1988, Gary continued to work on the localization of the various trimeric G protein subunits in neutrophils, their quantification, and characterization of Gn (the major neutrophil G protein subunit that is ADP-ribosylated) [23].

3

The Hunt for Low Molecular Weight G-Proteins in Neutrophils and HL-60 Cells

Around the mid-1980s Gary’s attention was turned to several reports stating the existence of 21-kDa proteins in different cell types that were able to activate phospholipase activity and exocytosis, and possessed sensitivity to ADP ribosylation by type D botulinum toxin (and not pertussis toxin) [24–26]. These reports intrigued Gary and led him and Dr. Charles Parkos to search for proteins, other than Gn, in human neutrophils that could bind radioactively labeled GTP ([α-32P]GTP). They found three new GTP-binding proteins having molecular weights of 22-, 24-, and 26-kD. The 22-Da protein appeared to be related to the ras proteins because it comigrated with H-ras on SDS-polyacrylamide gels and cross-reacted with a monoclonal antibody directed against the GTP-binding domain of known ras proteins. Furthermore, this new 22-kD GTP-binding protein served as a substrate for ADP-ribosylation by type D botulinum toxin suggesting it might have a role in exocytosis or other signaling pathways [27, 28]. To discover new low molecular weight G-proteins in other systems, Dr. Ralph Snyderman’s group at Genentech, with Gary as collaborator, screened a phage cDNA library from differentiated HL-60 cells using an oligonucleotide probe consisting of the sequence of the most conserved domain of all ras gene sequences known at that time. Using molecular biology methods, they were able to obtain the cDNA clones of two closely related genes. With Gary’s collaboration, the cDNAs were transfected into COS cells, and the two proteins were shown to be targets of ADP-ribosylation by C3 botulinum toxin, strongly supporting that these proteins were a new class of low molecular weight G proteins. They named these proteins, Rac1 and Rac2 (ras-related C3 botulinum toxin substrate). [29].

Despite the cloning of the rac1 and rac2 genes from HL-60 cells, it was still unknown whether a low molecular weight G-protein was involved in directly activating the respiratory burst. It had been reported by several groups that non-hydrolyzable analogs of GTP could increase oxidase activity [30–32], and the protein mediating this GTP-dependent activation appeared to be cytosolic. Therefore, one of Gary’s first postdocs, Dr. Ulla Knaus, fractionated the cytosolic fraction of human neutrophils and tested each fraction for its ability to bind radioactive GTP and for its ability to augment NOX2 activity as measured using the cell-free system that had been developed by Dr. Yael Bromberg and Dr. Edgar Pick [33, 34]. The active fraction was purified further, and immunoblot assays revealed that the purified protein did not react to antibodies developed for other known low molecular weight GTP-binding proteins (ras, rap1A, rho). However, the newly purified 22-kD cytosolic factor did react with the Rac2 antibody developed for the Rac2 protein of HL-60 cells. This novel finding was reported in Science at a time when very competitive research was being conducted to identify additional protein components of the NOX2 enzyme complex [35]. Dr. Edgar Pick and Dr. Arie Abo subjected a fraction of guinea pig macrophage cytosol derived by ammonium sulfate precipitation, known as σ1 [36], to purification to near homogeneity and found it to be a heterodimer of two proteins of 22 and 24 kDa [37]. Amino acid sequencing of the purified heterodimer identified the two proteins as Rac1 and RhoGDI, and the component required for NOX2 activation in a cell free system was shown to be Rac1 [38]. Ulla Knaus et al. reported further work on the purification and characterization of Rac2 from human neutrophils [39], the regulation of Rac2 by RhoGDI and RacGAP, the isoprenylation of Rac2, and the translocation of Rac2 to the neutrophil membrane [40–45].

4

The Balancing of Projects

One of Gary’s strongest assets was his amazing ability to keep several projects boiling hot at the same time. During 1990, he examined the nucleotide binding properties of Rap1a that he had purified from neutrophils [46], and then in 1991 Gary and his colleagues published their findings on the interaction of Rap1A with cytb558 in Science [47]. Dr. Mark Quinn, the senior author on that publication, and Gary were best friends at Scripps. I would venture to say that Mark was a friend as close as a brother to Gary based on their interactions that I witnessed in later years when I met Mark via Gary at conferences. In1994 a report in Nature identified a 65-kD (p65) serine/ threonine kinase (p21-activated kinase, PAK) in brain as an effector protein for Rac1 and Cdc42 GTPases using an overlay assay. Binding of active Rac1-GTP or Cdc42-GTP to the

8

Gary M. Bokoch, the Rac-n-Rho Man: His Fascination with Rho-GTPases

CRIB (Cdc42/Rac interactive binding) motif of PAK activates this kinase and leads to many downstream signaling events. Binding of active Rac or Cdc42 also induces autophosphorylation of PAK, resulting in the release of active Rac1 or Cdc42 and their freedom to bind other effector proteins [48]. Gary and Ulla must have heard about this kinase much earlier because by 1995 they reported in Science that they had identified PAK in neutrophils. They demonstrated that fMLF-receptor activation of neutrophils led to Rac2 and Cdc42 activation, and PAK-induced phosphorylation of 47phox [49]. Around the mid 1990’s Ulla Knaus was promoted to assistant professor at Scripps, but she continued to work with Gary. Together they developed a very successful research program on the topic of PAKs and eventually other kinases. While many members of Gary’s lab began working on PAKs, a few continued to work on neutrophils. One of them was Dr. Valerie Benard, a postdoc from France, who in 1999 published her work on the first pulldown assay for active Rac and Cdc42 using fMLF-stimulated neutrophils [50]. In this assay the p21(Rac/Cdc42)-binding domain (CRIB motif) of PAK1 was expressed in E.coli as a glutathione-S-transferase fusion protein bound to sepharose beads. The assay was first demonstrated to work using stimulated human neutrophils, but this assay also works in many cell types and has since been used by scientists worldwide as a commercial kit to determine if Rac and Cdc42 are activated in various cell types. Gary’s lab eventually designed as assay to detect active Rho as well [51].

5

My Story

I joined Gary’s lab in 1998. Whereas most post-docs do some amount of research before choosing a postdoc mentor, I did not. I did not know that Gary was an expert in RhoGTPases, that he studied under the Nobel Laureate Alfred Gilman, that he worked later and harder than his postdocs. Five years had passed since I graduated with my PhD because I decided to be a full-time mom during those five years. After those 5 years I traveled from Hawaii to attend a huge FASEB conference in Los Angeles to search for a postdoc position. Gary had a postdoc advertisement on a bulletin board, and I put my contact information in a file under his name. The next day at the conference, I had a hard copy message that I was scheduled for an interview with Gary. The interview was awkward. We introduced ourselves, but Gary made very little eye contact. He seemed more nervous than I as he explained that he was looking for a postdoc to work on neutrophils and NADPH oxidases. He seemed particularly interested in my background in mitochondrial electron transport and was impressed that I knew how to fractionate neutrophils and perform the cell-free assay. He asked if I could come down

127

to San Diego and see the lab that very week. Within the following 3 months, I sold my house in Honolulu, custombuilt by my father, pulled my husband out of his secure job, and tore my toddlers from their grandparents’ arms just so that I could work in Gary’s lab. (I have moved spontaneously twice more to continue work on NOX in other notable labs). When I started to work in Gary’s lab in 1998, I felt like Mr. Bean (the film actor, Rowan Atkinson), mistakenly being offered a postdoc position in a fast-paced lab in a well-known institute, The Scripps Research Institute. I knew nothing about RhoGTPases or purifying recombinant proteins or kinases or cell signaling. I only knew how to purify neutrophils and reconstitute the cytosol and membranes to measure superoxide production. I would not have survived in Gary’s lab had it not been for the well-organized lab and training infrastructure that he had established. He was a hands-on type of principal investigator who knew what was in every refrigerator and freezer box. He had binders with lists of every plasmid and mutation created in his lab (or the name of the person who donated it). The lists also stated the exact location: name of the freezer, box number, and row number. Gary was a Star Trek fan, so all the freezers were named after Star Trek characters. Everyone in the lab was assigned responsibilities such as preparing competent cells, buffers, etc. Gary still performed many experiments himself. Touching anything on Gary’s bench was tabooed, an instant death wish. At any given time, Gary had 2–3 well-trained technicians. Each one skilled in a certain area such as protein purification, molecular cloning, or tissue culture. Ben Bohl was Gary’s first technician, and he stayed with Gary until the end. Ben taught me how to purify neutrophils from large quantities of freshly donated human blood (a unit or more) that we picked up from Scripps Clinic, and he trained me to prepare neutrophil membrane and cytosolic fractions from these large quantities in a single day. Many late nights were spent in in the coldroom. Luckily for Ben, I was teachable, a fastlearner, and willing to work late to get the job done. Additionally, Ulla taught me how to use the FPLC to further purify recombinant proteins. Ben and Ulla taught me almost everything I needed to know to get my project started, but there was one thing that Gary insisted on teaching me himself; he wanted to teach me how to correctly load recombinant Rac protein with GTPγS. He placed 8.1 μL of 0.1 M EDTA on the bottom of a small glass tube, which he held up close to his face to observe the tiny drop. Then he carefully turned the tube on its side and placed 12.0 μL of recombinant Rac2 on the side of the tube to prevent pre-mature stripping of any pre-bound nucleotide. Lastly, he added 8.1 μL of 10 mM GTPγS while immediately tapping the tube on the bench to allow all three components to come together simultaneously. (I still have my lab book containing my notes from that day.) After incubating the mixture at 30 °C for 10 min, he added

128

B. A. Diebold

2 μL of 1.0 M MgCl2 to lock-in the bound nucleotide. “That’s how you load Rac!” he exclaimed with a beaming smile of a child, proudly showing others what he could do well.

6

The Difficult Project

Most of the projects that Gary gave his postdocs were reasonable. My project, in contrast, was rather difficult because it required biophysical methods that no one in his lab had any experience with. I say this because there wasn’t a doublebeamed spectrophotometer or fluorometer present in this lab. At first Gary wanted to see if I could detect electron transfer from NADPH to FAD using spectroscopy using neutrophil cytosol, a crude membrane fraction, GTPγS, FAD, and SDS. Using an old spectrophotometer in another lab, I struggled for a few months with this project. In addition, the composition of the reaction was very crude, and the background noise prevented me from seeing any change in the difference spectra of reduced vs. oxidized cytb558. We eventually used recombinant proteins and partially purified cytb558 from human neutrophil membranes. To determine if Rac2 had a potential role in regulating the first step of electron transfer, we used iodonitrotetrazolium (INT) as an artificial two-electron electron acceptor to measure the diaphorase activity of the system. INT was added to the cell-free system in place of ferricytochrome c to measure this first step (NADPH to FAD/ INT) of the electron transfer reaction, whereas reduction of ferricytochrome c was used to measure the complete electron transfer reaction including Step 2 (FADH2 to O2). We aimed to determine if Rac2 could independently regulate Step1 of the electron transfer pathway. We interpreted our results from those studies as an indication that Rac2 played a role in catalyzing the first step of electron transfer independently from p67phox because mutations in either Rac or p67phox, that prevented interaction between these proteins, still resulted in INT reduction [52]. Of course, our report generated much skepticism because Dr. David Lambeth’s group had shown that Step1 was the rate-limiting step and required the activation by p67phox [53]. We had also reported in that publication that the insert domain of Rac2 was necessary for the interaction of Rac with cytb558, but publications by other investigators [54] later showed that the insert domain of Rac is dispensable for its interaction with cytb558, in contrast to what we reported. Thus, it appears that the insert domain is not necessary (see ref. [55] for detailed review by Dr. Pick). We had also carried out additional experiments indicating that Rac2 could interact directly with cytb558 using a fluorescence assay based on the fluorescence of methylanthraniloyl guanosine-5′-[beta,gamma-imido]triphosphate (mantGppNHhp)-loaded Rac used by Dr. Lambeth’s group to

detect the interaction of p67phox with cytb558 [56]. We detected a shift of the fluorescence spectrum of Rac-mantGppNHhp when cytb558 was added. However, the data supporting this conclusion was also met with criticism due to marginal (10% changes) in the shift of the spectra. Several years later another postdoc in Gary’s lab mapped the region for the interaction between Rac2 and cytochrome b558 [57]. The main story here, is that Gary, as the senior author, handled the criticisms on his own and always supported my efforts on this difficult project. The subject of the role of Rac in regulation of NADPH Oxidases is covered in another chapter of this book (see Chap. 18 by Y. Lin and Y. Zheng).

7

New Beginnings: NOX1

When the new NOX family member, NOX1, was discovered in 1999 [58], I began to take interest in studying it and its regulation. At that time Gary was not enthusiastic about working on NOX1, so I started to develop preliminary data for a National Scientist Development Grant application from the American Heart Association on regulation of NOXA1 by phosphorylation and 14-3-3 proteins. After receiving the funding for the grant, I decided to move to Emory University because Dr. David Lambeth was there as well as Dr. Haian Fu, an expert on 14-3-3 proteins. My decision to move to Emory was a big shock to Gary, but he held back his anger as best he could in front of me. After a few years he forgave me, and we collaborated as co-senior authors to get the story about NOXA1 phosphorylation and 14-3-3 proteins published [59]. I was very surprised a year later when he called me to tell me that he was thinking of moving to Emory. He asked me if I would re-join his lab if he did. Thankfully, I never had to make that decision because his wife decided that she did not want to move. Soon thereafter Gary did take an interest in NOX1. His interest in NOX1 might have been ignited from reports about the Src protein kinase substrates, Tks5 and others, which had been identified from a phage cDNA library screen for new tyrosine kinase substrates (Tks) of Src back in 1998 at SUGEN, Inc. in California [60]. One of the Tks proteins identified wasTks5, a scaffolding protein resembling p47phox in structure, having a N-terminal PX domain and 5 SH3 domains. Dr. Sarah Courtneidge, a member of that group, continued to characterize Tks5 at the Van Andel Research Institute in Michigan where she discovered that it was involved in podosome formation. She eventually moved to the Burnham Institute across the street from Scripps. Dr. Courtneidge and Gary formed a collaboration to further understand the role of Tks4 and Tks5 proteins in invadopodia formation in cancer cells. In back-to-back articles in Science Signaling, Dr. Courtneidge and Gary reported their findings separately. Dr. Courtneidge’s group

8

Gary M. Bokoch, the Rac-n-Rho Man: His Fascination with Rho-GTPases

found that invadopodia formation was dependent upon NOX4-generated ROS which in turn was dependent on Tks5 phosphorylation by Src in C8161.9 melanoma cells [61]. Gary’s group reported that Tks4 and Tks5 were new organizing proteins that were required for NOX1-generated ROS that was required for invadopodia formation in DLD-1 colon carcinoma cells [62, 63]. Shortly before he became fatally ill, Gary in collaboration with Dr. Hugh Rosen performed a high throughput screen for NOX1 inhibitors with the help of The Scripps Research Institute Molecular Screening Center, part of the Molecular Libraries Probe Production Centers Network. From this screen, they identified the phenothiazine compound, ML-171, which was reported (in the year of his death) as a specific NOX1 inhibitor that was capable of inhibiting invadopodia formation in the colon adenocarcinoma cell line, DLD-1 [64, 65]. At that time, the pitfalls associated with screening for NOX inhibitors were only beginning to be addressed.

8

The Continuation of Gary’s Lab by Dr. Céline DerMardirossian

Even before Gary became very ill, he realized that he needed someone to carry-on the works of his laboratory. He wanted Dr. Céline DerMardirossian and me to share the future responsibilities of his laboratory. Dr. DerMardirossian would take on the cell biology aspects, and I would direct the work on NOX. Later, when I chose to leave the lab, Gary prepared Dr. DerMardirossian to take over his lab. Earlier, as a postdoc she had published several papers on RhoGDI, demonstrating that it was phosphorylated by PAK and by Src kinase [66–71]. She was also an author on the article about cofilin regulation by PAK [71, 72]. As Gary became weaker in 2009, Dr. DerMardirossian prepared herself to sustain Gary’s lab. With prolonged sorrow after his death, she successfully led the lab to complete and publish the remaining projects. She had both the support of the remaining postdocs and the tremendous help of Gary’s wife, who knew everything about Gary’s office.

9

Anecdotes About Gary

It has been a year since I began working on my essay about Gary. Each season brought back memories of the fun we had in Gary’s lab. The full tennis courts of summer reminded me of how Gary loved to play tennis. Quite often, he played tennis with members of our lab in the late afternoons. Gary was quite serious about tennis, and there were excellent tennis players among our postdocs from Germany, the UK, and France. I never saw Gary get so upset as he did whenever

129

he lost a match to his lab members. A lot of them thought they would surely get fired by winning a tennis match against Gary because he did threaten to fire them right after a game. However, by the following morning it was business as usual. In addition to tennis, Gary organized our hikes, kayaking, whale-watching trips, and picnicking. We were so privileged to live in San Diego where we were so close to the beautiful beaches of La Jolla and Del Mar. As the new school year began, I recalled that Gary had a very closely-knit family. Jan Bokoch supported Gary on many levels. In the laboratory, she kept Gary’s office and library of journals organized including reference card files. She and Gary were very involved in their church and their children’s schools. Every year, Gary looked forward to volunteering to do fun science demonstrations at his daughters’ schools. He often asked a few of us to assist him. One year I assisted Gary as he taught the children about cell organelles. Jan had baked small cakes for the children to decorate with candy as organelles. Gary and Jan also treated the lab members as an extension of their family, inviting us to see school plays or other performances by their daughters. Here in the US, the stores already have displays for the upcoming holidays which remind me of the holiday celebrations in Gary’s lab. On Halloween, Gary enjoyed scaring us. He arrived earlier than usual and placed frightening objects in hidden places in the lab. As examples, he placed a rubber mold of a decapitated human head on a platter of fake blood in the refrigerator, a dead rubber rat in the centrifuge, and other “creatures” in drawers and closets of the lab. He laughed with delight while sitting in his office as he heard screams coming from his lab. He also brought about 15 pumpkins to the lab each year for a pumpkin carving party, a tradition that many of us have kept in our current labs. Each Thanksgiving I am reminded of Gary’s kindnesses. One year my family and I were invited to join him and his family for Thanksgiving dinner. Knowing that we were a young family and not earning a lot, Gary suggested to his wife that she should check if they had some clothes and toys that his daughters (Jenny and Becca) had outgrown. That was so thoughtful that I still remember that evening vividly. There were many other occasions where Gary did little things that were very kind such as buying pizza for us when we were working late purifying proteins in the coldroom. Once as he was leaving for the night, he even offered me his lunch that he had not had time to eat. Every Christmas season, I am reminded of how Gary and Jan invited everyone in the lab and their families to their home for a huge party each year. Gary loved to make children happy, so he had one of his friends dress up as Santa. Days before the party, Gary asked all the parents to purchase something that Santa could give their child at the party. The

130

B. A. Diebold

Fig. 8.5 Gary’s poem on my lab coat. It took Gary several days to decide what to write as an autograph on my lab coat as I prepared to leave his lab. This lab coat is one of my most prized possessions Fig. 8.4 Gary on Santa’s lap. Gary was the last one to sit on Santa’s lap at the annual Christmas party at his home. Santa always gave him a bag of coal to make the children laugh

gifts were placed in Santa’s bag before the party started. When all the children had received their gifts from Santa, Gary would be the last one to sit in Santa’s lap. The gift was usually a bag of coal to make all the children laugh (Fig. 8.4). There were also snowball fights in the backyard using white, knotted socks as snowballs since it doesn’t snow in San Diego. To this day my children, now adults, still remember those snowball fights and what they received from Santa at those parties.

10

Conclusion

When I was completing this manuscript past bedtime one night, Gary’s spirit visited me again. “You’re still working on this! What has it been, over a year now?” he sighed with disbelief. He grabbed my computer’s mouse, and as he read it, he chuckled and even laughed out loud at some parts. “You

should change this, and you should change that,” he advised. “Too late, Gary, I’ve just submitted it, and if you don’t mind, I would like to go to bed now. I have a busy day tomorrow,” I said wearily while rubbing my eyes from exhaustion. “You’re busy? What are you working on?” he asked. “NOXs of course,” I replied. “You’re not working with RhoGTPases? Remember what I wrote on your lab coat when you left my lab?” he inquired disappointedly. “Yes, PHOX’s are red, NOX’s make you blue. Give me RhoGTPases, man! They will Rac-n-Rho you!” I recited. “That’s amazing man, you remembered that poem!” he exclaimed with delight. “Yes, Gary. You see, I do remember you.” I pointed to my old lab coat hanging in my room for 17 years (Fig. 8.5), then I fell asleep while hearing him still suggesting experiments that I should try. Acknowledgments I thank all the past members and collaborators of Dr. Bokoch’s laboratory, as well as his immediately family, for the wonderful memories.

Disclaimer The content of this chapter does not reflect the views or policies of the U.S. Department of Health and Human Services, nor does

8

Gary M. Bokoch, the Rac-n-Rho Man: His Fascination with Rho-GTPases

mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

References 1. Reed PW, Lardy HA (1972) A23187: a divalent cation ionophore. J Biol Chem 247:6970–6977 2. Knapp HR, Oelz O, Roberts LJ, Sweetman BJ, Oates JA, Reed PW (1977) Ionophores stimulate prostaglandin and thromboxane biosynthesis. Proc Natl Acad Sci U S A 74:4251–4255 3. Oelz O, Knapp HR, Roberts LJ, Oelz R, Sweetman BJ, Oates JA, Reed PW (1978) Calcium-dependent stimulation of thromboxane and prostaglandin biosynthesis by ionophores. Adv Prostaglandin Thromboxane Res 3:147–158 4. Reed PW, Knapp HR (1978) Prostaglandins and calcium. Ann N Y Acad Sci 307:445–447 5. Bokoch GM, Reed PW (1979) Inhibition of the neutrophil oxidative response to a chemotactic peptide by inhibitors of arachidonic acid oxygenation. Biochem Biophys Res Commun 90:481–487 6. Bokoch GM, Reed PW (1980) Stimulation of arachidonic acid metabolism in the polymorphonuclear leukocyte by an N-formylated peptide. Comparison with ionophore A23187. J Biol Chem 255:10223–10226 7. Bokoch GM, Reed PW (1981) Effect of various lipoxygenase metabolites of arachidonic acid on degranulation of polymorphonuclear leukocytes. J Biol Chem 256:5317–5320 8. Bokoch GM, Reed PW (1981) Evidence for inhibition of leukotriene A4 synthesis by 5,8,11,14-eicosatetraynoic acid in guinea pig polymorphonuclear leukocytes. J Biol Chem 256:4156–4159 9. Katada T, Ui M (1982) Direct modification of the membrane adenylate cyclase system by islet-activating protein due to ADP-ribosylation of a membrane protein. Proc Natl Acad Sci U S A 79:3129–3133 10. Katada T, Ui M (1982) ADP ribosylation of the specific membrane protein of C6 cells by islet-activating protein associated with modification of adenylate cyclase activity. J Biol Chem 257:7210–7216 11. Bokoch GM, Katada T, Northup JK, Ui M, Gilman AG (1984) Purification and properties of the inhibitory guanine nucleotidebinding regulatory component of adenylate cyclase. J Biol Chem 259:3560–3567 12. Katada T, Bokoch GM, Northup JK, Ui M, Gilman AG (1984) The inhibitory guanine nucleotide-binding regulatory component of adenylate cyclase. Properties and function of the purified protein. J Biol Chem 259:3568–3577 13. Katada T, Bokoch GM, Smigel MD, Ui M, Gilman AG (1984) The inhibitory guanine nucleotide-binding regulatory component of adenylate cyclase. Subunit dissociation and the inhibition of adenylate cyclase in S49 lymphoma cyc- and wild type membranes. J Biol Chem 259:3586–3595 14. Katada T, Northup JK, Bokoch GM, Ui M, Gilman AG (1984) The inhibitory guanine nucleotide-binding regulatory component of adenylate cyclase. Subunit dissociation and guanine nucleotidedependent hormonal inhibition. J Biol Chem 259:3578–3585 15. Bokoch GM, Gilman AG (1984) Inhibition of receptor-mediated release of arachidonic acid by pertussis toxin. Cell 39:301–308 16. Jesaitis AJ, Tolley JO, Painter RG, Sklar LA, Cochrane CG (1985) Membrane-cytoskeleton interactions and the regulation of chemotactic peptide-induced activation of human granulocytes: the effects of dihydrocytochalasin B. J Cell Biochem 27:241–253 17. Parkos CA, Cochrane CG, Schmitt M, Jesaitis AJ (1985) Regulation of the oxidative response of human granulocytes to chemoattractants. No evidence for stimulated traffic of redox enzymes between endo and plasma membranes. J Biol Chem 260: 6541–6547

131

18. Sklar LA, Hyslop PA, Oades ZG, Omann GM, Jesaitis AJ, Painter RG, Cochrane CG (1985) Signal transduction and ligand-receptor dynamics in the human neutrophil. Transient responses and occupancy-response relations at the formyl peptide receptor. J Biol Chem 260:11461–11467 19. Ross MI, Deems RA, Jesaitis AJ, Dennis EA, Ulevitch RJ (1985) Phospholipase activities of the P388D1 macrophage-like cell line. Arch Biochem Biophys 238:247–258 20. Bokoch GM (1987) The presence of free G protein beta/gamma subunits in human neutrophils results in suppression of adenylate cyclase activity. J Biol Chem 262:589–594 21. Bokoch GM, Sklar LA, Smolen JE (1987) Guanine nucleotide regulatory proteins as transducers of receptor-stimulated neutrophil activation. Int J Tissue React 9:285–293 22. Omann GM, Allen RA, Bokoch GM, Painter RG, Traynor AE, Sklar LA (1987) Signal transduction and cytoskeletal activation in the neutrophil. Physiol Rev 67:285–322 23. Bokoch GM, Bickford K, Bohl BP (1988) Subcellular localization and quantitation of the major neutrophil pertussis toxin substrate, Gn. J Cell Biol 106:1927–1936 24. Bar-Sagi D, Feramisco JR (1986) Induction of membrane ruffling and fluid-phase pinocytosis in quiescent fibroblasts by ras proteins. Science 233:1061–1068 25. Burgoyne RD (1987) G proteins: control of exocytosis. Nature 328: 112–113 26. Ohashi Y, Narumiya S (1987) ADP-ribosylation of a Mr 21,000 membrane protein by type D botulinum toxin. J Biol Chem 262: 1430–1433 27. Bokoch GM, Parkos CA (1988) Identification of novel GTP-binding proteins in the human neutrophil. FEBS Lett 227:66–70 28. Bokoch GM, Parkos CA, Mumby SM (1988) Purification and characterization of the 22,000-dalton GTP-binding protein substrate for ADP-ribosylation by botulinum toxin, G22K. J Biol Chem 263: 16744–16749 29. Didsbury J, Weber RF, Bokoch GM, Evans T, Snyderman R (1989) rac, a novel ras-related family of proteins that are botulinum toxin substrates. J Biol Chem 264:16378–16382 30. Seifert R, Rosenthal W, Schultz G (1986) Guanine nucleotides stimulate NADPH oxidase in membranes of human neutrophils. FEBS Lett 205:161–165 31. Gabig TG, English D, Akard LP, Schell MJ (1987) Regulation of neutrophil NADPH oxidase activation in a cell-free system by guanine nucleotides and fluoride. Evidence for participation of a pertussis and cholera toxin-insensitive G protein. J Biol Chem 262: 1685–1690 32. Ligeti E, Doussiere J, Vignais PV (1988) Activation of the O2(.-)generating oxidase in plasma membrane from bovine polymorphonuclear neutrophils by arachidonic acid, a cytosolic factor of protein nature, and nonhydrolyzable analogues of GTP. Biochemistry 27: 193–200 33. Bromberg Y, Pick E (1984) Unsaturated fatty acids stimulate NADPH-dependent superoxide production by cell-free system derived from macrophages. Cell Immunol 88:213–221 34. Bromberg Y, Pick E (1985) Activation of NADPH-dependent superoxide production in a cell-free system by sodium dodecyl sulfate. J Biol Chem 260:13539–13545 35. Knaus UG, Heyworth PG, Evans T, Curnutte JT, Bokoch GM (1991) Regulation of phagocyte oxygen radical production by the GTP-binding protein Rac 2. Science 254:1512–1515 36. Pick E, Kroizman T, Abo A (1989) Activation of the superoxideforming NADPH oxidase of macrophages requires two cytosolic components--one of them is also present in certain nonphagocytic cells. J Immunol 143:4180–4187 37. Abo A, Pick E (1991) Purification and characterization of a third cytosolic component of the superoxide-generating NADPH oxidase of macrophages. J Biol Chem 266:23577–23585

132 38. Abo A, Pick E, Hall A, Totty N, Teahan CG, Segal AW (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1. Nature 353:668–670 39. Knaus UG, Heyworth PG, Kinsella BT, Curnutte JT, Bokoch GM (1992) Purification and characterization of Rac 2. A cytosolic GTP-binding protein that regulates human neutrophil NADPH oxidase. J Biol Chem 267:23575–23582 40. Chuang TH, Bohl BP, Bokoch GM (1993) Biologically active lipids are regulators of Rac.GDI complexation. J Biol Chem 268:26206– 26211 41. Chuang TH, Xu X, Knaus UG, Hart MJ, Bokoch GM (1993) GDP dissociation inhibitor prevents intrinsic and GTPase activating protein-stimulated GTP hydrolysis by the Rac GTP-binding protein. J Biol Chem 268:775–778 42. Heyworth PG, Knaus UG, Settleman J, Curnutte JT, Bokoch GM (1993) Regulation of NADPH oxidase activity by Rac GTPase activating protein(s). Mol Biol Cell 4:1217–1223 43. Heyworth PG, Knaus UG, Xu X, Uhlinger DJ, Conroy L, Bokoch GM, Curnutte JT (1993) Requirement for posttranslational processing of Rac GTP-binding proteins for activation of human neutrophil NADPH oxidase. Mol Biol Cell 4:261–269 44. Quinn MT, Evans T, Loetterle LR, Jesaitis AJ, Bokoch GM (1993) Translocation of Rac correlates with NADPH oxidase activation. Evidence for equimolar translocation of oxidase components. J Biol Chem 268:20983–20987 45. Dorseuil O, Quinn MT, Bokoch GM (1995) Dissociation of Rac translocation from p47phox/p67phox movements in human neutrophils by tyrosine kinase inhibitors. J Leukoc Biol 58:108–113 46. Quilliam LA, Der CJ, Clark R, O'Rourke EC, Zhang K, McCormick F, Bokoch GM (1990) Biochemical characterization of baculovirus-expressed rap1A/Krev-1 and its regulation by GTPase-activating proteins. Mol Cell Biol 10:2901–2908 47. Bokoch GM, Quilliam LA, Bohl BP, Jesaitis AJ, Quinn MT (1991) Inhibition of Rap1A binding to cytochrome b558 of NADPH oxidase by phosphorylation of Rap1A. Science 254:1794–1796 48. Manser E, Leung T, Salihuddin H, Zhao ZS, Lim L (1994) A brain serine/threonine protein kinase activated by Cdc42 and Rac1. Nature 367:40–46 49. Knaus UG, Morris S, Dong HJ, Chernoff J, Bokoch GM (1995) Regulation of human leukocyte p21-activated kinases through G protein--coupled receptors. Science 269:221–223 50. Benard V, Bohl BP, Bokoch GM (1999) Characterization of rac and cdc42 activation in chemoattractant-stimulated human neutrophils using a novel assay for active GTPases. J Biol Chem 274:13198– 13204 51. Stofega M, DerMardirossian C, Bokoch GM (2006) Affinity-based assay of rho guanosine triphosphatase activation. Methods Mol Biol 332:269–279 52. Diebold BA, Bokoch GM (2001) Molecular basis for Rac2 regulation of phagocyte NADPH oxidase. Nat Immunol 2:211–215 53. Nisimoto Y, Motalebi S, Han CH, Lambeth JD (1999) The p67 (phox) activation domain regulates electron flow from NADPH to flavin in flavocytochrome b(558). J Biol Chem 274:22999–23005 54. Miyano K, Koga H, Minakami R, Sumimoto H (2009) The insert region of the Rac GTPases is dispensable for activation of superoxide-producing NADPH oxidases. Biochem J 422:373–382 55. Pick E (2014) Role of the rho GTPase Rac in the activation of the phagocyte NADPH oxidase: outsourcing a key task. Small GTPases 5:e27952 56. Nisimoto Y, Freeman JL, Motalebi SA, Hirshberg M, Lambeth JD (1997) Rac binding to p67(phox). Structural basis for interactions of

B. A. Diebold the Rac1 effector region and insert region with components of the respiratory burst oxidase. J Biol Chem 272:18834–18841 57. Kao YY, Gianni D, Bohl B, Taylor RM, Bokoch GM (2008) Identification of a conserved Rac-binding site on NADPH oxidases supports a direct GTPase regulatory mechanism. J Biol Chem 283: 12736–12746 58. Suh YA, Arnold RS, Lassegue B, Shi J, Xu X, Sorescu D, Chung AB, Griendling KK, Lambeth JD (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401:79–82 59. Kim JS, Diebold BA, Babior BM, Knaus UG, Bokoch GM (2007) Regulation of Nox1 activity via protein kinase A-mediated phosphorylation of NoxA1 and 14-3-3 binding. J Biol Chem 282:34787– 34800 60. Lock P, Abram CL, Gibson T, Courtneidge SA (1998) A new method for isolating tyrosine kinase substrates used to identify fish, an SH3 and PX domain-containing protein, and Src substrate. EMBO J 17:4346–4357 61. Diaz B, Shani G, Pass I, Anderson D, Quintavalle M, Courtneidge SA (2009) Tks5-dependent, nox-mediated generation of reactive oxygen species is necessary for invadopodia formation. Sci Signal 2:ra53 62. Gianni D, Diaz B, Taulet N, Fowler B, Courtneidge SA, Bokoch GM (2009) Novel p47(phox)-related organizers regulate localized NADPH oxidase 1 (Nox1) activity. Sci Signal 2:ra54 63. Gianni D, Taulet N, DerMardirossian C, Bokoch GM (2010) c-Srcmediated phosphorylation of NoxA1 and Tks4 induces the reactive oxygen species (ROS)-dependent formation of functional invadopodia in human colon cancer cells. Mol Biol Cell 21:4287– 4298 64. Gianni D, Nicolas N, Zhang H, Der Mardirossian C, Kister J, Martinez L, Ferguson J, Roush WR, Brown SJ, Bokoch GM, Hodder P, Rosen H (2010) Optimization and characterization of an inhibitor for NADPH oxidase 1 (NOX-1). Probe Reports from the NIH Molecular Libraries Program, Bethesda, MD 65. Gianni D, Taulet N, Zhang H, DerMardirossian C, Kister J, Martinez L, Roush WR, Brown SJ, Bokoch GM, Rosen H (2010) A novel and specific NADPH oxidase-1 (Nox1) small-molecule inhibitor blocks the formation of functional invadopodia in human colon cancer cells. ACS Chem Biol 5:981–993 66. DerMardirossian C, Rocklin G, Seo JY, Bokoch GM (2006) Phosphorylation of RhoGDI by Src regulates rho GTPase binding and cytosol-membrane cycling. Mol Biol Cell 17:4760–4768 67. DerMardirossian CM, Bokoch GM (2006) Phosphorylation of RhoGDI by p21-activated kinase 1. Methods Enzymol 406:80–90 68. DerMardirossian C, Schnelzer A, Bokoch GM (2004) Phosphorylation of RhoGDI by Pak1 mediates dissociation of Rac GTPase. Mol Cell 15:117–127 69. Golovanov AP, Chuang TH, DerMardirossian C, Barsukov I, Hawkins D, Badii R, Bokoch GM, Lian LY, Roberts GC (2001) Structure-activity relationships in flexible protein domains: regulation of rho GTPases by RhoGDI and D4 GDI. J Mol Biol 305:121– 135 70. DerMardirossian C, Bokoch GM (2005) GDIs: central regulatory molecules in rho GTPase activation. Trends Cell Biol 15:356–363 71. Khan AA, Mao XO, Banwait S, DerMardirossian CM, Bokoch GM, Jin K, Greenberg DA (2008) Regulation of hypoxic neuronal death signaling by neuroglobin. FASEB J 22:1737–1747 72. Delorme V, Machacek M, DerMardirossian C, Anderson KL, Wittmann T, Hanein D, Waterman-Storer C, Danuser G, Bokoch GM (2007) Cofilin activity downstream of Pak1 regulates cell protrusion efficiency by organizing lamellipodium and lamella actin networks. Dev Cell 13:646–662

9

History and Discovery of the Noxes: From Nox1 to the DUOXes Albert van der Vliet

Abstract

Aerobic organisms have adapted to living in an oxygenrich environment by using the chemical energy of molecular oxygen (O2). In addition to mitochondrial respiration in which O2 is used to convert energy stored in macronutrients into the universal cellular energy donor ATP, organisms also evolved other mechanisms to convert O2 to partially reduced reactive oxygen species (ROS) for distinct biological purposes. As the biological production of ROS became more broadly recognized for its roles in antimicrobial host defense, crosslinking of extracellular matrix proteins, biosynthetic pathways of certain hormones (thyroid hormone), and reversible redox-based regulation of cell signaling, the biomedical research field began searching for the enzymatic source(s) of ROS production in these various cases. Following the initial discovery of a dedicated NADPH oxidase complex that mediates ROS-dependent host defense in phagocytes, the availability of genome sequence data towards the end of the last century allowed for the discovery of multiple homologs of NADPH oxidase, leading to the description of the NOX superfamily we know today. This chapter will summarize the main research developments over the past century that ultimately led to the discovery and identification of this large NOX enzyme family. Keywords

NOX · DUOX · Hydrogen peroxide · Heme peroxidase · Thiol-based redox signaling

A. van der Vliet (✉) Department of Pathology and Laboratory Medicine, Larner College of Medicine, University of Vermont, Burlington, VT, USA e-mail: [email protected]

1

Introduction

Gradual rises in atmospheric oxygen (O2) on our planet allowed for the evolution of large multicellular organisms, which took advantage of the substantial chemical energy that is released during its reduction to water (H2O), allowing for maintenance of body temperature and other physiological functions that would otherwise be impossible. Since reduction of O2 to H2O occurs in a stepwise manner, and produces reactive intermediates such as superoxide anion (O2●-) and hydrogen peroxide (H2O2), commonly grouped under the term “reactive oxygen species (ROS)”, these organisms also evolved with various enzymatic detoxification mechanisms to minimize the inherent risk of such ROS production. However, ROS production by mammalian cells may also be deliberate, which was first recognized in the early 1930s by Baldridge and Gerard, when they observed a so-called respiratory burst due to marked consumption of O2 by canine neutrophils during phagocytosis of bacteria. This eventually led to the identification of the phagocyte NADPH oxidase system, which mediates the regulated production of ROS by mammalian organisms to serve a physiologically important function, namely in host defense against various pathogens [1]. However, the first observations that organisms can utilize O2 and mediate oxidative processes date back much earlier. The occurrence of oxidative processes in biological systems was already recognized as early as the middle of the nineteenth century, even though their physiological relevance was not clear. The German/Swiss chemist Christian Friedrich Schönbein reported in 1863 that various animal or plant tissues are capable turning colorless guaiacol tinctures into blue ones in the presence of H2O2, implying the presence of “peroxidase” activity [2]. This led Felix HoppeSeyler in Tübingen to develop the concept of “activation” of oxygen by living tissues in 1883. In 1900, Oscar Loew, while working at the United States Department of Agriculture, first described catalase as an enzyme dedicated to specifically metabolize H2O2 [3], even though the ability of biological

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_9

133

134

systems to actually generate H2O2 had not yet been demonstrated. This would take until 1928, when fungal glucose oxidase was the first identified enzyme capable of producing H2O2 as a co-product [4]. Over the past century, a number of seemingly unrelated observations were made in highly diverse research fields, which eventually led to the recognition of common biochemical pathways that involve enzymatic production of ROS as bioactive intermediates, which in turn exert their biological functions through oxidative chemistry involving heme peroxidases or redox-sensitive cysteine residues in target proteins. During the second half of the past century, it became increasingly apparent that biological systems can generate ROS in a deliberate and regulated manner for diverse physiological purposes, which range from antimicrobial host defense to oxidative cross-linking in the fertilization envelope of sea urchin eggs, thyroid hormone biosynthesis, and intracellular redox-based cell signaling. Due to increasing research activity in this field and advancing methodology to detect ROS in biological systems, it became apparent that many non-phagocytic cell types are also capable of generating O2●- or H2O2, in response to growth factors or cytokines, suggesting a much wider use of ROS in biology than initially recognized. Thus, the biomedical research community began to gradually embrace the idea that ROS can serve broad biological functions as a “second messenger”, through their ability to regulate diverse cellular and biochemical processes in a reversible redox-based manner, a process now commonly known as redox signaling. Not surprisingly, this strongly intensified interest into the biological source (s) for such ROS. While the phagocyte NADPH oxidase system had already been described for its role in antimicrobial host defense, a major breakthrough was made in the late 1990s with the identification of several homologs of the main NADPH oxidase catalytic subunit, gp91phox, made possible by the increasing availability of genomic sequence data and bioinformatics tools. Initially, these novel homologs were thought to exhibit cell- or tissue-specific functions, but we now understand that this NADPH oxidase (NOX) family is widely distributed throughout nature, with many cells and organisms expressing multiple NOX homologs. NOX enzymes are now recognized to possess highly diverse functional properties based on diverse subcellular location and dependent on unique modes of regulation. The discovery of the NOX enzyme family has transformed the field of redox biology, and has uncovered an ever-growing number of redox-dependent biological processes that are driven by these NOX homologs. It also sparked still ongoing research efforts towards a more complete understanding of their functional regulation and structural biology and their potential for therapeutic targeting in various disease conditions. It is the objective of this review to discuss the main seminal research observations over the past century that ultimately led to the

A. van der Vliet

discovery of the currently known NOX enzyme superfamily. Some of these selected milestones are schematically illustrated in a discovery timeline that spans more than 100 years (Fig. 9.1). Rather than attempting to fully describe the discovery of the various NOX enzymes and their diverse functions, which has been covered in several outstanding recent reviews [1, 5–7], I will provide a historical account of some of the major initial scientific observations in three diverse research areas that culminated in the identification of these various NOX homologs. In addition, I will also discuss some general concepts and perspectives with respect to the widespread importance of these NOX enzymes throughout nature, and which will likely form the basis for continued research in this research field.

2

Oxidative Biochemistry in Sea Urchin Reproduction

In the early 1900s, Otto Warburg, best known for his work on tumor metabolism and the “Warburg effect”, was a chemistry student at the University of Freiburg and performed summer internships at the Marine Station in Naples between 1908 to 1911, during which he became engaged in studies of the sea urchin. This marine organism has a robust exterior with long, dark purple spines, that offer ample protection against outside threats, but their reproduction relies on female urchins releasing up to several million large transparent eggs for external fertilization that lack shells or coatings and are therefore highly vulnerable. During his internship in 1908, Otto Warburg observed that these sea urchin eggs dramatically increased consumption of O2 (by about 6- to 7-fold) when they were fertilized [8]. This O2 consumption could be inhibited by cyanide and narcotics, leading Warburg to suggest the involvement of a respiratory enzyme. This is the very first report of a biological organism being able to consume O2 for a specific biological purpose, and preceded the discovery of mitochondrial respiration and oxidative phosphorylation by almost two decades. In fact, it would take many more decades before it was realized that this O2 consumption by sea urchin eggs, which is transient and reflects a respiratory burst, was not due to mitochondrial respiration but represented a distinct enzymatic process. In the intervening time period, the existence of a respiratory burst by activated phagocytes had been discovered, due to initial observation by Baldridge and Gerard in studies of dog phagocytes [9]. In 1941, Kjell Agner discovered a green enzyme from tuberculous empyema and characterized it as a heme peroxidase with antimicrobial properties. He named it verdoperoxidase because of its green color [10], but we now known it as myeloperoxidase. Although the production of ROS had not yet been directly demonstrated, this verdoperoxidase was assumed to function as a catalase in

9

History and Discovery of the Noxes: From Nox1 to the DUOXes

135

Fig. 9.1 Timeline of major milestones in ROS biology and discovery of NADPH oxidases. Selected major discoveries are highlighted with respect to biological production of ROS, specific biological actions via

heme peroxidases, reversible modes of redox signaling, and the identification of multiple NADPH homologs

detoxifying H2O2 production during the respiratory burst. It would not be until the 1960s that the respiratory burst was realized to be independent of mitochondrial respiration and associated with NADPH-dependent production of O2●-/ H2O2 as the initial antimicrobial product, involving oxidative chemistry catalyzed by myeloperoxidase. The key components of this NADPH oxidase enzyme system where eventually identified as a cytochrome b flavoenzyme (see also Fig. 9.1). Inspired by these various observations, Bennett Shapiro and his colleagues, as well as others, were able to demonstrate that the O2 consumption by sea urchin eggs also serves to produce H2O2 as a substrate for a specific heme peroxidase, named ovoperoxidase, which is secreted within seconds of fertilization from cortical granules located below the egg’s membrane surface [11]. Thus secreted ovoperoxidase was found to utilize H2O2 to catalyze oxidative cross-linking reactions within the glycoprotein coat on the egg surface, by forming di- and tri-tyrosine cross-links, thereby converting it into a rigid and relatively impermeable fertilization membrane. This oxidative chemistry thus serves

a unique biological function, and helps turn the sea urchin egg envelope into a hardened matrix that resists biochemical and mechanical challenges, in addition to enhancing spermicidal activity to prevent against polyspermy. The final question with respect to the identity of the (enzymatic) source of H2O2 would still remain unanswered for another four decades. Resembling the neutrophil NADPH oxidase, it was observed that H2O2 production by sea urchin eggs also requires a membrane-associated NADPH:O2 oxidoreductase [12], but it would not be until 2004 until the responsible enzyme was identified [13], after the existence of diverse NADPH oxidase homologs in mammalian systems had been demonstrated. Thus, the enzymatic system responsible for oxidative biochemistry in promoting sea urchin reproduction had now been described, and involves cooperative actions between an NADPH oxidase (as a source of H2O2) and a locally secreted heme peroxidase (ovoperoxidase), highly analogous to the cooperative actions of the phagocyte NADPH oxidase with myeloperoxidase to kill phagocytosed bacteria (Fig. 9.2). As we will see below,

136

A. van der Vliet

Fig. 9.2 Comparison of functional NOX-peroxidase partnerships in different biological settings. Activation of NOX/DUOX isoforms in three distinct scenarios, bacterial phagocytosis (left), fertilization of sea urchin eggs (center), or thyroid hormone synthesis (right), occurs in conjunction with local secretion of a heme peroxidase,

myeloperoxidase (MPO), ovoperoxidase (OVO), or thyroperoxidase, to catalyze oxidation of halides (Cl-, I-) or tyrosine residues. NOX/DUOX activation mechanisms are highlighted in red. Tg: thyroglobulin

we would soon learn of other examples of similar cooperative actions between NADPH oxidases and locally secreted heme peroxidases, which appear to have evolved throughout nature for various biological purposes [14].

circulation as functional hormones. The key chemical event in thyroid hormone synthesis is the local oxidation of iodide (I-), by a locally secreted heme peroxidase known as thyroid peroxidase (aka thyroperoxidase; TPO), to either molecular iodine (I2) or hypoiodous acid (HOI) [16], allowing its incorporation into tyrosyl residues of Tg, and also mediates the oxidative coupling of iodotyrosine residues within Tg, to eventually produce both T3 and T4 (Fig. 9.2). The critical nature of iodine as a dietary trace element rests primarily on its use in thyroid hormone, although iodine also has additional functions outside the thyroid. Indeed, iodide is an excellent substrate for other peroxidases, such as neutrophil myeloperoxidase, and its successful use as an antiseptic in wound care for over 160 years [17] is based largely on this chemistry. The oxidative iodination chemistry by TPO is also highly analogous to that of the NADPH oxidaseovoperoxidase system in the sea urchin, as both involve oxidation and cross-linking of tyrosine residues. It is also intriguing to note that, just like in the case of the fertilized sea urchin egg or the activated phagocyte, the oxidative chemistry and the responsible peroxidase was characterized well before the H2O2 source was identified. Indeed, TPO was first purified in the 1960s from porcine thyroid membranes and from human thyroid [16], as an additional member of the peroxidase-cyclooxygenase superfamily [18], and the oxidative biochemistry of thyroid hormone synthesis was soon established [16]. The identity of the H2O2-generating mechanism remained the last piece of the puzzle to be solved. After initial studies in the early 1970s revealing H2O2 production in the thyroid [19, 20], subsequent studies by several research groups, including those led by Jacques Dumont and Alain Virion, demonstrated that this involves a membrane-bound

3

Thyrocytes Produce H2O2 for Biosynthesis of Thyroid Hormone

Iodine is the heaviest of the stable halogens and is sparingly present in biological systems, but has long been recognized as an essential dietary nutrient. Iodine deficiency was identified as a major cause of hypothyroidism (goiter), which was especially widespread and prevalent in the nineteenth century, and led to the common use of iodized salt in most households. The presence of iodine in biology was first observed by Eugen Baumann, a chemist at the University of Freiburg in the late nineteenth century. He collected thyroid glands of 1000 sheep and boiled them in dilute sulfuric acid, after which he noted the presence of high levels of iodine in the resulting precipitate, which he therefore named “iodothyrin” [15]. We now know that iodine is used as a component of the thyroid hormones thyroxine (T4) and triiodothyronine (T3), important tyrosine-based hormones that are critical for regulating metabolism. The biosynthesis of both T4 and T3 depends on iodination of tyrosyl residues of thyroglobulin (Tg), which is produced by the follicular cells of the thyroid gland in response to thyroid-stimulating hormone (TSH) originating from the pituitary gland, and secreted into colloid in the lumen of the thyroid follicle. Iodinated Tg is then endocytosed by the thyrocyte and proteolytically processed to form T3 and T4, which are then secreted into the

9

History and Discovery of the Noxes: From Nox1 to the DUOXes

flavoprotein localized to the apical thyrocyte membrane requiring both calcium and NADPH [20–22], suggesting the involvement of an NADPH oxidase. However, it would not be until the end of 1999 that Corrine Dupuy in the research group of Alain Virion purified and cloned the enzyme responsible for this H2O2 production, and named it p138Tox, based on its homology to the phagocyte gp91Phox and the then just discovered mitogenic oxidase p65Mox [23, 24] (see also next section). In the same year, the research team led by Francoise Miot identified this thyroid homolog as two highly homologous but distinct cDNAs, and referred to them as thyroid oxidase (ThOX)-1 and -2 [25] (See also Chap. 14 by F. Miot and X. De Deken). Thus, all components of this thyroid hormone synthesis pathway were now identified, paving the way for further studies towards their regulation and involvement in thyroid-based disease. One important aspect was the discovery of a number of inactivating mutations in the ThOX2 gene in patients with congenital hypothyroidism [26], directly establishing its importance in thyroid hormone synthesis.

4

From “Oxidative Stress” to “Redox Signaling”: First Discovery of NADPH Oxidase Homologs

The previous sections highlight some specific examples in which ROS production by certain cell types can serve unique biological functions. However, the prevailing wisdom within the biomedical research field up until the 1990s was that ROS largely represent toxic byproducts, formed due to mitochondrial electron transport leak, or resulting from certain enzymatic pathways (e.g. glucose oxidase or amino acid oxidase) or metabolism of xenobiotics, and are potentially harmful due to their ability to induce irreversible oxidative damage to critical cell components. The continued production of H2O2 by intact mitochondria was first directly demonstrated in the 1960s [27, 28] and was considered a biological “accident”, which cells are fortunately able to tolerate due to the presence of detoxifying enzymes such as catalase. The discovery of a dedicated superoxide dismutase (SOD) and GSH peroxidase soon thereafter [29, 30] only helped to further establish this concept. In fact, these initial findings launched a vibrant research field dedicated to addressing the potential pathogenic roles of ROS in toxicology or disease, which led to development of various direct and indirect methods to detect the presence of ROS in various biological systems. Perhaps somewhat unintentionally, this also led to an increasing number of observations of measurable O2●- or H2O2 production by various non-phagocytic cell types in response to biological stimuli. Among the first such examples are observations in the late 1970s of epididymal fat cells, which were found to generate intracellular H2O2 upon stimulation with insulin via a putative NADPH oxidase, with a suspected

137

role in insulin signaling and glucose metabolism [31, 32]. Subsequent studies over the next decade demonstrated extracellular ROS production by fibroblasts, endothelial cells, and bronchial epithelial cells, in response to pro-inflammatory cytokines or activators of e.g. protein kinase C [33–36]. Moreover, various human tumor cells were found to generate considerable amounts of H2O2, even in the absence of exogenous stimulation [37]. In each of these cases, this ROS production was independent of mitochondria and was inhibitable by flavoprotein inhibitors such as diphenyleneiodonium, thus suggesting the involvement of an NADPH oxidase. In most of the examples summarized above, the purpose of such ROS production was not addressed at the time, and was typically believed to contribute to oxidative injury during e.g. inflammation, although others suggested that it may contribute to cell proliferation and tumor growth [33, 37]. Indeed, the field began to shift away from a common view of ROS as toxic byproducts of metabolism to their potential action as second messengers in growth factor or cytokine signaling. One of the first pathways that was found to be regulated in such manner is the activation of nuclear factor (NF)-κB, a critical transcription factor involved in pro-inflammatory signaling. Activation of NF-κB could in many cases be prevented by compounds that block or detoxify ROS, and could often also be mimicked by direct administration of ROS [38], and research over subsequent decades has identified several redox-sensitive components in this signaling pathway. Likewise, people also started to recognize an intricate relationship between ROS and tyrosine kinase signaling. First, Nicholas Tonks and coworkers demonstrated that protein tyrosine phosphatases are highly susceptible to inhibition by thiolreactive chemicals and by ROS, since they invariably contain a highly conserved cysteine residue that is essential for catalytic function [39]. Not long afterwards, Bauskin and coworkers discovered that an ER-localized tyrosine kinase, Ltk, could be directly activated in a ligand-independent manner by thiol-reactive compounds including ROS [40]. This was particularly intriguing, considering that tyrosine phosphorylation and the existence of tyrosine kinases are unique features of multicellular organisms that arose during periods of increased atmospheric O2, suggesting that tyrosine phosphorylation pathways may have co-evolved with ROS-producing mechanisms that regulate them [41]. Greg Downey and co-workers were among the first to address a direct link between NADPH oxidase activation and enhanced tyrosine phosphorylation in neutrophils, and found that such increased phosphorylation was NADPH-dependent and could be prevented by known flavoprotein inhibitors [42]. Soon thereafter, independent research in the laboratories of Toren Finkel and Sue Goo Rhee demonstrated that transient cellular production of H2O2 in non-phagocytic cells in response to growth factors, platelet-derived growth factor

138

(PDGF) or epidermal growth factor (EGF), was directly linked to enhanced tyrosine kinase signaling by these growth factors, and could be attributed to transient oxidative inhibition of protein tyrosine phosphatases [43, 44]. The various seminal observations described above gradually changed the landscape of ROS biology, and shifted the perception of ROS towards a role in reversible oxidative processes that regulate diverse physiological outcomes, a concept known as redox-based signaling. As this idea gained increasing interest, the main missing piece remained the identity of the enzymatic system(s) responsible for such ROS production. The suspected involvement of an NADPH oxidase was in most cases based on the ability of flavoprotein inhibitors to prevent such responses, detection of components of the phagocyte NADPH oxidase system in non-phagocytic cells using RT-PCR or antibodies that were available at the time, but the presence of the main business end of the NADPH oxidase complex, gp91phox, could typically not be established. It was fortunate that the 1990s also witnessed the rapid development of genomic analysis tools and increasing availability of genomic DNA sequence data from various organisms, which allowed researchers to search existing expressed sequence tag (EST) databases for potential NADPH oxidase homologs by sequence comparisons, and indeed led to the discovery of separate gene products that are homologous to gp91phox. The first such homolog was actually discovered in rice (Oryza sativa), based on a search fueled by observations that plants can also produce ROS as part of several plant defense responses [45]. The protein product was found to be about 62% similar and 37% homologous to gp91phox, and also displayed considerable similarity with the yeast ferric reductases FRE1 and FRP1, and was named respiratory burst oxidase homologue A (RbohA). Soon afterwards, the full-length rbohA cDNA was cloned from Arabidopsis thaliana, which indicated that this gene encoded for a protein with NADPH dehydrogenase and transmembrane domains similar to gp91phox, but also included a large unique hydrophilic N-terminal domain that contains two Ca2+-binding EF hand motifs with similarity to human GTPase activating protein [46] (See also Chap. 26 by G. Miller and R. Mittler). The presence of Ca2+-binding EF hand motifs suggested a mechanism for direct regulation by Ca2+, and the authors also recognized a similarity with the putative NADPH oxidase in thyrocytes which had not yet been identified at the time but was known to be regulated by Ca2+ [46, 47]. It would not be long until the first mammalian gp91phox homolog was also described by Dave Lambeth and co-workers, who sequenced this homolog from human colon cDNA and found it to be similar in size and 56% identical to gp91phox. They also noted that the tissue mRNA distribution of this homolog differed greatly from that of gp91phox, with high expression in the colon and enhanced expression in vascular smooth muscle cells in response to growth factors.

A. van der Vliet

Because of its apparent role in proliferation, Lambeth decided to name this homolog mitogenic oxidase (Mox1) [23]. Around the same time, Botond Banfi in the lab of Karl-Heinze Krause reported identification of the same gene and termed it NADPH oxidase homolog 1 (NOH-1). They also noted the existence of multiple splice variants of NOH-1 (NOH-1S and NOH-1L), with the short homolog lacking NADPH and FAD binding domains and postulated to represent a H+ channel, although it was later realized that this short isoform is not a naturally occurring splice variant but may have been observed artifactually [48]. Events followed in rapid succession. Just months after the initial report of Mox1, the aforementioned p138Tox homolog in thyrocytes was cloned and described by Dupuy as a unique large homolog that not only contains the basic gp91phox-like domains, but also an intracellular region with Ca2+-binding EF domain, similar to RbohA, and an additional N-terminal transmembrane region and extracellular domain that is homologous to heme peroxidases [24]. The following year, Miklos Geiszt in the research group of Thomas Leto searched mouse and human kidney libraries for an NADPH oxidase that could potentially participate in oxygen sensing mechanisms that regulate production of erythropoietin. They identified an additional NADPH homolog that differed from Mox1, with 40% sequence identity and 57% similarity to gp91phox and containing all conserved NADPH oxidase features, ie. six hydrophobic segments within the N-terminal portion and a C-terminal domain with binding sites for FAD and NADPH. Since this homolog was particulary highly expressed in proximal tubule epithelial cells in the kidney, it was named Renox [49].

5

The NOX Family Expands: From Plants to Fungi and Bacteria

It was quickly realized that these new homologs represented a larger NADPH oxidase family that is widely distributed throughout animal and plant kingdoms, with variable and homolog-specific tissue distribution [7, 50–52]. Initially, these new homologs were named based on their tissue of origin or protein size (e.g. p138Tox [24] or Renox [49]), and were also known by alternative names (e.g. Mox1 was also referred to as NOH-1 [48] or gp91-2 [53], and p138Tox was also called ThOX). This, combined with the fact that the name Mox1 was already in use for a homeobox gene family member [54], urged Dave Lambeth and other pioneers in this field to adopt a different and more consistent official nomenclature. They therefore established a consensus terminology based on the name NOX (for NADPH oxidase); NOX1 for Mox1, NOX2 for gp91phox, NOX3 for an as yet uncharacterized homolog, and NOX4 for Renox [51, 55], which was adopted by the HUGO Human Gene

9

History and Discovery of the Noxes: From Nox1 to the DUOXes

139

Fig. 9.3 Structural comparison of different NOX enzyme families. Mammalian NOX1-4 contain a conserved C-terminal core region consisting of 6 transmembrane α-helices (cylinders) and 2 heme groups (indicated by ‘Fe’), and form a heterodimer with p22phox that contains 2 transmembrane α-helices and a proline rich region (PRR). Activation of NOX1-3 also requires assembly with various cytosolic co-factors (not shown). The same core structure is also found in all RBOHs in plants, which also contain an extended N-terminal region with 2 EF-hand

motifs. Mammalian Nox5 has a similar N-terminal region with four EF hand motifs. Finally, mammalian Duoxes contain an additional transmembrane region and an extracellular N-terminal peroxidase homology domain (PHD). DUOXes also form heterodimers with Duoxa proteins (maturation factors) consisting of 5 transmembrane α-helices and a large extracellular loop that interacts with the PHD domains in DUOX1/2

Nomenclature Committee. Although NOX3, the closest homolog to gp91phox with 58% identity, was named before it was characterized, it was soon described to display unique expression in the inner ear with a putative function in the production of otoconia [56]. An additional NOX homolog, NOX5, was also soon described and found to contain a unique N-terminus with four EF-hand Ca2+ binding domains [57, 58], suggesting regulation by Ca2+ which was later more fully established. Similar to gp91phox, NOX1, NOX3 and

NOX4 were also found to depend on the presence of an additional transmembrane subunit, p22phox [5, 6], and homologs of p47phox and p67phox would also soon be discovered as additional cytosolic co-factors in the regulation of NOX1 and 3 [59]. The similarities and differences between the various NOX/RBOH families are schematically illustrated in Fig. 9.3. In spite of some initial resistance by researchers in the phagocyte field, this new terminology is now widely adopted

140

for mammalian NOX enzymes. The plant community has continued to utilize the name RBOH, which has expanded to ten members (RBOHA to RBOHH) in e.g. Arabidopsis with distinct expression patterns and functional roles in development and/or stress responses [50]. Also, the longer mammalian NADPH homologs that were initially cloned from thyrocytes (p138Tox/ThOX1/2) were not renamed to NOX6 and NOX7, but are instead referred to as Dual Oxidase (DUOX) 1 and 2, a name chosen based on the fact that they contain an additional extracellular protein domain that is homologous to heme peroxidases [60]. In fact, the DUOXes were also included in a recently constructed phylogenic tree of all 400 members of the peroxidase-cyclooxygenase superfamily, as cluster 7 [18]. Both oxidase and peroxidase activities were found to be intrinsic to the functional Duox homolog in Caenorhabditis elegans, CeDUOX1, as it was found to be capable of directly catalyzing tyrosine crosslinking within the extracellular matrix of the larval cuticle [60] (See also Chap. 27 by D.A. Garsin). This functional role in tyrosine cross-linking is analogous to the NADPH oxidase enzyme within sea urchin eggs, which was subsequently also identified as a dual oxidase homolog (Udx1, for “urchin dual oxidase 1”) [13]. However, the ability of Udx1 to promote tyrosine crosslinking required the presence of a separate heme peroxidase, ovoperoxidase [61], just like mammalian DUOX1/2 in thyrocytes also require a separate thyroperoxidase protein to aid in thyroid hormone synthesis (Fig. 9.2). This initially puzzled investigators with respect to the functional importance of the mammalian DUOX peroxidase homology domains (PHD), as they were found to lack critical amino acid residues involved in covalent heme binding and have no measurable peroxidase activity in contrast to the CeDUOX1 homologs. Interestingly, the tyrosine crosslinking properties of Ce-DUOX (also known as bli-3 because of a blistered cuticle phenotype associated with its deletion) were later also found to involve a separate heme peroxidase, MoLT-7 [62]. It was proposed that DUOX proteins may have evolved to adopt different functions in higher organisms, with a more heavy reliance on cooperative actions with locally secreted heme peroxidases in diverse biological contexts [14]. Also, the PHD domains in mammalian DUOX proteins were later realized to facilitate their appropriate subcellular localization and function at the apical plasma membrane, a notion also supported by observed associations of mutations within the human DUOX2 PHD domain with hypothyroidism [63]. Over the last 20 years, it has become clear that the NADPH oxidases form a large family that is widely distributed throughout biology, and can be found in plants, algae, fungi, amoebae, nematode worms, echinoderms, urochordates, insects, fish, reptiles, birds, and mammals, but they were originally not observed in prokaryotes [7, 52]. In their review in 2007, Takemoto and colleagues also

A. van der Vliet

highlighted the presence of several NOX homologs in some fungal species, but at the time no NOX genes had been observed in available genomes of hemiascomycete yeasts or other unicellular fungal species [64] (See also Chap. 25 by D. Takemoto and B. Scott). Thus, it was assumed that the NOX enzymes evolved at about the same time as single-cell eukaryotes, predating multicellularity by about 1.4 billion years [52], but the fact that some homologs were only observed in multicellular organisms suggested important roles in cell-specific functions and development. However, since NOX enzymes were noted to be homologous to yeast ferric reductases (FRE), which mediate transmembrane transfer from NADPH to reduce extracellular ferric iron (Fe3+) to promote uptake of this essential element, they are believed to be members of a much larger transmembrane ferric reductase domain (FRD) superfamily, which also includes the prokaryotic transmembrane protein YedZ (which lacks a dehydrogenase domain present in NOX or FRE) [65, 66]. This led to suggestions that the NOX and FRE enzymes may both have arisen by fusion of two ancestral genes, a transmembrane FRD domain homologous to cytochrome b and a cytosolic domain homologous to prokaryotic ferredoxin-NADP+ reductase (FNR) [1, 7, 65], and also fueled speculation that functional NOX enzymes might also be present in unicellular organisms and in prokaryotes. Indeed, more targeted studies of unicellular yeast species, Saccharomyces cerevisae or Candida albicans, indicated that some members of their FRE subfamilies actually act as genuine NADPH oxidases [67, 68] (See also Chap. 24 by M. Breitenbach et al.). More recently, using an updated algorithm using only essential eukaryotic motifs for NADPH/FAD binding and hemebinding His-X12–14-His sequences, approximately 1000 new NOX sequences were observed in bacterial genomes. One of these, Streptococcus pneumonia NOX (SpNOX), was directly demonstrated to function as a bona fide prokaryotic NOX homolog, containing a heme-binding transmembrane domain and a flavin-binding dehydrogenase domain, similar to NOX2 [69]. The function of such prokaryotic NOX enzymes still remains enigmatic, but could potentially be related to bacterial stress responses or biofilm formation or dynamics. Researchers in the NOX field have long been fascinated as to how they may have evolved in nature, and the consensus seems to be that the earliest ancestral NOX homolog likely arose in prokaryotes prior to the divergence of life into fungi, plants, and animals. This ancestral homolog likely possessed the basic NOX structure, consisting of 6 transmembrane domains (containing two asymmetrical hemes) and a long cytoplasmic C-terminal (containing the FAD and NADPH binding sites), and acquisition of Ca2+-binding EF-hand domains by an ancestral NOX would then have led to NOX5/RBOH-like isoforms when filamentous fungi and plants evolved. The DUOX isoforms may have developed

9

History and Discovery of the Noxes: From Nox1 to the DUOXes

from an early NOX5-like isoform through the additional acquisition of a peroxidase homology domain. Some NOX homologs may have appeared much later, such as the NOX3 homolog which is only found in reptiles, birds, and mammals, suggesting its emergence after the evolution of fish and amphibians and coinciding with adaptation of vertebrates to land. Also, while all other mammalian species contain seven NOX homologs (NOX1-5, and DUOX1/2), the NOX5 gene was apparently lost in rodents, thus fueling further speculation towards its biological function.

6

General Aspects of NOX Biology

Now that the NOX family has been (almost) fully described, there are still many remaining fundamental questions with respect to their biological role(s), and this forms an important topic of active ongoing research. It would certainly be impossible to adequately cover the current status of knowledge with respect to their regulation and functional properties, and these will be covered in more detail in other chapters in this book. Rather, I would like to use this final section to highlight some general aspects that are relevant to the NOX research field as a whole. First, one intriguing aspect is that many mammalian cell types express more than one functional NOX homolog, suggesting that different NOX enzymes play unique and non-redundant functions, depending on their subcellular location, their modes of activation, or their specific interactions with biological targets. This forms a critical current concept in the field of redox biology, which is far removed from the original concept of “cellular redox status” or the still common use of non-targeted (antioxidant) strategies to understand redox biology. Moreover, emerging studies also suggest the existence of various interrelationships between different NOX homologs, indicating that they work in a concerted manner, which has further complicated the identification of specific contributions of individual NOX homologs to biological outcomes. It seems that different NOX enzymes may play specific roles in diverse cellular redox networks, although they may also display some redundancy given the fact that genetic deletion of a single NOX homolog often doesn’t induce major functional defect. Another aspect that is often not fully considered is whether it is either O2●- or H2O2 that is the biologically relevant product of activated NOX enzymes. Although the primary product of most NOX enzymes is O2●-, H2O2 is most often viewed as the main biologically relevant product, because of its relative stability and diffusibility and ability to directly react with heme peroxidases or proteins with redoxsensitive cysteines. This notion is further supported by observations that activation or induction of NOX often coincides with recruitment or induction of superoxide dismutase (SOD) enzymes, which would efficiently convert

141

NOX-generated O2●- to H2O2 as the main bioactive mediator [68, 70]. Nevertheless, given the ability of O2●- to participate in diverse biologically relevant one-electron reactions, e.g. with nitric oxide (NO●) or with redox-active transition metals in proteins, its potential role in NOX biology should not be dismissed. The topology of NOX enzymes, which mediate electron transport across membranes to reduce extracellular O2, has also long puzzled investigators as it was difficult to envision how this is compatible with their role in tightly regulated oxidation of intracellular targets. The recent identification of some members of the large aquaporin water channel family as transmembrane transporters of H2O2 [71] has helped resolve this issue, and suggest that these channels provide an additional level of control in redox signaling by regulating access of H2O2 to select intracellular targets. Lastly, as is clear from the narrative above, the NOX enzymes were discovered because of their involvement in biological ROS production. In fact, it is often claimed that production of O2●- or H2O2 in a regulated fashion is the sole function of these enzymes. Yet, the biological outcomes of NOX activation may not fully rest on ROS production, and may also depend on cell responses to NADPH oxidation (resulting in localized pH changes) or to the induction of transmembrane charge alterations that lead to activation of ion channels to drive transmembrane fluxes of protons or other ions [72]. Although these additional aspects of NOX enzymology have long been recognized, their potential contribution to NOX-dependent biological outcomes is often not considered. Indeed, ROS production and associated oxidative events may be just one aspect of NOX biology, although this is receiving by far the most widespread attention.

7

Final Outlook

The past century has witnessed an interesting sequence of discoveries, which began with recognized ability of mammalian and plant cell types to produce ROS upon appropriate stimuli to mediate a variety of biological processes by redoxbased mechanisms, and eventually culminated into the discovery of widely distributed NOX enzymes. The discovery of NOX enzymes and their diverse roles in host defense, developmental aspects, and cell proliferation and differentiation, has transformed the overall field of redox biology. Discoveries of genetic mutations in various NOX homologs linked to e.g. chronic granulomatous disease (CGD) or congenital hypothyroidism, and the development of multiple model organisms with genetic deficiency of one or more NOX homologs, have contributed significantly to our overall current understanding with respect to their functional role(s). Recent advances with respect to their structural biology [73– 75] (see also Chap. 30 by J.-X. Wu et al. and Chap. 31 by W. Oosterheert et al.), combined with continued development of redox biology tools and application of single cell

142

transcriptomics and proteomics approaches, will likely lead to exciting new discoveries with respect to their specific functional role(s) which may find potential application in e.g. plant-based food manufacturing or as therapeutic targets in treatment of human disease.

References 1. Vermot A, Petit-Hartlein I, Smith SME, Fieschi F (2021) NADPH oxidases (NOX): an overview from discovery, molecular mechanisms to physiology and pathology. Antioxidants (Basel) 10(6) 2. Clark RA (2000) Peroxidases: a historical overview of milestones in research on myeloperoxidase. In: Petrides PE, Nauseef WM (eds) The peroxidase multigene family of enzymes. Springer, Berlin, Heidelberg, pp 1–10 3. Loew O (1900) A new enzyme of general occurrence in organismis. Science 11(279):701–702 4. Nathan C, Cunningham-Bussel A (2013) Beyond oxidative stress: an immunologist’s guide to reactive oxygen species. Nat Rev Immunol 13(5):349–361 5. Lambeth JD (2004) NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol 4(3):181–189 6. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87(1):245–313 7. Sumimoto H (2008) Structure, regulation and evolution of Nox-family NADPH oxidases that produce reactive oxygen species. FEBS J 275(13):3249–3277 8. Warburg O (1908) Beobachtungen uber die Oxydationsprozesse im Seeigelei. Z Physiol Chem 57:1–16 9. Baldridge CW, Gerard RW (1932) The extra respiration of phagocytosis. Am J Phys 103:235–236 10. Agner K (1947) Detoxicating effect of verdoperoxidase on toxins. Nature 159(4034):271–272 11. Foerder CA, Klebanoff SJ, Shapiro BM (1978) Hydrogen peroxide production, chemiluminescence, and the respiratory burst of fertilization: interrelated events in early sea urchin development. Proc Natl Acad Sci U S A 75(7):3183–3187 12. Heinecke JW, Shapiro BM (1989) Respiratory burst oxidase of fertilization. Proc Natl Acad Sci U S A 86(4):1259–1263 13. Wong JL, Creton R, Wessel GM (2004) The oxidative burst at fertilization is dependent upon activation of the dual oxidase Udx1. Dev Cell 7(6):801–814 14. Sirokmany G, Geiszt M (2019) The relationship of NADPH oxidases and Heme peroxidases: Fallin' in and out. Front Immunol 10:394 15. Niazi AK, Kalra S, Irfan A, Islam A (2011) Thyroidology over the ages. Indian J Endocrinol Metab 15(Suppl 2):S121–S126 16. Ruf J, Carayon P (2006) Structural and functional aspects of thyroid peroxidase. Arch Biochem Biophys 445(2):269–277 17. Cooper RA (2007) Iodine revisited. Int Wound J 4(2):124–137 18. Nicolussi A, Auer M, Sevcnikar B, Paumann-Page M, Pfanzagl V, Zamocky M, Hofbauer S, Furtmuller PG, Obinger C (2018) Posttranslational modification of heme in peroxidases - impact on structure and catalysis. Arch Biochem Biophys 643:14–23 19. Benard B, Brault J (1971) Production of peroxide in the thyroid. Union Med Can 100(4):701–705 20. Virion A, Michot JL, Deme D, Kaniewski J, Pommier J (1984) NADPH-dependent H2O2 generation and peroxidase activity in thyroid particular fraction. Mol Cell Endocrinol 36(1–2):95–105 21. Willems C, Rocmans P, Dumont JE (1971) Calcium requirements in the action of thyrotropin on the thyroid. FEBS Lett 14(5):323–325

A. van der Vliet 22. Deme D, Doussiere J, De Sandro V, Dupuy C, Pommier J, Virion A (1994) The Ca2+/NADPH-dependent H2O2 generator in thyroid plasma membrane: inhibition by diphenyleneiodonium. Biochem J 301(Pt 1):75–81 23. Suh YA, Arnold RS, Lassegue B, Shi J, Xu X, Sorescu D, Chung AB, Griendling KK, Lambeth JD (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401(6748):79–82 24. Dupuy C, Ohayon R, Valent A, Noel-Hudson MS, Deme D, Virion A (1999) Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. Cloning of the porcine and human cdnas. J Biol Chem 274(52):37265–37269 25. De Deken X, Wang D, Many MC, Costagliola S, Libert F, Vassart G, Dumont JE, Miot F (2000) Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275(30):23227–23233 26. Moreno JC, Bikker H, Kempers MJ, van Trotsenburg AS, Baas F, de Vijlder JJ, Vulsma T, Ris-Stalpers C (2002) Inactivating mutations in the gene for thyroid oxidase 2 (THOX2) and congenital hypothyroidism. N Engl J Med 347(2):95–102 27. Jensen PK (1966) Antimycin-insensitive oxidation of succinate and reduced nicotinamide-adenine dinucleotide in electron-transport particles. I. pH dependency and hydrogen peroxide formation. Biochim Biophys Acta 122(2):157–166 28. Murphy MP (2009) How mitochondria produce reactive oxygen species. Biochem J 417(1):1–13 29. McCord JM, Fridovich I (1969) Superoxide dismutase. An enzymic function for erythrocuprein (hemocuprein). J Biol Chem 244(22): 6049–6055 30. Loschen G, Flohe L, Chance B (1971) Respiratory chain linked H(2) O(2) production in pigeon heart mitochondria. FEBS Lett 18(2): 261–264 31. May JM, de Haen C (1979) Insulin-stimulated intracellular hydrogen peroxide production in rat epididymal fat cells. J Biol Chem 254(7):2214–2220 32. Mukherjee SP, Lynn WS (1977) Reduced nicotinamide adenine dinucleotide phosphate oxidase in adipocyte plasma membrane and its activation by insulin. Possible role in the hormone’s effects on adenylate cyclase and the hexose monophosphate shunt. Arch Biochem Biophys 184(1):69–76 33. Cross AR, Jones OT (1991) Enzymic mechanisms of superoxide production. Biochim Biophys Acta 1057(3):281–298 34. Kinnula VL, Adler KB, Ackley NJ, Crapo JD (1992) Release of reactive oxygen species by Guinea pig tracheal epithelial cells in vitro. Am J Phys 262(6 Pt 1):L708–L712 35. Meier B, Radeke HH, Selle S, Younes M, Sies H, Resch K, Habermehl GG (1989) Human fibroblasts release reactive oxygen species in response to interleukin-1 or tumour necrosis factor-alpha. Biochem J 263(2):539–545 36. Matsubara T, Ziff M (1986) Increased superoxide anion release from human endothelial cells in response to cytokines. J Immunol 137(10):3295–3298 37. Szatrowski TP, Nathan CF (1991) Production of large amounts of hydrogen peroxide by human tumor cells. Cancer Res 51(3): 794–798 38. Schreck R, Baeuerle PA (1991) A role for oxygen radicals as second messengers. Trends Cell Biol 1(2–3):39–42 39. Tonks NK, Diltz CD, Fischer EH (1988) Characterization of the major protein-tyrosine-phosphatases of human placenta. J Biol Chem 263(14):6731–6737 40. Bauskin AR, Alkalay I, Ben-Neriah Y (1991) Redox regulation of a protein tyrosine kinase in the endoplasmic reticulum. Cell 66(4): 685–696 41. Dustin CM, Heppner DE, Lin MJ, van der Vliet A (2020) Redox regulation of tyrosine kinase signalling: more than meets the eye. J Biochem 167(2):151–163

9

History and Discovery of the Noxes: From Nox1 to the DUOXes

42. Fialkow L, Chan CK, Grinstein S, Downey GP (1993) Regulation of tyrosine phosphorylation in neutrophils by the NADPH oxidase. Role of reactive oxygen intermediates. J Biol Chem 268(23): 17131–17137 43. Sundaresan M, Yu ZX, Ferrans VJ, Irani K, Finkel T (1995) Requirement for generation of H2O2 for platelet-derived growth factor signal transduction. Science 270(5234):296–299 44. Lee SR, Kwon KS, Kim SR, Rhee SG (1998) Reversible inactivation of protein-tyrosine phosphatase 1B in A431 cells stimulated with epidermal growth factor. J Biol Chem 273(25):15366–15372 45. Groom QJ, Torres MA, Fordham-Skelton AP, Hammond-Kosack KE, Robinson NJ, Jones JD (1996) rbohA, a rice homologue of the mammalian gp91phox respiratory burst oxidase gene. Plant J 10(3): 515–522 46. Keller T, Damude HG, Werner D, Doerner P, Dixon RA, Lamb C (1998) A plant homolog of the neutrophil NADPH oxidase gp91phox subunit gene encodes a plasma membrane protein with Ca2+ binding motifs. Plant Cell 10(2):255–266 47. Gorin Y, Leseney AM, Ohayon R, Dupuy C, Pommier J, Virion A, Deme D (1997) Regulation of the thyroid NADPH-dependent H2O2 generator by Ca2+: studies with phenylarsine oxide in thyroid plasma membrane. Biochem J 321(Pt 2):383–388 48. Banfi B, Maturana A, Jaconi S, Arnaudeau S, Laforge T, Sinha B, Ligeti E, Demaurex N, Krause KH (2000) A mammalian H+ channel generated through alternative splicing of the NADPH oxidase homolog NOH-1. Science 287(5450):138–142 49. Geiszt M, Kopp JB, Varnai P, Leto TL (2000) Identification of renox, an NAD(P)H oxidase in kidney. Proc Natl Acad Sci U S A 97(14):8010–8014 50. Chapman JM, Muhlemann JK, Gayomba SR, Muday GK (2019) RBOH-dependent ROS synthesis and ROS scavenging by plant specialized metabolites to modulate plant development and stress responses. Chem Res Toxicol 32(3):370–396 51. Lambeth JD, Cheng G, Arnold RS, Edens WA (2000) Novel homologs of gp91phox. Trends Biochem Sci 25(10):459–461 52. Kawahara T, Quinn MT, Lambeth JD (2007) Molecular evolution of the reactive oxygen-generating NADPH oxidase (Nox/Duox) family of enzymes. BMC Evol Biol 7:109 53. Kikuchi H, Hikage M, Miyashita H, Fukumoto M (2000) NADPH oxidase subunit, gp91(phox) homologue, preferentially expressed in human colon epithelial cells. Gene 254(1–2):237–243 54. Mankoo BS, Collins NS, Ashby P, Grigorieva E, Pevny LH, Candia A, Wright CV, Rigby PW, Pachnis V (1999) Mox2 is a component of the genetic hierarchy controlling limb muscle development. Nature 400(6739):69–73 55. Shiose A, Kuroda J, Tsuruya K, Hirai M, Hirakata H, Naito S, Hattori M, Sakaki Y, Sumimoto H (2001) A novel superoxideproducing NAD(P)H oxidase in kidney. J Biol Chem 276(2): 1417–1423 56. Banfi B, Malgrange B, Knisz J, Steger K, Dubois-Dauphin M, Krause KH (2004) NOX3, a superoxide-generating NADPH oxidase of the inner ear. J Biol Chem 279(44):46065–46072 57. Banfi B, Molnar G, Maturana A, Steger K, Hegedus B, Demaurex N, Krause KH (2001) A Ca(2+)-activated NADPH oxidase in testis, spleen, and lymph nodes. J Biol Chem 276(40):37594–37601 58. Cheng G, Cao Z, Xu X, van Meir EG, Lambeth JD (2001) Homologs of gp91phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269(1–2):131–140 59. Schroder K, Weissmann N, Brandes RP (2017) Organizers and activators: cytosolic Nox proteins impacting on vascular function. Free Radic Biol Med 109:22–32

143 60. Edens WA, Sharling L, Cheng G, Shapira R, Kinkade JM, Lee T, Edens HA, Tang X, Sullards C, Flaherty DB et al (2001) Tyrosine cross-linking of extracellular matrix is catalyzed by Duox, a multidomain oxidase/peroxidase with homology to the phagocyte oxidase subunit gp91phox. J Cell Biol 154(4):879–891 61. Wong JL, Wessel GM (2008) Renovation of the egg extracellular matrix at fertilization. Int J Dev Biol 52(5–6):545–550 62. Thein MC, Winter AD, Stepek G, McCormack G, Stapleton G, Johnstone IL, Page AP (2009) Combined extracellular matrix cross-linking activity of the peroxidase MLT-7 and the dual oxidase BLI-3 is critical for post-embryonic viability in Caenorhabditis elegans. J Biol Chem 284(26):17549–17563 63. Grasberger H (2010) Defects of thyroidal hydrogen peroxide generation in congenital hypothyroidism. Mol Cell Endocrinol 322(1–2): 99–106 64. Takemoto D, Tanaka A, Scott B (2007) NADPH oxidases in fungi: diverse roles of reactive oxygen species in fungal cellular differentiation. Fungal Genet Biol 44(11):1065–1076 65. Zhang X, Krause KH, Xenarios I, Soldati T, Boeckmann B (2013) Evolution of the ferric reductase domain (FRD) superfamily: modularity, functional diversification, and signature motifs. PLoS One 8(3):e58126 66. Shatwell KP, Dancis A, Cross AR, Klausner RD, Segal AW (1996) The FRE1 ferric reductase of Saccharomyces cerevisiae is a cytochrome b similar to that of NADPH oxidase. J Biol Chem 271(24): 14240–14244 67. Rinnerthaler M, Buttner S, Laun P, Heeren G, Felder TK, Klinger H, Weinberger M, Stolze K, Grousl T, Hasek J et al (2012) Yno1p/ Aim14p, a NADPH-oxidase ortholog, controls extramitochondrial reactive oxygen species generation, apoptosis, and actin cable formation in yeast. Proc Natl Acad Sci U S A 109(22):8658–8663 68. Rossi DCP, Gleason JE, Sanchez H, Schatzman SS, Culbertson EM, Johnson CJ, McNees CA, Coelho C, Nett JE, Andes DR et al (2017) Candida albicans FRE8 encodes a member of the NADPH oxidase family that produces a burst of ROS during fungal morphogenesis. PLoS Pathog 13(12):e1006763 69. Hajjar C, Cherrier MV, Dias Mirandela G, Petit-Hartlein I, Stasia MJ, Fontecilla-Camps JC, Fieschi F, Dupuy J (2017) The NOX Family of proteins is also present in bacteria. mBio 8(6) 70. Spencer NY, Engelhardt JF (2014) The basic biology of redoxosomes in cytokine-mediated signal transduction and implications for disease-specific therapies. Biochemistry 53(10): 1551–1564 71. Henzler T, Steudle E (2000) Transport and metabolic degradation of hydrogen peroxide in Chara corallina: model calculations and measurements with the pressure probe suggest transport of H(2)O (2) across water channels. J Exp Bot 51(353):2053–2066 72. Segal AW (2016) NADPH oxidases as electrochemical generators to produce ion fluxes and turgor in fungi, plants and humans. Open Biol 6(5) 73. Magnani F, Nenci S, Millana Fananas E, Ceccon M, Romero E, Fraaije MW, Mattevi A (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci U S A 114(26):6764–6769 74. Wu JX, Liu R, Song K, Chen L (2021) Structures of human dual oxidase 1 complex in low-calcium and high-calcium states. Nat Commun 12(1):155 75. Sun J (2020) Structures of mouse DUOX1-DUOXA1 provide mechanistic insights into enzyme activation and regulation. Nat Struct Mol Biol 27(11):1086–1093

Part II Canonical NADPH Oxidases

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

10

Thomas L. Leto and Miklós Geiszt

Abstract

Keywords

NOX1 was the first of the NOX family of NADPH oxidases described with sequence homologous to the well-known microbicidal NADPH oxidase of phagocytes. NOX1 not only shows close structural similarities with the phagocytic NOX2-based prototype, but also exhibits striking functional similarities as a regulated multicomponent enzyme complex. Early studies explored proposed roles for NOX1 ranging from reactive oxygen species (ROS)-related signaling functions in responses to activated oncogenes, growth factors, and vascular agonists to voltage-gated proton transport. However, most current evidence supports notions of NOX1 functioning primarily in roles related to mucosal innate immunity, particularly in the colon epithelium where it exhibits its highest expression. Like its phagocytic counterpart, it acts as a tightly regulated ROS generator responsive to a variety of microbial patterns and induced by inflammatory cytokines. Defects in NOX1 in humans are associated with inflammatory bowel disease, consistent with roles of NOX1 related to innate immunity. Recent cancer transcriptomic analysis has not supported proposed links between high NOX1 expression and RAS mutations or cancer progression. This review provides an historical account of research developments on the NOX1-based NADPH oxidase and offers critical perspectives on these findings in the broader context of the redox biology field.

NOX1 · NOX · NADPH oxidase · Innate immunity · Inflammation · Cancer

T. L. Leto (✉) Laboratory of Clinical Immunology and Microbiology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA e-mail: [email protected] M. Geiszt Faculty of Medicine, Department of Physiology, Semmelweis University, Budapest, Hungary e-mail: [email protected]

1

Introduction

The initial discovery of NOX1 was based on searches for proteins with sequence similarities with that of gp91phox, the core component of the phagocyte NADPH oxidase (later renamed NOX2) [1]. The identification of other sequencerelated NOX family isozymes soon followed, bringing the total number of human NOX family genes to seven, including NOX1-NOX5, DUOX1, and DUOX2 (see reviews [2–4] and Chap. 9 by A. van der Vliet). Several of these were discovered based on leads obtained from rapidly expanding human gene sequence databases that became available by the late 1990s. These were exciting times as they coincided with other important developments in the emerging field of redox biology. Prior to that, the robust NOX2-based NADPH oxidase of phagocytic blood cells was well recognized as a critical source of reactive oxygen species (ROS) produced as toxic agents to fight microbial infection. Deficiencies in phagocytic oxidase activity were known to cause chronic granulomatous disease (CGD) resulting from genetic defects in any one of several essential NOX2 components, leading to enhanced susceptibility to bacterial and fungal infection and dysregulated inflammation (see Chap. 32 by M.J. Stasia and D. Roos). Aside from serving the dedicated purpose of oxidant-based microbial killing by phagocytes, ROS were otherwise widely recognized as by-products of oxidative metabolism, where the incidental generation of oxygen free radicals was considered mainly in the context of pathways leading to cell and tissue damage. By the late 1990s, there was a newfound appreciation of ROS serving as cellular second messengers that could act through signaling pathways stimulated by growth factors,

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_10

147

148

T. L. Leto and M. Geiszt

hormones, vascular agonists, or malignant transformation of cells [5–7]. Thus, the notion that NOX1 (originally called mox1, or ‘mitogenic oxidase 1’ [1]) could act as a ‘deliberate’ generator of ROS with proposed roles in mitogenesis was enthusiastically embraced particularly in the areas of cancer and vascular biology research. The field moved rapidly, as NOX1 has been one of the most extensively studied of all the non-phagocytic NADPH oxidases. Some 20 years later, research investigations on NOX1 continue with sustained interests in its potential roles in cancer and inflammatory disease processes of vascular, central nervous, and gastrointestinal systems. Indeed, the search for specific NOX inhibitors as compounds with therapeutic potential for treating cancer and vascular disease continues (see Chap. 21 by C.M. Dustin, E. Cifuentes-Pagano, and P.J. Pagano), although it remains uncertain whether any of the nonphagocytic NOX homologues produce ROS at levels comparable to that generated by NOX2 in inflammatory cells. This chapter provides a historical account of major developments in our basic understanding of NOX1 functions with a particular emphasis on roles of NOX1 in disease pathogenesis, while offering some critical perspectives on past and current directions of this research.

2

NOX1, the Close Cousin of the NOX2-Based Oxidase of Phagocytes

2.1

Structural Features of the NOX1-Based Flavocytochrome

Human NOX1 shows 56% amino acid sequence identity with that of gp91phox (a.k.a. NOX2) with corresponding total sequence lengths of 564 and 570 amino acids, respectively [1]. The structural basis of NOX2 function was well understood at that time, and it was clear that NOX1 shares common topographical features considered essential for transmembrane electron transport and superoxide generation (Fig. 10.1). NOX1 exhibits its highest structural conservation with sites within NOX2 involved in the binding of cytosolic NADPH and FAD in its C-terminal domain dehydrogenase domain; the N-terminal domain of NOX1 also comprises 6 hydrophobic, membrane-spanning helical segments, which include 4 conserved histidine residues in helix 3 and 5 required for ligation of two hemes. The importance of these conserved sites was reaffirmed recently once the X-ray crystallographic structure of NOX5 was described [8]. All NOX family NADPH oxidases share these minimal structural features required for binding of NADPH to the cytoplasmic dehydrogenase domain, and one electron transfer from NADPH to FAD (semiquinone and FADH2) and two membrane-imbedded Fe+3 hemes, ultimately donating them to molecular oxygen to produce superoxide anion.

Shortly after the description of mox1, a shorter “alternatively-spliced” mRNA derived variant of NOX1 was reported called NOH-1S, which when heterologously expressed appeared to exhibit properties similar to voltage-gated proton channels [9]. The cDNA of NOH-1S encodes a protein 191 amino acids in length encompassing only the first 4 hydrophobic transmembrane segments (I–IV) fused with a sequence derived from the 3′ untranslated region. The fusion appeared to involve intraexonic sites that lack appropriate splice donor or acceptor sequences. Follow up studies examining NOX1 transcripts in detail could not authenticate the origins of NOH-1S as a native spliced variant of NOX1 [10, 11]; rather, NOH-1S was shown to be a product synthesized in vitro solely by low-temperature AMV reverse transcriptase reactions involving template switching between repetitive sequences that flank the proposed splice sites. This work also confirmed production of only one major alternatively spliced mRNA variant of NOX1, called NOX1 Lv (a.k. a., NADPH oxidase 1, isoform 2), that lacks critical exon 11 sequence encoding conserved FAD-binding residues [10]. Deletion of exon 11 resulted in the loss of NADPH oxidase activity. It remains unclear whether this inactive variant prevalent in healthy human colon and colon epithelial cancer cell lines can affect the activity of full-length NOX1. Banfi et al. [12] later acknowledged the artifactual basis of NOH-1S cDNA synthesis in vitro, and interests in NOX isozymes functioning as proton transporters were abandoned once the identity of a distinct human gene encoding the voltage-sensitive proton channel, HV1, was described in 2006 [13]. The initial excitement over NOX1 functioning as a pro-mitogenic ROS generator capable of inducing tumor formation in transfected NIH-3T3 fibroblasts was dampened when the authors of these earlier findings revealed that they were unable to reproduce these findings in cells transfected with NOX1 alone [2]; it was further noted that they detected mutant oncogenic KRas-Val12 in the original NIH-3T3 cell clones used to demonstrate tumor formation as well as pro-mitogenic and pro-angiogenic activities that were previously attributed to transfected Mox1 alone [14, 15]. These revelations coincided with findings by other groups described below showing that NOX1 oxidase activity is strictly dependent on several cytosolic regulatory components related to those already known as critical in supporting activity of the NOX2-based enzyme.

2.2

NOX1: A Regulated Multicomponent Oxidase Complex

NOX1 was initially noted for its high expression in the colon, with lower transcript levels detected in prostate, uterus, and vascular smooth muscle [1]. Continued bioinformatics searches eventually identified expressed sequence tags

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

149

NOX2 (gp91phox)

NOX1

Heme

Heme

p22phox

p22phox

Heme RAC1 or 2

FAD p47phox NADPH

p67phoxx

p40phox

Heme FAD NOXO1 N NADPH

RAC1

NOXA1

Fig. 10.1 Schematic representation of components of the NOX2- and NOX1-based NADPH oxidase complexes. The core catalytic flavocytochrome b component of each oxidase (dark blue) has binding sites for NADPH, FAD, and two membrane imbedded heme moieties involved in electron transport and superoxide generation. Defects or

deficiencies of any of the phagocytic NOX2 components shown result in chronic granulomatous disease. Full activation of the NOX1-based oxidase also involves RAC1 and closely related homologues of p47phox and p67phox, known as NOX organizer 1 (NOXO1) and NOX activator 1 (NOXA1), respectively

(ESTs) homologous to the essential cytosolic regulators of NOX2 which were designated as NOX organizer 1 (NOXO1; p47phox homologue) and NOX activator 1 (NOXA1; p67phox homologue) [16–18; also see Chap. 15 by P.M. Dang and J. El-Benna and Chap. 16 by H. Sumimoto, A. Kohda, J. Hayase, and S. Kamakura]. In situ hybridization with antisense probes of NOXO1 and NOXA1 transcripts revealed prominent expression in mouse colon epithelial cells. Human NOXO1 and NOXA1 transcripts were detected in abundance in colon, although the more widespread expression pattern of human NOXA1 suggested its functions are not limited to supporting NOX1 activation in humans (Fig. 10.2) [17]. NOX1 reconstituted cell model studies by several groups confirmed that a similar hierarchy of interactions between NOX1 components dictates their subcellular targeting, assembly, and activation in a manner analogous to what was understood about NOX2 function (Fig. 10.1; see reviews [20, 21]). NOX1 forms a stabilized heterodimeric complex with p22phox that allows it to escape beyond the endoplasmic reticulum and undergo maturation into an active superoxide generating enzyme on the plasma membrane, where it presents a docking site for interactions with the cytosolic regulators [22, 23]. NOXO1 serves as a modular adaptor molecule that interacts with high affinity with the C-terminal, proline-rich tail of p22phox and membrane phosphatidic acid (PA) and phosphoinositide (PI) lipids, through its two SH3 domains and PX domain, respectively [24]. It also bridges interactions between the membrane and NOXA1 by binding to the SH3 domain of NOXA1 through

its proline-rich, C-terminal sequence. Homologous domain interactions were already recognized as critical in the assembly of the NOX2 complex such that p47phox-deficient CGD patients fail to translocate p67phox to the membrane in activated neutrophils. The N-terminal tetratricopeptide repeats (TPRs) of NOXA1 serve as a scaffold that presents RAC1-binding sequence [23, 25, 26]. NOXA1 also has a conserved ‘activation domain’ sequence motif immediately following the TPRs as seen within p67phox. Initial demonstrations of regulation of NOX1 by RAC1 were challenging because most reconstituted cell models have a significant pool of active GTP-bound RAC1, unlike phagocytes, where RAC-GDP is maintained in an inactive state in a one-to-one complex bound to Rho GDP Dissociation Inhibitor (Rho-GDI) [27]. The best evidence supporting the dependence of NOX1 activity on RAC1 included experiments demonstrating effects of overexpressed active or inactive forms of RAC1, RAC1 silencing, and the effects of mutating sites that inhibit interaction of RAC1 and NOXA1 [23, 25, 26]. Similar approaches revealed regulation of NOX3 by RAC1, although NOX3 is less dependent on RAC1 and NOX activator proteins than NOX1 or NOX2 isozymes [23]. The close structural and functional similarities between NOX1- and NOX2-based oxidases were further demonstrated by studies showing that cytosolic components of either oxidase can cross-function in supporting oxidase activity of either NOX1 or NOX2 enzymatic core component when co-expressed in heterologous reconstituted cell models [28]. Some studies suggest mixed NOX1 and NOX2

150

T. L. Leto and M. Geiszt

NOX1

NOXO1

NOXA1

Fig. 10.2 Tissue-specific expression patterns of NOX1 components showing high colon-specific expression of NOX1 and NOXO1 transcripts, but more widespread expression of NOXA1, suggesting alternative functions beyond NOX1 activation. Data obtained through

HPA RNA-seq transcriptomics of 27 normal human tissues available through the Human Protein Atlas (www.proteinatlas.org/about/down load) [19]

component-based oxidases function in certain cell types, (i.e., vascular or immune cells) [29–31] In general, NOX2 is a more tightly regulated oxidase complex than NOX1 as it appears that phagocytic cells maintain NOX2 components in a latent, dissociated state until its assembly and robust activation is triggered for the purpose of microbial killing following phagocytosis. In contrast, human NOX1 exhibits significant constitutive oxidase activity which is enhanced 1.5- to 3-fold by treatments that stimulate protein kinase C

(PKC) activity [17, 18]. Key differences in the two enzymes that account for this can be summarized as follows: (1) The core flavocytochrome of NOX2 is stored in granule membranes in unstimulated cells, whereas NOX1 is found on the plasma membrane and appears to cycle with endomembrane compartments [32, 33]. (2) The polarized structure of p47phox maintains this protein in a ‘closed’ conformation, due to electrostatic and SH3 domain interactions with its C-terminal autoinhibitory region, while

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

bound in a complex with other cytoplasmic components [21]. Hyperphosphorylation of p47phox following cellular activation is needed to expose the SH3 domains, enabling p47phox translocation to the membrane through interactions with p22phox. In contrast, NOXO1 is not as extensively phosphorylated as p47phox; weak intramolecular contacts have been detected involving the SH3 domains and PX domain of NOXO1 [24], although it lacks an autoinhibitory region seen in p47phox and is detected on the plasma membrane even in unstimulated cells [23, 26]. Phosphorylation of NOXO1 by protein kinase C (PKC) in response to phorbol 12-myristate 13-acetate (PMA) stimulation at Ser-154 and Thr-341 supports higher NOX1 oxidase activity by promoting interactions of NOXO1 with p22phox and NOXA1 [34, 35]. PKC-beta-mediated phosphorylation of NOX1 promotes its interaction with NOXA1 [36], however, protein kinase A (PKA) and mitogen-activated protein kinase (MAPK) mediated phosphorylation of NOXA1 suppresses its interaction with NOXO1 resulting to lower NOX1 activity [37, 38] (3) Membrane assembly of NOX1- and NOX2-based cytosolic oxidase components also involves PX domain recognition of distinct membrane phospholipids. The PX domains of NOXO1 and p47phox have two lipid binding pockets: one conserved pocket in both proteins binds PA, while the other favors binding of distinct phosphoinositide lipids to p47phox (PI(3,4)P2) and NOXO1 ((PI(4)P, PI(5)P, PI(3,5)P2) [39, 40] (reviewed in [20]). Alternatively spliced PX domain variants of NOXO1 exhibit different lipid binding specificities with the most prevalent spliced variants (β and γ) favoring subcellular targeting to the plasma membrane [40–42]. Binding and retention of the NOX2 cytosolic complex on phagosomes is attributed to the p40phox PX domain binding affinity for PI(3)P [43, 44], whereas the NOX1 enzyme has no homologous counterpart to p40phox.

2.3

Other NOX1 Regulators

Another interacting partner that supports NADPH oxidase activities of both NOX1 and NOX2 is peroxiredoxin 6 (PRDX6), which is a multi-functional enzyme exhibiting glutathione peroxidase as well as phospholipase A2 and lysophosphatidyl choline (lysoPC) acyltransferase activities [45]. All three activities promote oxidized membrane repair by eliminating phospholipid hydroperoxides. However, the phospholipase activity of PRDX6 appears to be critical for NOX2 activation [46, 47]. Here, generation of lysoPC, which converts to lysoPA, then acts through lysoPA receptor signaling in pulmonary endothelial and alveolar macrophages [48]. In the NOX1 complex PRDX6 binds to and stabilizes NOXA1 and, when overexpressed in colon epithelial cells, both its phospholipase A2 and peroxidase activities support higher NOX1 oxidase activity as well as NOX1-dependent

151

epithelial cell migration [49]. Thus, this enzyme serves roles in antioxidant defense and membrane lipid turnover and repair, while also supporting higher ROS generation by NOX1 or NOX2. As an intracellular target of hydrogen peroxide, the PRDX6 associated with NOX1 or NOX2 enzyme complexes may relay intracellular redox signals more efficiently from these direct sources of ROS.

3

NOX1 Functions in Innate Immunity

Abundant expression of NOX1 and its cytosolic regulators in colon epithelium suggests a potential role of NOX1 in innate immunity. NOX1 mRNA was initially detected in lower portions of mouse colon crypts and was later confirmed to be restricted to proliferating colon stem cells [28, 50]. However, it appears that highest NOX1 protein levels accumulate at the apical aspect of polarized, mature, differentiated colon epithelial cells exposed at the mucosal surface of normal human, mouse, and guinea pig large intestine [51– 54]. Human and mouse NOX1 expression increases along the length of large intestine from proximal to distal sites, correlating with increased microbial burden in the colon [31, 50]. A growing body of evidence has steadily emerged further supporting roles for NOX1 in colon innate immune functions in interactions with the gut microbe. This includes induction of NOX1 by cytokines and other inflammatory mediators, NOX1 induction and activation in responses to microbial patterns, NOX1 participation in colon epithelial restitution and homeostasis (proliferation, cell migration and wound healing), and the identification of NOX1 lossof-function (LoF) genetic variants associated with human inflammatory bowel disease (IBD). NOX1 transcript levels in human colon epithelial lines are induced markedly by interferon-γ, resulting in enhanced superoxide generation, effects correlated with identification of an interferon-γ response element within the NOX1 promoter that binds to STAT1 [28, 55]. This treatment causes growth arrest along with induced differentiation of these cells, suggesting these higher levels of NOX1 are not functionally linked to mitogenesis. Interleukin-1-beta (IL-1β) and tumor necrosis factor-alpha (TNF-α) treatments also induce NOX1 activity in colon epithelial cells; the effects of TNF-α were attributed to AP-1 binding sites within the NOXO1 promoter [56]. Later studies in human colon epithelial cells and IL-10 knock-out (KO) mice indicated that this antiinflammatory cytokine acts by inhibiting interferon-γ induction of NOX1 and TNF-α induction of NOXO1 [57]. Other recent work showed TNF-α and IL-17 together induce NOXO1 and higher NOX1 activity which increases colon epithelial production of the antimicrobial factor lipocalin2 [58]. NOX1 also is modestly induced by treatment of colon epithelial tumor lines with Th2 cytokines (IL-4 or

152

IL-13) through a pathway involving Janus kinase-signal transducer and activator of transcription (JAK1/STAT6) signaling, STAT3 activation and its binding to the NOX1 promoter [59]. Early studies highlighted NOX1 induction in guinea pig gastric pit cells following infection with Helicobacter pylori, where lipopolysaccharide (LPS) was shown to act through Toll-like receptor 4 (TLR4) to trigger excess ROS generation [60]. The same group showed recombinant flagellin from Salmonella enteritidis also enhances NOX1 expression and ROS production through TLR5 in colon epithelial cells expressing NOXO1 and NOXA1, leading to enhanced IL-8 release [61]. Despite the high expression of NOX1 in normal mouse colon, initial studies describing Nox1 KO mouse models did not report any profound colon phenotypes such as spontaneous intestinal inflammation or enhanced susceptibility to microbial infection (see reviews [62, 63]). NOX1 is likely a significant contributor to overall redox homeostasis in the colon, since colitis induced by the excess oxidative stress in glutathione peroxidase 1 and 2 double-knockout mice is attenuated when these mice are crossed with Nox1 knockout mice [64]. NOX1-deficient mouse colons showed a shift with increased numbers of goblet cells that secrete mucins and a corresponding reduction in colonocytes, suggesting the absence of NOX1-derived ROS results in suppressed phosphatidylinositol-AKT-Wnt-beta-catenin and Notch1 signaling that affects differentiation from colon progenitor cells [65]. NOX1-/- and Il-10-/- double knockout mice show increased spontaneous colitis which was attributed to excess endoplasmic reticulum stress [66]. Studies using intestinal epithelial cell-targeted Nox1 knockout (Nox1IEC-/-) mice showed impaired epithelial restitution in vivo in response to scratch wounding which was attributed to decreased epithelial cell proliferation and migration in the absence of NOX1 [67]. A similar defective intestinal wound healing phenotype was observed in formyl peptide receptor1deficient (Fpr1-/-) mice [68]. Together, these findings suggest optimum colon epithelial barrier function and restitution involves a FPR1-NOX1-ROS-dependent pathway that is stimulated by formyl peptide presented by commensal Lactobacillus species in the gut microbiome. Annexin-1, another endogenous Fpr1 agonist released during wounding, also appears to promote NOX1-dependent epithelial restitution [69]. Another study using a complete Nox1 knockout mouse model showed that epithelial repair in response to dextran sulfate sodium (DSS)-induced colitis is impaired due to diminished cell proliferation, survival, migration, and terminal differentiation; enhanced NOX1 expression and ROS production was detected during the epithelial restitution process in response to DSS treatment in control wild type animals [70]. Finally, recent work focused on NOX1 transcript in colon stem cells in vivo and in vitro (maintained in spheroids) showed that microbial ligands for TLR2, TLR4,

T. L. Leto and M. Geiszt

and TLR5 induce NOX1 expression and NOX1-dependent stem cell proliferation through epidermal growth factor receptor (EGFR) transactivation [50]. Thus, NOX1 through interactions with the gut microbiome or other wound healing mediators appears to support colon epithelial homeostasis and barrier functions in mice. Beyond the role of NOX1 in colon epithelium restitution, recent work comparing the effects of LPS in wild type versus NOX1-deficient mice indicated both Nox1 and inducible nitric oxide synthetase (NOS2 or iNOS) are co-induced in macrophages, which together appear to represent a significant source of peroxinitrite production (generated by superoxide, formed by Nox1, reacting with nitric oxide, formed by iNOS) in response to endotoxemia [71]. Following intraperitoneal LPS administration, NOX1 and 3-nitrotyrosine (reflecting peroxinitrite production) were detected in macrophages in the gut lamina propria. Nox1-/- mice failed to induce Nos2 or activate matrix metallopeptidase 9 (MMP9), and showed diminished gut hyperpermeability responses to LPS, confirming that iNOS induction is dependent on NOX1. Bone marrow transplantation experiments in chimeric WT/Nox1-deficient animals confirmed that macrophage NOX1 and NOS2 mediate these gut permeability effects of endotoxemia in mice. Furthermore, early studies using in situ hybridization-based detection of NOX1 transcript in healthy and diseased human colon tissues also found an enrichment of NOX1 positive lymphocytes within ulcerative colitis and Crohn’s disease lesions [31].

4

NOX/DUOX Deficiencies in Inflammatory Bowel Disease (IBD)

4.1

Human NOX1 Deficiencies in IBD

Deficiencies in three NOX family isoforms have been associated with inflammatory bowel disease in humans, including NOX1, NOX2, and DUOX2, all of which appear to serve innate immune or barrier functions in the GI tract (see recent reviews [62, 63, 72, 73]). Partial or complete NOX1 deficiencies associated with IBD have been described by several groups. Most cases occur as very early onset IBD (VEO-IBD) in males, consistent with the X-chromosome location of the NOX1 gene. In one large study, several defective NOX1 missense variants associated with IBD were characterized in several systems, including primary biopsy tissue, colon organoids, and heterologous NOX1reconstituted cell models [74]. Several variants are rare and exhibit significant losses of NADPH oxidase activity in comparison to wild-type NOX1, although one exception is a common NOX1 p.D360N variant exhibiting only partial LoF (oxidase activity 50–75% of WT NOX1). This variant was linked to disease only in male Ashkenazi Jewish

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

populations but is otherwise tolerated in healthy male European populations with a high allelic frequency (0.0266), while showing a similar frequency in European male IBD subjects. Functional studies characterizing NOX1 p.D360N and p.P330S variants in reconstituted COS-7 cells showed enhanced microbial cell invasion by the enteric pathogen, Campylobacter jejuni, in comparison with WT NOX1 reconstituted cells [75]. Another partially active NOX1 hemizygous p.R241C variant was linked to IBD in a patient who was also homozygous for a CYBA (p22phox) p.Y72H variant [76]. Cells expressing these NOX1 and p22phox variants together exhibited significantly lower superoxide generation than WT NOX1/p22phox-expressing control cells and were more susceptible to invasion by Listeria monocytogenes; here, the lower oxidase activity was correlated with diminished nucleotide-binding oligomerization domain-containing protein 2 (NOD2) signaling, nuclear factor kappa-lightchain-enhancer of activated B cells NF-κB activation, and IL-8 release in response to the microbial NOD2 ligand, muramyl dipeptide. Recent studies on a truncated NOX1 variant associated with IBD (p.R54*) showed the NOX1 null cells exhibited reduced migration in response to microbial-derived N-formyl-methionine-leucyl-phenylalanine (fMLF) or EGF in comparison with cells expressing full-length WT NOX1 [77]. In summary, NOX1 appears to serve signaling functions in colon epithelial homeostasis, as well as innate antimicrobial defense functions in humans that appear to be compromised with a variety of human NOX1 genetic variants associated with IBD. Population-based studies suggest these variants can be tolerated without IBD manifestations, suggesting the NOX1 defects have a low disease penetrance that is likely influenced by compensating roles of other NOX-based ROS generators or scavengers in the colon mucosal epithelium, as well as a variety of other environmental, dietary, and genetic factors.

4.2

Human NOX2 Deficiencies in IBD

A role for NOX2 deficiency in IBD has been appreciated for some time, since Crohn’s-like colitis symptoms have been a well-recognized clinical feature of CGD observed in about 40–50% of patients [62, 72, 73]. The basis for this disease phenotype in this subset of NOX2-deficient patients is not well understood, but may relate to other proinflammatory complications of CGD, including arthritis and lupus-like autoimmune symptoms. The overall survival of CGD patients has been closely correlated with residual oxidase function, however, the IBD phenotype in CGD patients is not tightly correlated with oxidase loss of function [78]. The burden of other established IBD-related risk alleles is higher in CGD patients with IBD than in those without IBD [79]. Among CGD patients, those with IBD also show a

153

higher frequency of infections, including enteric pathogens, although overall survival among CGD patients with IBD is not different from those without IBD [80]. Several heterozygous hypomorphic variants of NOX2 components were also associated with IBD in patients lacking any manifestations of enhanced susceptibility to microbial infections characteristic of CGD, and some of these variants were shown to affect protein-protein interactions between NOX2 oxidase components [81]. IBD symptoms in affected CGD patients are resolved following allogenic hematopoietic stem cell transplantation, therefore NOX2 deficiency appears to be an independent genetic risk factor for IBD in humans attributable to NOX2 defects in the hematopoietic compartment [82]. It remains unclear how NOX2 deficiency favors the enhanced inflammation observed only in a subset of CGD patients with IBD. Neutrophils are present in the colon mucosa in high numbers particularly under pro-inflammatory conditions. Their capacity for excess NOX2-derived ROS generation can impact intestinal health in several ways, either by serving protective innate antimicrobial (microbicidal) functions, promoting inflammation through enhanced oxidant-mediated host tissue damage, or even contributing to the resolution of inflammation by creating a hypoxic environment [62, 83]. Furthermore, NOX2 expressed in lymphoid and dendritic cells could affect adaptive immune functions related to IBD pathology in CGD.

4.3

Human DUOX2 Deficiencies in IBD

The dual oxidases, DUOX1 and DUOX2, were first described in thyroid tissue and were proposed to serve as sources of hydrogen peroxide supporting thyroperoxidasemediated thyroid hormone biosynthesis [84, 85] (see Chap. 14 by F. Miot and X. De Deken). Many defective human DUOX2 variants have since been linked to severe or transient congenital hypothyroidism [86]. Geiszt et al. [87] detected high levels of DUOX1 and DUOX2 expression in mucosal and exocrine tissues and proposed a role related to host defense, suggesting DUOX derived hydrogen peroxide could support the antimicrobial activity of lactoperoxidase in exocrine secretions. DUOX2 expression is particularly high in the human rectum, where the microbial burden is greatest [87]. Several groups demonstrated that DUOX isozymes are critical in suppressing enteric infections in Drosophila, Caenorhabditis elegans (C. elegans), and zebrafish [88– 90]. Several rare human DUOX2 missense variants were detected in IBD patients by whole exome sequencing that exhibit compromised oxidase function in DUOX2/DUOXA2 reconstituted cell models and, in some cases, cells reconstituted with these defective variants appear more susceptible to microbial invasion [75, 91]. DUOX2 is among the most inducible genes detected in IBD patients and in mice

154

T. L. Leto and M. Geiszt

with intestinal dysbiosis [92, 93]. A recent large populationbased study detected correlations between elevated plasma levels of IL-17c and detection of mutant DUOX2 variants [94]. This association was further validated in a casecontrolled study of IBD patients who exhibit an increased DUOX2 mutation burden along with elevated plasma IL-17c levels and increase ileal expression of IL-17c. Thus, elevated plasma IL-17c levels were proposed as a useful biomarker of mucosal dysbiosis related to abnormal microbiota-host mucosal interactions associated with rare DUOX2 variants that predict IBD risk in humans.

4.4

Mouse Models of IBD Related to NOX1, NOX2 and DUOX2 Deficiency

Studies investigating colitis susceptibility in various NOX1-, NOX2-, and DUOX-deficient mouse models have provided limited insight on the mechanistic basis of intestinal inflammatory phenotypes observed in human IBD patients with these oxidase defects. Unfortunately, none of the NOX-deficient models maintained in specific pathogen-free facilities develop colitis spontaneously, which may reflect the absence of important environmental and dietary factors that contribute to IBD in humans. Experimental outcomes exploring colitis susceptibility can vary depending on the NOX component knockout or methods for colitis induction (i.e., DSS- or 2,4,6-trinitrobenzene sulfonic acid-induced, or infectious challenges) [See reviews: 62, 63, 73]. Variables related to animal care facility conditions, diets, strains or substrains of knockout lines, breeding and housing strategies, etc., are difficult to decipher in accounting for discrepancies between studies. Moreover, extrapolating the observations from mice to human disease could be misleading when considering how mice and humans have diverged and coevolved with their respective microbiomes. Duox2-deficient mice have severe congenital hypothyroidism, requiring thyroid hormone supplementation in any studies exploring potential roles of Duox2 in mucosal tissue host defense. Despite these challenges, studies exploring adaptations in gut microbiome composition in response to various NADPH oxidase deficiencies have revealed remarkable complexities in hostmicrobiome interactions that can profoundly influence intestinal health. Duox2 intestinal epithelial expression in mice is enhanced by exposure to normal gut microbiota or pathogens, and at sites of active ulcerative colitis [95– 98]. Duoxa1/2-deficient mice on thyroid hormone supplementation are more susceptible to Helicobacter felis infection [96]. Several mouse models with NOX1 deficiency exhibit changes in cell populations in the mucosal epithelium (i.e., enhanced goblet cell numbers) along with significant alterations in gut microbiome composition and susceptibility to infection [65, 99, 100]. One study examining effects of

Nox1 or Nos2 knockouts on mouse gut ROS generation suggested that NOX1-derived superoxide together with NOS2-derived nitric oxide produce enough microbicidal peroxynitrite to affect the composition of the ileal microbiome, which in Nox1 and Nos2 KO mice more closely resembles the normal caecum microbiome [99]. The authors concluded both ROS generators act together in controlling microbial burden and reflux in the ileum. Some mouse models of CGD caused by disruption or defects in Cybb, Ncf1, Ncf4, or Rac2 genes showed enhanced susceptibility to induced colitis, along with increased bacterial translocation from the gut, consistent with expectations related to NOX2 innate immune dysfunction in CGD phagocytes (see reviews [62, 73]). Ncf1-/- mice are more susceptible to DSSor Citrobacter rodentium-induced colitis, which was accompanied by changes in host inflammatory signatures (i.e., enhanced IL-6, IL-10, IL-17A, TNF-α, and IFN-γ levels) [101, 102]. In one study, the effects of p47phox/ Ncf1 deficiency were not reversed by NOX2 restoration in chimeric mice transplanted with wild-type bone marrow; these findings were explained by the alterations in the gut microbiome composition detected in Ncf1-/- mice [101]. In other studies, disrupting global Cyba expression, affecting the p22phox partner of several oxidases (NOX1-4), surprisingly produced mice that appeared resistant to C. rodentium or L. monocytogenes infection [100]. Mice with complete Cyba or epithelial-targeted Cyba disruption produced similar results, suggesting that the basis for resistance to infection by these gut pathogens is attributable to epithelial NOX1 deficiency. These unexpected findings correlated with significant gut microbiome alterations, including a significant enrichment in probiotic, hydrogen peroxide-producing Lactobacillus species. The authors proposed a model of ‘defensive mutualism’ in which the overgrowth of these hydrogen peroxide producing bacterial species associated with epithelial NOX1 deficiency was suggested to compensate for the host ROS generation defect, which would thereby directly impact virulence pathways of specific enteric pathogens. However, alternative interpretations were proposed to account for the enhanced resistance of NOX1-deficient mice to this pathogen, emphasizing unique capabilities of C. rodentium to not only tolerate, but favorably adapt to, high levels of hydrogen peroxide on the epithelial surface [103]; here, it was suggested that the pathogen can adhere to surface epithelium using the type-III secretion system and respire using the host epithelial NOX1-derived hydrogen peroxide. Thus, in the absence of functional NOX1, the pathogen would lose its advantage over other anaerobes and fail to thrive in the mucosal epithelial microenvironment. In another study examining effects of a different hypomorphic Cyba variant in mice that reduces both NOX1 and NOX2 activities, but not that of NOX4, these mice exhibited a disrupted mucus layer, microbial penetration into crypts, and defective antimicrobial

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

defense, findings entirely different from those with epithelial NOX1 deficiency [104]. The above studies in NOX1-, NOX2-, and DUOX-deficient mice demonstrate an impressive range of unanticipated consequences arising from host NADPH oxidase defects that appear to reshape the microbiome is different ways. Regardless of the interpretations, or the limitations of extrapolating such findings to human disease, these findings raise interesting possibilities for the prospects of therapeutic interventions aimed at manipulating microbiome compositions in IBD and CGD patients, either with antibiotics or fecal transplantation. Another important finding uncovered in the studies on murine responses to C. rodentium enteric infection is that NOX1 also appears to serve a critical role in the induction of colon epithelial DUOX2 during infection, which would in effect amplify ROS release by the colon epithelium [100]. Both global and intestinal epithelial-targeted Cyba-/- mice fail to show robust Duox2 induction following infection. Intestinal epithelial-targeted IL-17RA and IL-17RC disrupted mice are also susceptible to C. rodentium infection and fail to induce NOX1 and Duox2 [105, 106]. A similar dependence of epithelial DUOX2 induction on NOX1 was observed in human colon epithelial cell lines such that NOX1 silencing appears to suppress induction of DUOX2 by IL-17 [54]. Thus, NOX1 deficiency in mice and humans may act as a double-hit affecting activities of both NOX1- and DUOX2-based epithelial oxidases. In summary, IBD susceptibility is influenced by a complex interplay of multiple factors, including genetic, environmental, diet, and host-microbial interactions. Genetic defects in at least three NOX family NADPH oxidases (NOX1, NOX2 and DUOX2) have been linked to IBD, which were suggested to result from partial or complete loss of NADPH oxidase function. Much remains to be learned about the interrelationship of multiple, seemingly redundant NADPH oxidases to better understand how inflammatory disease is manifested, particularly in cases of oxidase variants ranging from completely inactive to almost fully active. The low disease penetrance of several inactive variants may involve redox-based compensatory mechanisms either within the host or through altered host-microbiome interactions that maintain a proper balance required for intestinal health. The basis of inflammatory disease observed with heterozygous NOX or DUOX variants exhibiting partial loss of function is more difficult to understand in mechanistic terms. Induction of epithelial NOX1 is closely linked to DUOX2 and NOS2, in that induction of the latter two sources of microbicidal ROS depends on NOX1 activity, effectively amplifying any ROS

155

signal from NOX1. Furthermore, NOX1 and NOX2, as close structural homologues of epithelial and phagocytic compartments, respectively, share functional similarities, including some of the same modes of activation, responsiveness to common microbial patterns, and induction by some of the same cytokines. Figure 10.3 provides a broad hypothetical scheme depicting how the defective NADPH oxidases could impact their distinct yet closely interrelated roles in maintaining innate anti-microbial defenses and mucosal barrier functions which, if compromised, would lead to inflammatory bowel disease. The defective oxidases result in altered gut microbiome compositions that can be either favorable or unfavorable to the host. Unfavorable changes, frequently referred to as dysbiosis, could result in microbial invasion of the mucosal layer that normally excludes or repels bacteria, causing enhanced microbial sensing, local immune activation, and release of inflammatory cytokines and chemokines. Some of these cytokines provide a positive feedback signal to induce higher oxidase expression in response to oxidase loss of function, while some chemokines can act in recruiting inflammatory cells mobilized to fight infection, while potentially inflicting host tissue damage. Microbial translocation beyond the gut to more distal sites may trigger systemic immune activation and cytokine release that can, in turn, cause systemic inflammatory disease.

5

NOX1 in Other Systemic Inflammatory Diseases

The levels of NOX1 in the colon far exceed those detected in any other tissue (Fig. 10.2), however, investigations in the field of vascular biology have suggested roles for NOX1 in mice in the development of inflammatory disease processes in vascular tissues. These studies have been reviewed in detail elsewhere and are beyond the scope of this chapter [107–109]. For example, proinflammatory cytokines such as TNF-α and IL-1β activate NOX1 signaling in endosomes leading to NF-KB activation in vascular smooth muscle cells [33]. Several studies based, in large part, on observations comparing vascular phenotypes in NOX1deficient or transgenic NOX1 overexpressing mice versus WT mice have highlighted roles of NOX1 in hypertensive responses of vascular smooth muscle cells to angiotensin II, where NOX1-derived superoxide was proposed to limit the bioavailability of nitric oxide. NOX1- as well as NOX2deficient mice also appear to be protected against atherogenesis and ischemia-reperfusion injury (i.e., stroke), likely related to distinct but potentially similar roles in vascular

156

T. L. Leto and M. Geiszt

Fig. 10.3 Mechanistic scheme depicting the functional overlap of several gut NOX family NADPH oxidases, their interplay with the gut microbiome, and consequences of dysfunction leading to inflammatory bowel disease. All three oxidases are induced or activated by exposure to microbial patterns, have roles in shaping the gut microbiome, and are responsive to some of the same inflammatory cytokines produced

downstream of microbial invasion of or translocation beyond the gut mucosal epithelium. The interdependent, overlapping, and potentially redundant functions of both epithelial and hematopoietic oxidases may, in part, account for the observed low disease penetrance of genetic defects or deficiencies in any one NOX isozyme (see text for details)

and inflammatory cells. The extent to which the observations in mice pertain to vascular disease development in humans remains controversial [110]. To our knowledge, there have been no studies described to date linking any human NOX1 LoF genetic variants to altered (i.e., protective) vascular phenotypes.

that links of higher NOX1 expression to cancer are limited for the most part to intestinal tumors [54]. Thus, early studies suggesting links between EGFR stimulation, ROS generation, and proliferation in many cancer cell types could involve other NOX family oxidases or other modulators of intracellular ROS levels. For example, DUOX1 was shown to be activated and produces hydrogen peroxide downstream of EGFR stimulation of keratinocytes, whereas NOX2 is activated by EGFR stimulation in lung epithelial cells [112, 113]. Furthermore, NOX1 expression in mutant RAS vs. WT RAS containing intestinal tumors is only modestly increased (1.6-fold vs. 1.9-fold for mutant and WT KRAS tumors, respectively) in comparison with the adjacent normal colon tissues [54]. The same study examined a panel of some 89 established colorectal cancer (CRC) tumor cell lines and observed no significant difference between NOX1 transcript levels in tumor lines with WT vs. activated, mutant forms of RAS. Of the 27 cell lines from the American Type Culture Collection examined in detail, NOX1 transcript levels varied widely and only a fraction of these lines demonstrated detectable NOX1 protein or PMA-stimulated superoxide release. A monoclonal antibody demonstrating specificity for NOX1 confirmed that NOX1 protein levels closely correlated with NOX1 transcript levels in various tumor lines, as well as in normal and cancerous tissues [54]. The results indicated that

6

NOX1 in Cancer Revisited

Links between NOX1 expression and cell transformation, proliferation or cancer progression have been proposed and debated for more than two decades. As indicated above, initial reports linking ectopically overexpressed NOX1 to tumorigenesis, cell proliferation, and angiogenesis were flawed in that they failed to account for requirements for NOX1-supportive cofactors (NOXO1 and NOXA1) for oxidase activity [1, 14, 15] or the unexplained detection of oncogenic mutant Ras in the NIH-3T3 cloned lines studied [2]. Follow-up studies in other cell lines suggested growth factor receptors and oncogenic activated forms of RAS can induce NOX1 expression through a common Ras-Raf-MEKERK signaling pathway (see review [111]). However, recent NOX1 transcriptomic analysis of human primary tumor specimens in The Cancer Genome Atlas (TCGA) confirmed

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

157

Fig. 10.4 Comparison of NOX1 transcript expression patterns in differentiated (a, b) and advanced (c, d) human colon adenocarcinoma sections detected by NOX1 antisense in situ hybridization (reproduced with permission [28]). Autoradiographic exposure of silver grains, detected by white light scattering under darkfield microscopy, reveals

abundant NOX1 transcript levels in polarized epithelial layers of a differentiated tumor (b, arrowheads), whereas lower NOX1 transcript levels are detected in a more advanced tumor (d). (a, c) Corresponding H & E staining of the same sections presented in b and d, respectively

NOX1 overexpression is limited to human colon, small intestine and, to a lesser extent, gastric epithelial tumors at various stages of progression, including adenocarcinomas, adenomas, and polyps. Early-stage tumors showed a higher percentage of epithelial cells staining positive, with an accumulation of NOX1 protein on the luminal plasma membrane. Particularly noteworthy, high NOX1 protein levels were not a prominent immunohistological feature of prostate, lung, breast, or ovarian adenocarcinomas. Earlier studies by Geiszt et al. [28] reached similar conclusions based on in situ hybridization experiments probing NOX1 transcript levels in 1200 histological samples in NCI TARP1 and TARP2 human tumor tissue arrays: high NOX1 expression was limited to colon tumors and was detected at highest levels within more differentiated (less advanced) tumors, where epithelial cells were organized into polarized monolayers (Fig. 10.4). Here, comparatively low NOX1 transcript levels were

detected in human prostate, lung, breast, and ovarian tumors [28]. Fukuyama et al. [51], using an anti-NOX1 polyclonal antibody also detected highest NOX1 protein levels in differentiated colorectal tumors (adenomas and adenocarcinomas), where it was localized to the apical pole in perinuclear regions in non-proliferating (Ki67-negative) cells, whereas Ki67-positive cells produced lower levels of NOX1 protein. Thus, both studies considered NOX1 a marker of polarized, differentiated epithelial cells of the colon and disassociated high NOX1 expression from proliferation. Studies by Szanto et al. [31] detected NOX1 transcript in adenomas as well as adenocarcinomas and found no correlation of NOX1 levels with differentiation state. Laurent et al. [52] detected elevated NOX1 transcript and protein levels at stages I–III of colon cancer development but did not describe any expression bias correlated with differentiation state; higher NOX1 expression was detected in tumors

158

with activating RAS mutations in their analysis of a limited number of colon tumors, whereas no correlations were detected between TP53 mutation status and NOX1 expression. These results stand in contrast to a recent TCGA pan-cancer survey of >20,000 primary human tumors (encompassing 33 cancer types), in which NOX4 transcript levels were positively correlated with mutant TP53 transcript levels, along with several programs associated with cancer progression, including enhanced cell migration and invasiveness (downstream consequences of the epithelial-to-mesenchymal transition (EMT)), as well as angiogenesis, proliferation, and anti-apoptotic gene expression profiles [114]. Thus, patients with tumors with high NOX4 expression and wild type TP53 (which encodes a tumor suppressor protein (p53)) survive longer than those exhibiting high NOX4 expression within mutant TP53 tumors. In the latter case, mutant p53 has distinct effects on histone H3 acetylation within the NOX4 promoter, resulting in higher NOX4 expression, which was correlated with these cancer progression programs [115]. The recent informatics-based analysis of a colorectal adenocarcinoma cohort in TCGA also explored possible correlations between NOX1 expression and patient survival [54]. Here, Kaplan-Meier curves defined two simple groups of patients with tumors with either ‘high’ or ‘low’ levels relative to a median NOX1 expression cutoff value and compared the two groups for overall survival or diseasefree survival (n = 373 and 329, respectively). The authors observed no significant difference in overall survival or disease-free survival between ‘high’ and ‘low’ NOX1 expression groups, concluded that higher NOX1 expression does not contribute to the course of advanced disease, and speculated that NOX1 may have roles in earlier stages of carcinogenesis leading to CRC, as was suggested earlier [116]. We interrogated an updated cohort of colorectal adenocarcinoma patients in TCGA (597 subjects with survival information) using the open access resource of the Human Pathology Atlas of the Human Protein Atlas (HPA; www. proteinatlas.org/pathology) which allows the user to explore a range of ‘high’ and ‘low’ tumor gene expression groupings to establish a cutoff with maximum statistical significance [117]. This analysis (Fig. 10.5) found a significant difference in overall survival between the ‘high’ NOX1 expression group (n = 133 subjects; upper 22.3 percentile) and ‘low’ expression group (n = 464 subjects; lower 77.7 percentile). These results imply that high expression of NOX1 represents a favorable prognostic indicator for overall survival, a classification consistent with suggestions that NOX1 serves roles associated with colon epithelial cell differentiation. A link between sustained inflammation in colitis patients and development of CRC has long been appreciated [118, 119]. Proinflammatory induction of several colon epithelial NOX family ROS generators including NOX1 is

T. L. Leto and M. Geiszt

Fig. 10.5 Kaplan-Meier analysis of overall survival of patients in the colorectal adenocarcinoma cohort of TCGA performed using the open access resource of the Human Pathology Atlas of the Human Protein Atlas (www.proteinatlas.org/pathology) [117]. Patient groups designated with high and low tumor NOX1 transcript expression were assigned based on the determination of the cutoff providing maximum statistically significant differences between groups (P = 0.00097). The complete cohort dataset and resource that generated this graphic are available online: https://www.proteinatlas.org/ENSG00000007952NOX1/pathology/colorectal

worthwhile considering in this context as they represent potential mediators of oxidative stress that can result in excess DNA damage and cell transformation in these patients. Recent work with transgenic mice in which constitutively activated TLR4 was targeted to colon epithelial cells (under the control of the villin promoter) has provided interesting insights on potential involvement of NOX1 and other co-regulated ROS generators in colitis-associated cancer development [120]. NOX1, DUOX2, and NOS2 were among the most differentially overexpressed genes detected in villin-TLR4 colons. These observations are consistent with changes in gene expression profile datasets observed in patients with Crohn’s disease and CRC, which showed enhanced expression of TLR4 as well as NOX1 and DUOX2. The effects of TLR4 activation in villin-TLR4 mice included gut microbiome alterations that were associated with enhanced DUOX2-derived colon epithelial hydrogen peroxide production and increased numbers and sizes of colon tumors that developed in the azoxymethane (mutagenic)/DSS (colitogenic) treatment model of tumorigenesis. Furthermore, transfer of the altered mucosal microbiome from these villin-TLR4 mice was sufficient to

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

induce increased ROS production and susceptibility to tumorigenesis in the colons of recipient wild-type mice. Thus, it appears that both the microbiome alterations and excess colon epithelial ROS have roles in colonic tumorigenesis in mice even in the absence of active inflammation. These findings require further research including corroborative observations in IBD patients to define precise roles of NOX1 and other induced ROS generators in IBD-associated cancer before considering potential therapeutic interventions targeting these ROS generators or the microbiome.

7

Summary and Conclusions

In summary, NOX1 is one of several ROS generators in the gut that is intimately involved in a complex innate immune crosstalk between the host and gut microbiome. NOX1 is the prevalent epithelial NADPH oxidase detected in healthy uninflamed colon that is induced at higher levels along with other ROS generators in response to a variety of inflammatory signals, including cytokines and exposure to microbial patterns. ROS generated by NOX1 clearly serve signaling roles affecting colon epithelial cell proliferation and migration involved in mucosal epithelial cell homeostasis and barrier functions, and can also orchestrate other host defense responses, including induction of epithelial DUOX2 and NOS2 and other innate antimicrobial factors. While it is unclear whether the ROS generated by NOX1 serve a direct microbicidal role, as observed with NOX2-derived ROS in phagocytes, its role in controlling DUOX2 and NOS2 production may provide a means for significant microbicidal ROS generation that could alter or reshape the composition of the gut microbiome. Genetic defects in NOX1, DUOX2, and NOX2 (in phagocytes) have been linked to inflammatory bowel disease, which are all manifested with microbiome alterations associated with enhanced inflammatory responses. Thus, it appears that all three NOX enzymes together serve interrelated innate immune functions in maintaining a balanced gut microbiome required for intestinal health. Links between high NOX1 expression and cancer were proposed based initially on a functional association of NOX1 with cell proliferation. Surveys of expression profiles of primary human tumor databases and tumor tissue arrays indicate that high NOX1 expression is limited to tumors of the gastrointestinal tract. Higher NOX1 expression was not associated with altered disease-free or overall survival times in patients with colorectal adenocarcinomas, providing little support for the notion that NOX1 induction is related to latestage disease progression. However, given the association between cancer and chronic inflammation, the induction of NOX1 by pro-inflammatory agonists and microbial patterns,

159

and the potential for excess ROS mediating DNA damage, roles for NOX1 and other co-regulated ROS generators in early-stage carcinogenesis remain plausible. Thus, under conditions of unresolved chronic infection or dysbiosis, overstimulation by microbial factors would induce excess NOX-derived ROS generation, which could, in part, explain the observed enhanced susceptibility of ulcerative colitis and Crohn’s disease patients for developing colorectal cancers. Much remains to learned to account for the enhanced NOX1 expression observed in CRC and the possible involvement of NOX1 and other co-expressed ROS generators potentially linking colitis to enhanced risk for carcinogenesis. Acknowledgments This work was supported, in part, by the Intramural Research Program of the National Institute of Allergy and Infectious Diseases, NIH (T.L.L.). The research of M.G. is supported by a grant from the National Research, Development, and Innovation Office (K133002) and by grant VEKOP-2.3.2-16-2016-00002. His work is also supported by the Ministry of Innovation and Technology of Hungary from the National Research, Development and Innovation Fund, financed under the MOLORKIV funding scheme.

References 1. Suh YA, Arnold RS, Lassegue B, Shi J, Xu X, Sorescu D, Chung AB, Griendling KK, Lambeth JD (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401:79–82 2. Lambeth JD (2004) NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol 4:181–189 3. Geiszt M, Leto TL (2004) NOX family of NAD(P)H oxidases: host defense and beyond. J Biol Chem 279:51715–51718 4. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313 5. Yu SM, Ferrans Z-X, Irani VJ, Finkel K (1995) Requirement for generation of H2O2 for platelet-derived growth factor signal transduction. Science 270:296–299 6. Irani K, Xia Y, Zweier JL, Sollott SJ, Der CJ, Fearon ER, Sundaresan M, Finkel T, Goldschmidt-Clermont PJ (1997) Mitogenic signaling by oxidants in Ras-transformed fibroblasts. Science 275:1649–1652 7. Bae YS, Kang SW, Seo MS, Baines IC, Tekle E, Chock PB, Rhee SG (1997) Epidermal growth factor (EGF)-induced generation of hydrogen peroxide. Role in EGF receptor-mediated tyrosine phosphorylation. J Biol Chem 272:217–221 8. Magnani F, Nenci S, Millana Fananas E, Ceccon M, Romero E, Fraaije MW, Mattevi A (2017) Crystal structures and atomic model of NADPH oxidases. Proc Natl Acad Sci U S A 114:6764–6769 9. Banfi B, Maturana A, Jaconi S, Arnaudeau S, Laforge T, Sinha B, Ligeti E, Demaurex N, Krause KH (2000) A mammalian H + channel generated through alternative splicing of the NADPH oxidase homolog NOH-1. Science 287:138–142 10. Geiszt M, Lekstrom K, Leto TL (2004) Analysis of mRNA transcripts from the NAD(P)H oxidase 1 (Nox1) gene. Evidence against production of the NADPH oxidase homolog-1 short (NOH-1S) transcript variant. J Biol Chem 279:51 661–668 11. Harper RW, Xu CH, Soucek K, Setiadi H, Eiserich JP (2005) A reappraisal of the genomic organization of human Nox1 and its splice variants. Arch Biochem Biophys 435:323–330

160 12. Banfi B et al (2005) Corrections and clarifications. Science 307:44. https://doi.org/10.1126/science.307.5706.44 13. Ramsey IS, Moran MM, Chong JA, Chapman DE (2006) A voltage-gated proton-selective channel lacking the pore domain. Nature 440:1213–1216 14. Arnold RS, Shi J, Murad E, Whalen AM, Sun CQ, Polavarapu R, Parthasarathy S, Petros JA, Lambeth JD (2001) Hydrogen peroxide mediates the cell growth and transformation caused by the mitogenic oxidase nox1. Proc Natl Acad Sci U S A 98:5550–5555 15. Arbiser JL, Petros J, Klafter R, Govindajaran B, McLaughlin ER, Brown LF, Cohen C, Moses M, Kilroy S, Arnold RS, Lambeth JD (2002) Reactive oxygen generated by Nox1 triggers the angiogenic switch. Proc Natl Acad Sci U S A 99:715–720 16. Banfi B, Clark RA, Steger K, Krause KH (2003) Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 278:3510–3513 17. Geiszt M, Lekstrom K, Witta J, Leto TL (2003) Proteins homologous to p47phox and p67phox support superoxide production by NAD(P)H oxidase 1 in colon epithelial cells. J Biol Chem 278: 20006–20012 18. Takeya R, Ueno N, Kami K, Taura M, Kohjima M, Izaki T, Nunoi H, Sumimoto H (2003) Novel human homologues of p47phox and p67phox participate in activation of superoxideproducing NADPH oxidases. J Biol Chem 278:25234–25246 19. Fagerberg L, Hallstrom B, Oksvold P et al (2014) Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics. Mol Cell Proteomics 13:397–406. https://doi.org/10.1074/mcp.M113.035600 20. Leto TL, Morand S, Hurt D, Ueyama T (2009) Targeting and regulation of reactive oxygen species generation by Nox family NADPH oxidases. Antioxid Redox Signal 11:2607–2619 21. Sumimoto H, Minakami R, Miyano K (2019) Soluble regulatory proteins for activation of NOX Family NADPH oxidases. In: Knaus UG, Leto TL (eds) NADPH oxidases: methods and protocols. Methods in molecular biology, vol 1982. Springer, New York, pp 121–128 22. Ambasta RK, Kumar P, Griendling KK, Schmidt H, Busse R, Brandes RP (2004) Direct interaction of the novel nox proteins with p22phox is required for the formation of a functionally active NADPH oxidase. J Biol Chem 279:45935–45941 23. Ueyama T, Geiszt M, Leto TL (2006) Involvement of Rac1 in activation of multicomponent Nox1- and Nox3-based NADPH oxidases. Mol Cell Biol 26:2160–2174 24. Dutta S, Rittinger K (2010) Regulation of NOXO1 activity through reversible interactions with p22phox and NOXA1. PLoS One 5: e10478 25. Miyano K, Ueno N, Takeya R, Sumimoto H (2006) Direct involvement of the small GTPase Rac in activation of the superoxide producing NADPH oxidase Nox1. J Biol Chem 281:21857–21868 26. Cheng G, Diebold BA, Hughes Y, Lambeth JD (2006) Nox1dependent reactive oxygen generation is regulated by Rac1. J Biol Chem 281:17718–17726 27. Abo A, Pick E, Hall A, Totty N, Teahan CG, Segal AW (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21 rac1. Nature 353:668–670 28. Geiszt M, Lekstrom K, Brenner S, Hewitt SM, Dana R, Malech HL, Leto TL (2003) NAD(P)H oxidase 1, a product of differentiated colon epithelial cells, can partially replace glycoprotein 91phox in the regulated production of superoxide by phagocytes. J Immunol 171:299–306 29. Lavigne M, Holland SM, Leto TL (2001) Genetic demonstration of p47phox-dependent superoxide anion production in murine vascular smooth muscle cells. Circulation 104:79–84 30. Li JM, Wheatcroft S, Fan LM, Kearney MT, Shah AM (2004) Opposing roles of p47phox in basal versus angiotensin II–stimulated

T. L. Leto and M. Geiszt alterations in vascular O2- production, vascular tone, and mitogenactivated protein kinase activation. Circulation 109:1307–1313 31. Szanto I, Rubbia-Brandt L, Kiss P et al (2005) Expression of NOX1, a superoxide generating NADPH oxidase, in colon cancer and inflammatory bowel disease. J Pathol 207:164–176 32. Borregaard N, Heiple JM, Simons ER et al (1983) Subcellular localization of the b-cytochrome component of the human neutrophil microbicidal oxidase: translocation during activation. J Cell Biol 97:52–61 33. Miller FJ Jr, Filali M, Huss GJ, Stanic B, Chamseddine A, Barna TJ, Lamb FS (2007) Cytokine activation of nuclear factor kappa B in vascular smooth muscle cells requires signaling endosomes containing Nox1 and CIC-3. Circ Res 101:663–671 34. Yamamoto A, Takeya R, Matsumoto M, Nakayama KI, Sumimoto H (2013) Phosphorylation of Noxo1 at threonine 341 regulates its interaction with Noxa1 and the superoxide producing activity of Nox1. FEBS J 280:5145–5159 35. Debbabi M, Kroviarski Y, Bournier O, Gougerot-Pocidalo MA, El-Benna J, Dang PM (2013) NOXO1 phosphorylation on serine 154 is critical for optimal NADPH oxidase 1 assembly and activation. FASEB J 27:1733–1748 36. Streeter J, Schickling BM, Jiang S, Stanic B, Thiel WH, Gakhar L, Houtman JC, Miller FJ Jr (2014) Phosphorylation of Nox1 regulates association with NoxA1 activation domain. Circ Res 115:911–918 37. Kim JS, Diebold BA, Babior BM, Knaus UG, Bokoch GM (2007) Regulation of Nox1 activity via protein kinase A-mediated phosphorylation of NoxA1 and 14-3-3 binding. J Biol Chem 282: 34787–34800 38. Kroviarski Y, Debbabi M, Bachoual R, Perianin A, GougerotPocidalo MA, El-Benna J, Dang PM (2010) Phosphorylation of NADPH oxidase activator 1 (NOXA1) on serine 282 by MAP kinases and on serine 172 by protein kinase C and protein kinase A prevents NOX1 hyperactivation. FASEB J 24:2077–2092 39. Karathanassis D, Stahelin RV, Bravo J, Perisic O, Pacold CM, Cho WW, Williams RL (2002) Binding of the PX domain of p47 phox to phosphatidylinositol 3,4-bisphosphate and phosphatidic acid is masked by an intramolecular interaction. EMBO J 21(19): 5057–5068 40. Ueyama T, Lekstrom K, Tsujibe S, Saito N, Leto TL (2007) Subcellular localization and function of alternatively spliced Noxo1 isoforms. Free Radic Biol Med 42:180–190 41. Cheng G, Lambeth JD (2004) NOXO1, regulation of lipid binding, localization, and activation of Nox1 by the Phox homology (PX) domain. J Biol Chem 279:4737–4742 42. Takeya R, Taura M, Yamasaki T, Naito S, Sumimoto H (2006) Expression and function of Noxo1γ, an alternative splicing form of the NADPH oxidase organizer 1. FEBS J 273:3663–3677 43. Tian W, Li XJ, Stull ND, Ming W, Suh CI, Bissonnette SA, Yaffe MB, Grinstein S, Atkinson SJ, Dinauer MC (2008) FcgRstimulated activation of the NADPH oxidase: phosphoinositidebinding protein p40phox regulates NADPH oxidase activity after enzyme assembly on the phagosome. Blood 112:3867–3877 44. Ueyama T, Kusakabe T, Karasawa S, Kawasaki T, Shimizu A, Son J, Leto TL, Miyawaki A, Saito NJ (2008) Sequential binding of cytosolic Phox complex to phagosomes through regulated adaptor proteins: evaluation using novel monomeric Kusabiragreen system and live imaging of phagocytosis. J Immunol 181: 629–640 45. Fisher AB (2017) Peroxiredoxin 6 in the repair of peroxidized cell membranes and cell signaling. Arch Biochem Biophys 617:68–83 46. Chatterjee S, Feinstein SI, Dodia C, Sorokina E, Lien YC, Nguyen S, Debolt K, Speicher D, Fisher AB (2011) Peroxiredoxin 6 phosphorylation and subsequent phospholipase A2 activity are required for agonist-mediated activation of NADPH oxidase in mouse

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

pulmonary microvasculature endothelium and alveolar macrophages. J Biol Chem 286:11696–11706 47. Ambruso DR, Ellison MA, Thurman GW, Leto TL (2012) Peroxiredoxin 6 translocates to the plasma membrane during neutrophil activation and is required for optimal NADPH oxidase activity. Biochim Biophys Acta 1823:306–315 48. Vázquez-Medina JP, Dodia C, Weng L et al (2016) The phospholipase A2 activity of peroxiredoxin 6 modulates NADPH oxidase 2 activation via lysophosphatidic acid receptor signaling in the pulmonary endothelium and alveolar macrophages. FASEB J 30: 2885–3898 49. Kwon J, Wang A, Burke DJ, Boudreau HE, Lekstrom KJ, Korzeniowska A, Sugamata R, Kim YS, Yi L, Ersoy I, Jaeger S, Palaniappan K, Ambruso DR, Jackson SH, Leto TL (2016) Peroxiredoxin 6 (Prdx6) supports NADPH oxidase1 (Nox1)based superoxide generation and cell migration. Free Radic Biol Med 96:99–115 50. van der Post S, Birchenough GMH, Held JM (2021) NOX1dependent redox signaling potentiates colonic stem cell proliferation to adapt to the intestinal microbiota by linking EGFR and TLR activation. Cell Rep 35:108949 51. Fukuyama M, Rokutan K, Sano T, Miyake H, Shimada M, Tashiro S (2005) Overexpression of a novel superoxide producing enzyme, NADPH oxidase 1, in adenoma and well differentiated adenocarcinoma of the human colon. Cancer Lett 221:97–104 52. Laurent E, McCoy JW 3rd, Macina RA, Liu W, Cheng G, Robine S, Papkoff J, Lambeth JD (2008) Nox1 is over-expressed in human colon cancers and correlates with activating mutations in K-Ras. Int J Cancer 123:100–107 53. Matsumoto M, Katsuyama M, Iwata K, Ibi M, Zhang J, Zhu K, Nauseef WM, Yabe-Nishimura C (2014) Characterization of N-glycosylation sites on the extracellular domain of NOX1/ NADPH oxidase. Free Radic Biol Med 68:196–204 54. Lu J, Jiang G, Wu Y, Antony S, Meitzler JL, Juhasz A, Liu H, Roy K, Makhlouf H, Chuaqui R, Butcher D, Konaté MM, Doroshow JH (2020) NADPH oxidase 1 is highly expressed in human large and small bowel cancers. PLoS One 15:e0233208 55. Kuwano Y, Kawahara T, Yamamoto H, Teshima-Kondo S, Tomiga K, Masuda K, Kishi K, Morita K, Rokutan K (2006) Interferon-gamma activated transcription of NADPH oxidase 1 gene and upregulates production of superoxide anion by human large intestinal epithelial cells. Am J Phys Cell Physiol 290:c433–c443 56. Kuwano Y, Tominaga K, Kawahara T, Sasaki H, Takeo K, Nishida K, Masuda K, Kawai T, Teshima-Kondo S, Rokutan K (2008) Tumor necrosis factor alpha activates transcription of NADPH oxidase organizer 1 (NOXO) gene and upregulates superoxide production in colon epithelial cells. Free Radic Biol Med 45: 1642–1652 57. Kamizato M, Nishida K, Masuda K, Takeo K, Yamamoto Y, Kawai T, Teshima-Kondo S, Tanahashi T, Rokutan K (2009) Interleukin 10 inhibits interferon gamma- and tumor necrosis factor alpha-stimulated activation of NADPH oxidase 1 in human colonic epithelial cells and the mouse colon. J Gastroenterol 44:1172–1184 58. Makhezer N, Ben Khemis M, Liu D, Khichane Y, Marzaioli V, Tlili A, Mojallali M, Pintard C, Letteron P, Hurtado-Nedelec M, El-Benna J, Marie JC, Sannier A, Pelletier AL, Dang PM (2019) NOX1-derived ROS drive the expression of Lipocalin-2 in colonic epithelial cells in inflammatory conditions. Mucosal Immunol 12: 117–131 59. Liu H, Antony S, Roy K, Juhasz A, Wu Y, Lu J, Meitzler JL, Jiang G, Polley E, Doroshow JH (2017) Interleukin-4 and interleukin-13 increase NADPH oxidase 1-related proliferation of human colon cancer cells. Oncotarget 8:38113–38135

161

60. Kawahara T, Teshima S, Oka A, Sugiyama T, Kishi K, Rokutan K (2001) Type I helicobacter pylori lipopolysaccharide stimulates toll-like receptor 4 and activates mitogen oxidase 1 in gastric pit cells. Infect Immun 69:4382–4389 61. Kawahara T, Kuwano Y, Teshima-Kondo S et al (2004) Role of nicotinamide adenine dinucleotide phosphate oxidase 1 in oxidative burst response to toll-like receptor 5 signaling in large intestinal epithelial cells. J Immunol 172:3051–3058 62. Aviello G, Knaus UG (2018) NADPH oxidases and ROS signaling in the gastrointestinal tract. Mucosal Immunol 11:1011–1023 63. Dang PM, Rolas L, El-Benna J (2020) The dual role of reactive oxygen species-generating nicotinamide adenine dinucleotide phosphate oxidases in gastrointestinal inflammation and therapeutic perspectives Antioxid & Redox. Signals 33: 354–373 64. Esworthy RS, Kim BW, Chow J, Shen B, Doroshow JH, Chu FF (2014) Nox1 causes ileocolitis in mice deficient in glutathione peroxidase-1 and -2. Free Radic Biol Med 68:315–325 65. Coant N, Ben Mkaddem S, Pedruzzi E et al (2010) NADPH oxidase 1 modulates WNT and NOTCH1 signaling to control the fate of proliferative progenitor cells in the colon. Mol Cell Biol 30: 2636–2650 66. Tréton X, Pedruzzi E, Guichard C, Ladeiro Y, Sedghi S, Vallée M, Fernandez N, Bruyère E, Woerther PL, Ducroc R et al (2014) Combined NADPH oxidase 1 and interleukin 10 deficiency induces chronic endoplasmic reticulum stress and causes ulcerative colitis-like disease in mice. PLoS One 9:e101669 67. Jones RM, Luo L, Ardita CS, Richardson AN, Kwon YM, Mercante JW et al (2013) Symbiotic lactobacilli stimulate gut epithelial proliferation via Nox-mediated generation of reactive oxygen species. EMBO J 32:3017–3028 68. Alam A, Leoni G, Wentworth CC, Kwal JM, Wu H, Ardita CS, Swanson PA, Lambeth JD, Jones RM, Nusrat A, Neish AS (2014) Redox signaling regulates commensal-mediated mucosal homeostasis and restitution and requires formyl peptide receptor 1. Mucosal Immunol 7:645–655 69. Leoni G, Alam A, Neumann PA, Lambeth JD, Cheng G, McCoy J, Hilgarth RS, Kundu K, Murthy N, Kusters D, Reutelingsperger C, Perretti M, Parkos CA et al (2013) Annexin A1, formyl peptide receptor, and NOX1 orchestrate epithelial repair. J Clin Invest 123: 443–454 70. Kato M, Marumo M, Nakayama J, Matsumoto M, YabeNishimura C, Kamata T (2016) The ROS-generating oxidase Nox1 is required for epithelial restitution following colitis. Exp Anim 65:197–205 71. Liu J, Iwata K, Zhu K, Matsumoto M, Matsumoto K, Asaoka N, Zhang X, Ibi M, Katsuyama M, Tsutsui M, Kato S, YabeNishimura C (2020) NOX1/NADPH oxidase in bone marrowderived cells modulates intestinal barrier function. Free Radic Biol Med 147:90–101 72. Stenke E, Bourke B, Knaus UG (2019) NADPH oxidases in inflammatory bowel disease. In: Knaus UG, Leto TL (eds) NADPH oxidases: methods and protocols. Methods in molecular biology, vol 1982. Springer, New York, pp 695–713 73. Falcone EL, Holland SM (2019) Gastrointestinal complications in chronic granulomatous disease. In: Knaus UG, Leto TL (eds) NADPH oxidases: methods and protocols. Methods in molecular biology, vol 1982. Springer, New York, pp 573–586 74. Schwerd T, Bryant RV, Pandey S, Capitani M, Meran L, Cazier JB et al (2018) NOX1 loss-of-function genetic variants in patients with inflammatory bowel disease. Mucosal Immunol 11:562–574 75. Hayes P, Dhillon S, O’Neill K, Thoeni C, Hui KY, Elkadri A et al (2015) Defects in NADPH oxidase genes NOX1 and DUOX2 in very early onset inflammatory bowel disease. Cell Mol Gastroenterol Hepatol 1:489–502

162 76. Lipinski S, Till A, Sina C et al (2009) DUOX2-derived reactive oxygen species are effectors of NOD2-mediated antibacterial responses. J Cell Sci 122:3522–3530 77. Khoshnevisan R, Anderson M, Babcock S, Anderson S, Illig D, Marquardt B, Sherkat R, Schröder K, Moll F, Hollizeck S, Rohlfs M, Walz C, Adibi P, Rezaei A, Andalib A, Koletzko S, Muise AM, Snapper SB, Klein C, Thiagarajah JR, Kotlarz D (2020) NOX1 regulates collective and planktonic cell migration: insights from patients with pediatric-onset IBD and NOX1 deficiency. Inflamm Bowel Dis 26:1166–1176 78. Kuhns DB, Alvord WG, Heller T, Feld JJ, Pike KM, Marciano BE et al (2010) Residual NADPH oxidase and survival in chronic granulomatous disease. N Engl J Med 363:2600–2610 79. Huang C, De Ravin SS, Paul AR, Heller T, Ho N, Wu Datta L, Zerbe CS, Marciano BE, Kuhns DB, Kader HA, Holland SM, Malech HL, Brant SR, NIDDK IBD Genetics Consortium (2016) Genetic risk for inflammatory bowel disease is a determinant of Crohn’s disease development in chronic granulomatous disease. Inflamm Bowel Dis 22:2794–2801 80. LaBere B, Gutierrez MJ, Wright H, Garabedian E, Ochs HD, Fuleihan RL, Secord E, Marsh R, Sullivan KE, CunninghamRundles C, Notarangelo LD, Chen K (2022) Chronic granulomatous disease with inflammatory bowel disease: clinical presentation, treatment, and outcomes from the USIDNET Registry. J Allergy Clin Immunol Pract 10:1325–1333 81. Dhillon SS, Fattouh R, Elkadri A et al (2014) Variants in nicotinamide adenine dinucleotide phosphate oxidase complex components determine susceptibility to very early onset inflammatory bowel disease. Gastroenterology 147:680–689 82. Marsh RA, Leiding JW, Logan BR, Griffith LM, Arnold DE, Haddad E, Falcone EL, Yin Z et al (2019) Chronic granulomatous disease-associated IBD resolves and does not adversely impact survival following allogeneic HCT. J Clin Immunol 39:653–667 83. Campbell EL, Bruyninckx WJ, Kelly CJ, Glover LE, McNamee EN, Bowers BE et al (2014) Transmigrating neutrophils shape the mucosal microenvironment through localized oxygen depletion to influence resolution of inflammation. Immunity 40:66–77 84. Dupuy C, Ohayon R, Valent A et al (1999) Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. Cloning of the porcine and human cDNAs. J Biol Chem 274:37265–37269 85. De Deken X, Wang D, Many MC et al (2000) Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275:23227–23233 86. DeDeken X, Miot F (2019) Duox defects and their roles in congenital hypothyroidism. In: Knaus UG, Leto TL (eds) NADPH oxidases: methods and protocols. Methods in molecular biology, vol 1982. Springer, New York, pp 667–693 87. Geiszt M, Witta J, Baffi J, Lekstrom K, Leto TL (2003) Dual oxidases represent novel hydrogen peroxide sources supporting mucosal surface host defense. FASEB J 17:1502–1504 88. Ha EM, Oh CT, Bae YS et al (2005) A direct role for dual oxidase in drosophila gut immunity. Science 310:847–850 89. Chavez V, Mohri-Shiomi A, Garsin DA (2009) Ce-Duox1/BLI-3 generates reactive oxygen species as a protective innate immune mechanism in Caenorhabditis elegans. Infect Immun 77:4983– 4989 90. Flores MV, Crawford KC, Pullin LM et al (2010) Dual oxidase in the intestinal epithelium of zebrafish larvae has anti-bacterial properties. Biochem Biophys Res Commun 400:164–168 91. Parlato M, Charbit-Henrion F, Hayes P, Tiberti A, Aloi M, Cucchiara S et al (2017) First identification of biallelic inherited DUOX2 inactivating mutations as a cause of very early onset inflammatory bowel disease. Gastroenterology 153:609–611.e3 92. Haberman Y, Tickle TL, Dexheimer PJ, Kim MO, Tang D, Karns R, Baldassano RN, Noe JD, Rosh J, Markowitz J, Heyman

T. L. Leto and M. Geiszt MB, Griffiths AM, Crandall WV, Mack DR, Baker SS, Huttenhower C, Keljo DJ, Hyams JS, Kugathasan S, Walters TD, Aronow B, Xavier RJ, Gevers D, Denson LA (2014) Pediatric Crohn disease patients exhibit specific ileal transcriptome and microbiome signature. J Clin Invest 124:3617–3633 93. Grasberger H, Gao J, Nagao-Kitamoto H, Kitamoto S, Zhang M, Kamada N et al (2015) Increased expression of DUOX2 is an epithelial response to mucosal dysbiosis required for immune homeostasis in mouse intestine. Gastroenterology 149:1849–1859 94. Grasberger H, Magis AT, Sheng E, Conomos MP, Zhang M, Garzotto LS, Hou G, Bishu S, Nagao-Kitamoto H, El-Zaatari M, Kitamoto S, Kamada N, Stidham RW, Akiba Y, Kaunitz J, Haberman Y, Kugathasan S, Denson LA, Omenn GS, Kao JY (2021) DUOX2 variants associate with preclinical disturbances in microbiota-immune homeostasis and increased inflammatory bowel disease risk. J Clin Invest 131(9):e141676 95. Corcionivoschi N, Alvarez LA, Sharp TH, Strengert M, Alemka A, Mantell J, Verkade P, Knaus UG, Bourke B (2012) Mucosal reactive oxygen species decrease virulence by disrupting Campylobacter jejuni phosphotyrosine signaling. Cell Host Microbe 12:47–59 96. Grasberger H, El-Zaatari M, Dang DT et al (2013) Dual oxidases control release of hydrogen peroxide by the gastric epithelium to prevent Helicobacter felis infection and inflammation in mice. Gastroenterology 145:1045–1054 97. Sommer F, Backhed F (2015) The gut microbiota engages different signaling pathways to induce Duox2 expression in the ileum and colon epithelium. Mucosal Immunol 8:372–379 98. MacFie TS, Poulsom R, Parker A, Warnes G, Boitsova T, Nijhuis A et al (2014) DUOX2 and DUOXA2 form the predominant enzyme system capable of producing the reactive oxygen species H2O2 in active ulcerative colitis and are modulated by 5-aminosalicylic acid. Inflamm Bowel Dis 20:514–524 99. Matziouridou SD, Roch C, Haabeth OA, Rudi K, Carlsen H, Kielland A (2018) iNOS- and NOX1-dependent ROS production maintains bacterial homeostasis in the ileum of mice. Mucosal Immunol 11:774–784 100. Pircalabioru G, Aviello G, Kubica M, Zhdanov A, Paclet MH, Brennan L, Hertzberger R, Papkovsky D, Bourke B, Knaus UG (2016) Defensive mutualism rescues NADPH Oxidase inactivation in gut infection. Cell Host Microbe 19:651–663 101. Falcone EL, Abusleme L, Swamydas M, Lionakis MS, Ding L, Hsu AP et al (2016) Colitis susceptibility in p47(phox-/-) mice is mediated by the microbiome. Microbiome 4:13 102. Rodrigues-Sousa T, Ladeirinha AF, Santiago AR, Carvalheiro H, Raposo B, Alarcao A et al (2014) Deficient production of reactive oxygen species leads to severe chronic DSS-induced colitis in Ncf1/p47phox-mutant mice. PLoS One 9(5):e97532 103. Miller BM, Liou MJ, Zhang LF, Nguyen H, Litvak Y, Schorr EM, Jang KK, Tiffany CR, Butler BP, Bäumler AJ (2020) Anaerobic respiration of NOX1-Derived hydrogen Peroxide licenses bacterial growth at the colonic surface. Cell Host Microbe 28:789–797 104. Aviello G, Singh AK, O'Neill S, Conroy E, Gallagher W, D'Agostino G, Walker AW, Bourke B, Scholz D, Knaus UG (2019) Colitis susceptibility in mice with reactive oxygen species deficiency is mediated by mucus barrier and immune defense defects. Mucosal Immunol 12:1316–1326 105. Matsunaga Y, Clark T, Wanek AG, Bitoun JP, Gong Q, Good M, Kolls JK (2021) Intestinal IL-17R Signaling controls secretory IgA and oxidase balance in Citrobacter rodentium infection. J Immunol 206:766–775 106. Kumar P, Monin L, Castillo P, Elsegeiny W, Horne W, Eddens T, Vikram A, Good M, Schoenborn AA, Bibby K et al (2016) Intestinal interleukin-17 receptor signaling mediates reciprocal control of the gut microbiota and auto-immune inflammation. Immunity 44: 659–671

10

NADPH Oxidase 1: At the Interface of the Intestinal Epithelium and Gut Microbiota

107. Gimenez M, Schickling BM, Lopes LR, Miller FJ Jr (2016) Nox1 in cardiovascular diseases: regulation and pathophysiology. Clin Sci (Lond) 130:151–165 108. Sirokmany G, Donko A, Geiszt M (2016) Nox/Duox family of NADPH oxidases: lessons from knockout mouse models. Trends Pharmacol Sci 37:318–327 109. Lamb FS, Choi H, Miller MR, Stark RJ (2020) TNFα and reactive oxygen signaling in vascular smooth muscle cells in hypertension and atherosclerosis. Am J Hypertens 33:902–913 110. Schröder K (2010) Isoform specific functions of Nox proteinderived reactive oxygen species in the vasculature. Curr Opin Pharmacol 10:122–126 111. Kamata T (2009) Roles of Nox1 and other Nox isoforms in cancer development. Cancer Sci 100:1382–1388 112. Sirokmany G, Pató A, Zana M, Donkó A, Bíró A, Nagy P, Geiszt M (2016) Epidermal growth factor-induced hydrogen peroxide production is mediated by dual oxidase 1. Free Radic Biol Med 97:204–211 113. Heppner D, Hristova M, Dustin CM, Danyal K, Habibovic A, van der Vliet A (2016) The NADPH oxidases DUOX1 and NOX2 play distinct roles in redox regulation of epidermal growth factor receptor signaling. J Biol Chem 291:23282–23293 114. Ma WF, Boudreau HE, Leto TL (2021) Pan-cancer analysis shows TP53 mutations modulate the association of NOX4 with genetic

163

programs of cancer progression and clinical outcome. Antioxidants (Basel) 10:235 115. Boudreau HE, Ma WF, Korzeniowska A, Park JJ, Bhagwat MA, Leto TL (2017) Histone modifications affect differential regulation of TGFβ- induced NADPH oxidase 4 (NOX4) by wild-type and mutant p53. Oncotarget 8:44379–44397 116. Rokutan K, Kawahara T, Kuwano Y, Tominaga K, Sekiyama A, Teshima-Kondo S (2006) NADPH oxidases in the gastrointestinal tract: a potential role of Nox1 in innate immune response and carcinogenesis. Antioxid Redox Signal 8:1573–1582 117. Uhlem M, Zhang C, Lee S et al (2017) A pathology atlas of the human cancer transcriptome. Science 357. https://doi.org/10.1126/ science.aa2507 118. Ekbom A, Helmick C, Zack M et al (1990) Ulcerative colitis and colorectal cancer. A population-based study. N Engl J Med 323: 1228–1233 119. Rubio CA, Befrits R (1997) Colorectal adenocarcinoma in Crohn’s disease: a retrospective histologic study. Dis Colon Rectum 40: 1072–1078 120. Burgueño JF, Fritsch J, Gonzalez EE et al (2020) Epithelial TLR4 signaling activates DUOX2 to induce microbiota-driven tumorigenesis. Gastroenterology 160:797–808

Physiological Functions and Pathological Significance of NADPH Oxidase 3

11

Yoko Nakano and Botond Bánfi

Abstract

NADPH oxidase 3 (NOX3) is the catalytic subunit of a superoxide-producing enzyme complex in the inner ear. Other subunits of this complex include p22phox and NOX organizer 1 (NOXO1). Both of these accessory proteins are essential for the enzymatic activity of NOX3, with p22phox stabilizing, and NOXO1 activating it. In mice, deleterious mutations in the genes that encode any of these subunits cause a balance defect and a complete lack of calcium carbonate crystals (otoconia) in the gravity-sensing organs (utricle and saccule) of the inner ear. Consistent with a role in the genesis of otoconia, the NOX3-containing enzyme complex is expressed adjacent to the utricle and saccule, in the endolymphatic sac and duct. NOX3 and p22phox are also expressed in the cochlea, and NOX3 is likely enzymatically active and pathogenic at this location because inactivating mutations protect mice from distinct types of acquired hearing loss. Thus, NOX3 plays both developmental and pathological roles in the inner ear. Finally, a few studies have linked NOX3 to physiological and pathological processes outside the inner ear. In this chapter we summarize key discoveries related Y. Nakano Department of Anatomy and Cell Biology, Carver College of Medicine, University of Iowa, Iowa City, IA, USA Inflammation Program, Carver College of Medicine, University of Iowa, Iowa City, IA, USA e-mail: [email protected] B. Bánfi (✉) Department of Anatomy and Cell Biology, Carver College of Medicine, University of Iowa, Iowa City, IA, USA Inflammation Program, Carver College of Medicine, University of Iowa, Iowa City, IA, USA Department of Otolaryngology – Head and Neck Surgery, Carver College of Medicine, University of Iowa, Iowa City, IA, USA Department of Internal Medicine, Carver College of Medicine, University of Iowa, Iowa City, IA, USA e-mail: botond-banfi@uiowa.edu

to the expression pattern, interacting partners, physiological functions, and pathological effects of NOX3. Keywords

NOX3 · p22phox · NOXO1 · Otoconia · Endolymphatic sac · Cisplatin-induced ototoxicity

1

Introduction

NOX enzymes produce superoxide, a reactive molecule and precursor of other reactive oxygen species (ROS), including hydrogen peroxide (H2O2), hypochlorous acid, and hypothiocyanite (Fig. 11.1a). Superoxide is generated by one-electron reduction of molecular oxygen. To catalyze this reaction, NOX enzymes transfer electrons from cytosolic nicotinamide adenine dinucleotide phosphate (NADPH) to exoplasmic oxygen (reviewed in [1]). The electron transfer is mediated by NOX-anchored flavin adenine dinucleotide (FAD) and two heme groups (Fig. 11.1b). NADPH and FAD are bound to the cytosolic C-terminal tail of NOX proteins, and the two heme groups are anchored to the N-terminal transmembrane region (Fig. 11.1b) [2]. Electrons flow along this route, from NADPH to FAD, from FAD to the FAD-proximal (inner) heme, from the inner to the outer heme, and finally from the outer heme to molecular oxygen [2]. Thus, NOX proteins harbor a complete molecular apparatus for electron transport across the membrane. The human genome encodes five NOX proteins (NOX1– NOX5) and two NOX-related ‘dual oxidases’ (DUOX1 and DUOX2) (reviewed in [3]). NOX1–3 and NOX5 generate and release superoxide. NOX4 and the DUOX proteins generate superoxide as an intermediate product and ultimately release H2O2 (reviewed in [4]). The first identified member of the NOX–DUOX family was NOX2 (formerly named gp91phox), the catalytic subunit of the phagocyte NOX complex [5]. The superoxide generated by NOX2 is vital for killing invading bacteria (reviewed in [6]). Although

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_11

165

166

Y. Nakano and B. Bánfi

A

B .-

2O2 + NADP+ + H+ transmembrane

NOX

outer heme (2)

.-

2O2 + 2H+

-

(3) H2O2 + Cl + H

(4)

(SOD)

+

MPO

-

LPO

H2O2 + SCN

H2O2 + O2 inner heme

HOCl + H2O N -

OSCN + H2O

FAD NADPH

C

cytosolic

(1) 2O2 + NADPH

Fig. 11.1 Chemical reactions catalyzed by selected redox enzymes and overall structure of NADPH oxidase (NOX) proteins. (a) Redox reactions catalyzed by NOX proteins, superoxide dismutase (SOD), myeloperoxidase (MPO), and lactoperoxidase (LPO). SOD is parenthesized because spontaneous dismutation of superoxide is quite rapid. On each line, the molecular formulas of the resulting ROS are indicated by red, i.e., superoxide in (1), hydrogen peroxide in (2),

hypochlorous acid in (3), and hypothiocyanite in (4). (b) Schematic of the overall structure of the NOX1–NOX4 proteins. NOX5 is similar except that its cytosolic N-terminal region is 20-fold longer. Transmembrane helices (cylinders) are arranged based on crystallographic data for NOX5 [2]. Positions of the NOX-bound hemes (orange), FAD (red), and NADPH (green) are indicated. ‘N’ and ‘C’ represent the N-terminal and C-terminal ends, respectively

superoxide is not particularly bactericidal, it is a precursor of more toxic ROS, including H2O2 and hypochlorous acid. H2O2 is a product of superoxide dismutation, which can occur either spontaneously or be catalyzed by superoxide dismutase (Fig. 11.1a), and hypochlorous acid is a product of chloride oxidation, a reaction catalyzed by the phagocytic enzyme myeloperoxidase (MPO) (Fig. 11.1a) (reviewed in [7]). Of the NOX2-related proteins, only DUOX1 has been proposed to play a NOX2-like role in bacterial killing [8– 11]. It produces H2O2 at the apical plasma membrane of airway epithelial cells. There, the MPO-like enzyme lactoperoxidase catalyzes the reaction of H2O2 with the secreted ion thiocyanate, producing the potent antibacterial molecule hypothiocyanite (Fig. 11.1a) [9, 10, 12, 13]. The roles of other NOX2-related proteins include supporting the formation of otoconia (NOX3), hormone synthesis (DUOX2), or signal transduction (NOX1, NOX4 and NOX5) (reviewed in [14]). Here we focus on the expression pattern, regulation, functions, and pathological significance of NOX3.

[15, 16]. In addition, the Tabula Muris, a compendium of single-cell transcriptome data from Mus musculus, indicates that Nox3 expression is minimal in all 81 cell types analyzed [17]. The inner ear was not among the organs that were used for cell isolation in the Tabula Muris project [17]. These data indicate that, under physiological conditions, Nox3 is expressed in the inner ear and not in most other organs.

2

Expression and Subunit Composition of a NOX3-Containing Enzyme Complex in the Inner Ear

2.1

Pattern of Nox3 Expression Across Organs

Nox3 expression is high in the inner ear and embryonic kidney, but minimal in tissues from 19 other organs that have been tested for Nox3 expression using RT-PCR

2.2

Subunit Composition of the NOX3-Containing Enzyme Complex

Most members of the NOX family function in multisubunit complexes. NOX5 is the only member of this family that is not known to require a dedicated interacting protein for its conformational stability [18–20]. Other members of the NOX family (i.e., NOX1–NOX4) require interaction with the transmembrane protein p22phox for conformational stability, and three members (i.e., NOX1–NOX3) require the recruitment of additional proteins to activate ROS production (reviewed in [3]). NOX3 and p22phox stabilize one another. p22phox was discovered as a NOX2-interacting transmembrane protein that is essential for superoxide production in phagocytes [21, 22]. Later studies revealed that p22phox also interacts with NOX3 when the two proteins are co-expressed in transfected cells, and that p22phox forms a heterodimer with nascent NOX3 in the endoplasmic reticulum (ER) [23– 25]. This dimerization stabilizes both NOX3 and p22phox, and it is required for the export of NOX3 from the ER to the Golgi apparatus en route to the plasma membrane

11

Physiological Functions and Pathological Significance of NADPH Oxidase 3

167

A

B

bone cochlea

semicircular canals (Vs) utricle (Vs)

tympanic cavity

ampulla (Vs)

Nox3 expression in single cells (log2[FPKM + 1])

Fig. 11.2 Diagram of endolymphatic and perilymphatic compartments. Endolymphatic spaces (pale purple), perilymphatic spaces (cyan), temporal bone (gray), cochlea, tympanic cavity, endolymphatic sac and duct, and components of the vestibular system (Vs) are indicated. The endolymphatic sac and duct are continuous with the vestibular system but not part of it. Red shading highlights the organs that contain otoconia

n = 44

n = 41

n = 82

n = 46

10

5

0

E12.5

E16.5

P5

P30

Fig. 11.3 Expression of Nox3, Cyba, and Noxo1 in epithelial cells of the endolymphatic sac. (a) Violin plot of levels of Nox3 expression in individual epithelial cells isolated from the endolymphatic sac of mice at the indicated developmental timepoints, as determined based on published scRNA-seq data [28]. Each dot represents an individual cell.

% of Nox3+ cells that express genes of interest (GOI)

[23, 25]. In the absence of p22phox expression, NOX3 is targeted to the ER-associated degradation pathway [23]. Nox3 and p22phox are expressed in the endolymphatic sac and duct during development. The endolymphatic sac is an epithelium-lined pouch that is connected to the gravitysensing organs (utricle and saccule) by the endolymphatic duct (Fig. 11.2). In mice, the p22phox protein is expressed in the apical plasma membrane of many (but not all) epithelial cells in the endolymphatic sac and duct, from approximately embryonic day (E)14 to postnatal day (P)12 [26]. During the

same period, the Nox3 promoter is active in these endolymphatic compartments [27]. To determine whether the p22phoxencoding mRNA (Cyba) and Nox3 are expressed in the same cells, we re-analyzed previously published single-cell RNA-sequencing (scRNA-seq) data that had been generated using cells from the endolymphatic sac of prenatal and young postnatal mice [28]. This analysis revealed that Nox3 expression is the highest around E16.5 (Fig. 11.3a), and that Cyba is expressed in all Nox3-expressing cells at E16.5 (Fig. 11.3b). Thus, gene expression data are consistent with the model that p22phox and NOX3 form a complex in vivo. However, the NOX3 and p22phox proteins have not been co-localized in the inner ear. Such an analysis has been hindered by a lack of validated anti-NOX3 antibodies. In transfected cells, NOX3 is activated by the NOX organizer 1 protein (NOXO1). In cells transfected with NOX3 and CYBA, the NOX3-p22phox dimer produces very little superoxide [24]; however, when the cells are additionally co-transfected with human NOXO1, the production of superoxide is robust (Table 11.1) [24, 25, 29, 30]. NOXO1 was originally identified based on its similarity to p47phox, a cytosolic regulatory subunit of the phagocyte NOX complex [31–34]. Both p47phox and NOXO1 contain two Src homology 3 (SH3) domains, which bind to p22phox [35, 36], as well as a Phox homology (PX) domain, which binds to specific phosphorylated forms of phosphatidylinositol (PIPs) in the plasma membrane [37, 38]. In addition, each of these proteins has been suggested to interact directly with various NOX family members (i.e., NOX1–NOX3) (reviewed in [39]). Each is also known to interact with the cytosolic NOX2 activator protein p67phox, as well as with the p67phox-like protein NOX activator 1 (NOXA1), when expressed in the

endolymphatic sac endolymphatic duct saccule (Vs)

100 80 60 40 20 0

GOI: Cyba Noxo1 Cyba Ncf1 Noxa1 Ncf2 Noxa1 + + Noxo1 Ncf2 Red horizontal lines indicate medians. Numbers of single cells in each group are indicated (n). (b) Co-expression of Nox3 with the indicated genes in epithelial cells isolated from the endolymphatic sac of E16.5 mice, as determined based on published scRNA-seq data [28]

168

Y. Nakano and B. Bánfi

Table 11.1 Effects of NOX-interacting proteins on NOX3 enzymatic activity in transfected cells Tested combinations of human (h) and mouse (m) NOX complex subunits. Organizer Activator Stabilizer subunita subunita NOXa subunitb hNOX3 hNOX3 hp22phox -

Superoxide production. None, -; low, +; high, ++++ – vs.c +

PKC agonist dependent (Yes/No) N N

hNOX3

hp22phox

hNOXO1

-

++++

N

hNOX3

hp22phox

-

hNOXA1

+

N

hNOX3

hp22phox

hNOXO1

hNOXA1

++ vs.c ++++

N

hNOX3 hNOX3 hNOX3

hp22phox hp22phox hp22phox

hp47phox hp47phox

hp67phox hp67phox

– vs.c + ++ vs.c ++++ ++++

Y N Y

hNOX3 hNOX3 mNOX3 mNOX3 mNOX3 mNOX3 mNOX3 mNOX3 mNOX3 mNOX3

hp22phox hp22phox hp22phox hp22phox hp22phox hp22phox hp22phox KDd hp22phox hp22phox hp22phox

hNOXO1 hp47phox mNOXO1 mNOXO1 mNOXO1 hp47phox hp47phox

hp67phox hNOXA1 mNOXA1 mNOXA1 mNOXA1 hp67phox hp67phox

+ vs.c ++++ ++++ + + + ++++ ++ + + ++++

N Y Y Y Y N N Y Y Y

Transfected cell lines CHO HEK293, CHO, COS-7 HEK293, CHO, COS-7 HEK293, CHO, COS-7 HEK293, CHO, COS-7 HEK293, CHO HEK293, CHO HEK293, CHO, COS-7 HEK293, CHO HEK293, CHO HEK293 HEK293 HEK293 HEK293 HEK293 HEK293 HEK293 HEK293

References [24] [24, 25, 29, 30] [24, 25, 29, 30] [24, 25, 29, 30] [24, 25, 29, 30] [24, 25, 29] [24, 25, 29] [24, 25, 29] [24, 29] [24, 29] [15] [15, 30] [15, 30] [15, 23, 30] [23] [15] [15] [15]

a

These subunits were expressed in the transfected cell lines using expression plasmids CHO cells, but not HEK293 and COS-7 cells, were transfected with hp22phox-encoding plasmids because the CHO cell line (in contrast to the other two cell lines) does not express p22phox endogenously [24] c Versus (vs.) indicates discrepancies among data reported by different groups d Protein expression was knocked down (KD) using siRNA b

same cell [33, 35, 40, 41]. The major difference between p47phox and NOXO1 is that p47phox requires a conformational switch, brought about by protein kinase C (PKC)-catalyzed phosphorylation, to bind to p22phox and PIPs, whereas NOXO1 binds to p22phox and PIPs independent of PKC activity (reviewed in [39]). In cells transfected with NOX3, CYBA, and NOXO1, the encoded proteins form a constitutively active superoxide-producing complex in the plasma membrane [24, 25, 30]. In transfected cells, some NOX-interacting proteins can substitute for NOXO1 in regulating NOX3 activity. NOX organizer (i.e., NOXO1 and p47phox) and activator (i.e., NOXA1 and p67phox) proteins have been tested for their effects on NOX3 activity in transfected cells, both individually and in various combinations. In these analyses, co-expression of human NOX3 with human p22phox and NOXO1 resulted in robust production of superoxide in the absence of NOXA1 and p67phox (Table 11.1) [24, 25, 29, 30]. However, when mouse NOX3 was expressed in a mixedspecies complex with human p22phox, its robust activation

required co-expression of both NOXO1 and NOXA1 of mouse origin (Table 11.1) [15]. Regardless of whether the co-expressed subunits originated in one or two species (i.e., human and mouse), co-expression of NOX3 and p22phox with p47phox led to PKC-dependent production of superoxide at a very low rate, and this activity was much higher when the cells were additionally transfected with either NOXA1 or p67phox (Table 11.1) [15, 24, 25, 29]. Finally, co-expression of the NOX3-p22phox dimer with NOXA1 resulted in minimal superoxide production, whereas co-expression of this dimer with p67phox resulted in moderate superoxide production (Table 11.1) [24, 25, 29]. Thus, several combinations of NOX organizer and activator proteins can support the enzymatic activity of p22phox-stabilized NOX3 in transfected cell lines. In transfected cells, the effect of the small GTPase RAC1 on NOX3 activity is dependent on the available NOX organizer subunit. Many studies have shown that, depending on the cell type tested, activation of the phagocyte NOX complex requires either the RAC1 or RAC2 GTPase

11

Physiological Functions and Pathological Significance of NADPH Oxidase 3

(reviewed in [42, 43]). GTP-bound RAC activates the phagocyte NOX by directly interacting with p67phox and inducing a conformational change that increases the affinity of p67phox for NOX2 [44, 45]. GTP-bound RAC also enhances the activity of the NOX1-containing oxidase complex by directly interacting with NOXA1 [25, 40, 46]. These studies motivated two research groups to evaluate the effect of RAC1 on the activation of human NOX3 in transfected cells that also expressed various NOX-interacting proteins of human origin. Their data revealed that the ability of RAC1 to regulate NOX3 activity depends on the NOX organizer subunit that is present in the cell. Specifically, in the presence of p47phox and either p67phox or NOXA1, RAC1 enhances the superoxide-producing activity of p22phox-stabilized NOX3 [25, 47]; however, in the presence of NOXO1, RAC1 does not enhance the activity of p22phox-stabilized NOX3 regardless whether NOXA1 or p67phox is co-expressed [24, 47]. Thus, the effect of RAC1 on NOX3 activity depends on which NOX organizer is part of the NOX3-containing complex. These data are consistent with the notion that the NOX3-p22phox-NOXO1 complex does not contain a RAC1interacting subunit (i.e., p67phox or NOXA1), whereas the NOX3-p22phox-p47phox-p67phox and NOX3-p22phox-p47phoxNOXA1 complexes do. NOX3, p22phox, and NOXO1 are required for balance in mice. Homozygosity for deleterious mutations in Nox3, Cyba, and Noxo1 causes balance defects in mice [26, 30, 48]. Thus, the physiological functions of NOX3, p22phox, and NOXO1 overlap [26]. In addition, the three proteins are likely to form a complex in vivo. If this is true, Nox3, Cyba, and Noxo1 must be expressed in the same cells. Given that previous studies had not evaluated Noxo1 expression in the endolymphatic sac and duct, we again re-analyzed the published scRNA-seq data set that was generated using cells from the endolymphatic sac [28]. This analysis revealed that Noxo1 is expressed in 89% of the cells that express Nox3 and Cyba at E16.5 (Fig. 11.3b). Thus, Nox3, Cyba, and Noxo1 are expressed in many of the same cells in the endolymphatic sac during development of the inner ear. These data support the model that NOXO1 and p22phox are subunits of a NOX3-containing complex in the endolymphatic sac. p47phox, p67phox, and NOXA1 are not required for balance in mice. Phenotypic characterizations have not demonstrated any balance defect in mice that are homozygous for deleterious mutations in the p47phox-encoding gene Ncf1, the p67phox-encoding gene Ncf2, or Noxa1 [49– 51]. These data do not rule out the possibility that the structurally similar proteins p67phox and NOXA1 act redundantly to activate NOX3 in the inner ear. To test this possibility, we once more re-analyzed the endolymphatic sac scRNA-seq data [28]. Our analysis revealed that Ncf2 and Noxa1 are co-expressed in only 14% of the Nox3-expressing cells in this endolymphatic compartment at E16.5 (Fig. 11.3b). Thus,

169

p67phox and NOXA1 are unlikely to act redundantly in epithelial cells of the endolymphatic sac. Further analysis of the scRNA-seq data also revealed that Ncf1 is expressed in only 3% of the Nox3-expressing cells in the endolymphatic sac at E16.5 (Fig. 11.3b). These data are consistent with the model that p67phox, NOXA1, and p47phox are not essential, either individually or in combination, for the physiological function of NOX3 in the inner ear.

3

Physiological Functions of NOX3

3.1

Physiological Function of NOX3 in the Inner Ear

The NOX3-containing enzyme complex is required for the genesis of otoconia in mice. Homozygosity for loss-offunction mutations in Nox3, Cyba, or Noxo1 is sufficient to cause complete agenesis of calcium carbonate (CaCO3) bio-crystals (i.e., otoconia) in the utricle and saccule in mice [26, 30, 48]. In wild-type mice, tens of thousands of otoconia are anchored above mechanosensory hair cells in these structures (Fig. 11.4a, b) (reviewed in [52]). Due to their relatively high density and inertia, otoconia deflect microvilli-like projections (i.e., stereocilia) of the underlying hair cells when the head accelerates, decelerates, or changes position relative to the direction of gravity [53]. The resulting mechanical stimulation of hair cells is fundamental for the detection of linear acceleration and gravity in the inner ear (reviewed in [54]). In mice and rats, otoconial agenesis is associated with balance defects. Balance defects are readily noticeable in mouse lines that are homozygous for inactivating mutations in Nox3, Cyba, or Noxo1. These mice often tilt their head, curl up when lifted by the tail, cannot swim, and cannot stay on a slowly rotating rod longer than a few seconds [26, 30, 48]. In contrast, wild-type mice hold their head straight, stretch out their legs when lifted by the tail, can swim, and can stay on a slowly rotating rod for minutes. A loss-offunction variant of Cyba has also been discovered in the rat strain MES [55], and homozygosity for this mutation is likewise associated with otoconial agenesis and tilted head posture [55]. However, balance and motor coordination have not been evaluated further in this strain. Although swim and rotarod tests are useful for documenting defects in balance and motor coordination, other assessments are required when the goal is to determine whether these defects originate in the utricle and saccule or the central nervous system (CNS). Most frequently, in rodent studies the function of the utricle and saccule are evaluated by measuring vestibular evoked potentials (VsEPs), electrical responses to linear acceleration of the head. These potentials are initiated within the gravitysensing organs and are generated by afferent neurons (early

170

Y. Nakano and B. Bánfi

Fig. 11.4 Scanning electron microscopy images of otoconia. (a) Low- and (b) high-magnification images of saccular otoconia of a wild-type mouse at P1. Arrowheads in panel a indicate stereocilia bundles. Scale bars, 4 μm

peaks in VsEP recordings) and central relays (later peaks) [56, 57], and homozygosity for loss-of-function mutations in any of the above-mentioned genes (Nox3, Cyba, or Noxo1) is associated with their complete absence [26, 58]. Otoconial agenesis-causing mutations in other genes are also associated with a complete absence of VsEPs and with balance defects [59]. These data support the notion that inactivation of the NOX3-p22phox-NOXO1 complex causes balance defects in mice by preventing the formation of otoconia. In humans, p22phox deficiency does not cause symptomatic dysfunction of balance organs. In both mice and rats, homozygosity for p22phox-inactivating mutations is associated with a balance defect and an immune disorder. These defects are consistent with the loss of function of NOX3 and NOX2 [26, 55]. In humans, p22phox-inactivating mutations are also associated with symptoms of an immune disorder, chronic granulomatous disease (CGD). However, they are not associated with symptoms of a balance defect (reviewed in [60]). Thus, it is possible that either p22phox is not necessary for the genesis of otoconia in humans, or otoconia are not required for balance in humans. Although neither possibility can be ruled out based on published data, indirect evidence indicates that otoconial agenesis is compensated to a large extent by the human CNS. To control balance, the CNS integrates sensory inputs from the somatosensory, visual, and vestibular systems. In healthy subjects tested in a well-lit room with a firm floor, balance relies 70% on somatosensory inputs, 10% on visual inputs, and 20% on vestibular inputs [61], and if the environment changes (e.g., lights are turned off) the CNS rapidly re-weighs these inputs.

Long-term re-weighing between vestibular and somatosensory inputs has also been demonstrated with the help of patients who participated in a ‘no visual input’ (i.e., closed eyes) balance test after losing vestibular functions. Specifically, in patients who lost the functions of all balance organs (i.e., utricle, saccule, and semicircular canals) as infants, displacement of the head evoked normal compensatory responses in trunk muscles [62]. However, this was not the case in patients who lost the functions of all of their balance organs as adults [62]. Based on these data we speculate that the human CNS can compensate for prenatal (and early postnatal) loss of function of the utricle and saccule, using somatosensory and visual inputs. Otoconial agenesis has not been diagnosed in the human population. Genetic causes of balance defects in humans have been explored by analyzing DNA samples from subjects who reported symptoms of vestibular dysfunction (reviewed in [63]). These studies have not revealed pathogenic mutations in the orthologs of genes that are required for the genesis of otoconia in mice (i.e., Cyba, Nox3, Noxo1, the Ca2+ pump-encoding gene Atp2b2, the autophagy regulator-encoding gene Atg4b, and the proton channel-encoding gene Otop1) [26, 30, 64–69]. Given the lack of identified cases of otoconial agenesis in the human population, the clinical signs and symptoms of otoconial agenesis are unknown and suitable diagnostic tests have not been described. In mice, VsEP and video ocular counter-roll (vOCR) tests are suitable for demonstrating deficits in utricular and saccular function [70, 71]. Of these, only vOCR has been optimized for use in humans [72, 73]. Therefore, we

11

Physiological Functions and Pathological Significance of NADPH Oxidase 3

171

suggest that vOCR testing in combination with genetic analysis holds the most promise for the identification of cases of otoconial agenesis in the human population. Humans may compensate better than mice for the congenital loss of function of the utricle and saccule. In mice, numerous genes are required for both hearing and balance. However, in humans the orthologs of many of these genes are required specifically for hearing rather than for balance (e.g., WHRN, TMPRSS3, TMIE, TMHS, and LRTOMT) [74–85]. Based on these data, we and others have suggested that the human CNS compensates for congenital impairment of vestibular function more effectively than that of mice [26, 86–88]. Humans and mice also differ with respect to the quality of sensory inputs that regulate the balance output. Mice see very poorly relative to humans. Indeed, if mice were humans, they would qualify as legally blind (reviewed in [89, 90]). Therefore, the effect of visual input on balance is likely smaller in mice. Together, these data support the notion that humans and mice differ with respect to the extent to which they re-weigh visual, somatosensory, and vestibular inputs to compensate for the loss of function of the utricle and saccule in the context of otoconial agenesis.

formation of submicron-sized otoconial precursors [30]. Similarly, scanning electron microscopy analyses of the inner ear of NOX3-deficient and p22phox-deficient mice at later time points failed to detect otoconia of any size [26, 48]. Two models have been proposed to explain the effects of NOX3, p22phox, and NOXO1 on the genesis of otoconia. According to the first to be proposed [48], NOX3-dependent oxidation of otoconial proteins causes conformation changes that trigger either the nucleation or aggregation of CaCO3 nanocrystals. According to the second model that was proposed [26], the dismutation of NOX3-generated superoxide to H2O2 and O2 (Fig. 11.1a, second equation) consumes H+ from the endolymph and creates pH conditions that favor the acid-base reaction between Ca2+ and bicarbonate. We suggest that these two hypotheses are compatible. Specifically, NOX3-generated superoxide may elevate the endolymphatic pH by consuming H+ in the dismutation reaction, and the H2O2 product from this reaction may alter sulfhydryl or other residues in otoconial proteins such that they favor the nucleation or aggregation of CaCO3 nanocrystals.

3.2

Several studies reported that NOX3 generates ROS in various cell types that do not reside in the inner ear, including spermatogonial stem cells, oligodendrocytes, and B lymphocytes [97–99]. In the same studies, the NOX3generated ROS were proposed to regulate the self-renewal of spermatogonial stem cells, differentiation of oligodendrocytes, and activation of B lymphocytes. These potential functions of NOX3 have not been validated in vivo using Nox3 mutant animals, and their physiological importance is currently unknown.

Proposed Roles of NOX3 in the Genesis of Otoconia

The genesis of otoconia is limited to a specific period during development. Fully formed otoconia are ~10 μm calcite bio-crystals that consist of CaCO3 (~95% by weight) and proteins (~5% by weight) [91]. In mice, otoconial nucleation starts at E14.5, and it is limited to the space above the sensory epithelium in the utricle and saccule (reviewed in [92]). Thus, local concentrations of ions or secreted proteins within these endolymphatic compartments favor the formation of calcite bio-crystals. Nucleation of CaCO3 crystals starts at aminoacyl moieties or oligosaccharide side chains of secreted glycoproteins, and the nucleating nanocrystals incorporate both CaCO3 and glycoproteins (reviewed in [93]). During the growth phase, the nanocrystals agglomerate to form calcite mesocrystals, through a process that is regulated by several otoconial proteins (reviewed in [54]), and the mineral growth persists until ~P7 [94, 95]. Analyses of otoconial morphology and CaCO3 turnover have not demonstrated the nucleation of new otoconial crystals after P7 in mice [96]. Thus, new otoconia are likely not produced in mammals after a specific stage in development. The NOX3-containing enzyme complex is necessary for some of the earliest steps in the genesis of otoconia. Transmission electron microscopy-based analysis of the saccule of E16.5 NOXO1-deficient mice showed that inactivation of the NOX3-containing enzyme complex prevents the

3.3

Physiological Functions of NOX3 Outside of the Inner Ear

4

Pathological Significance of NOX3

4.1

Pathological Effects of NOX3 in the Inner Ear

Cisplatin causes hearing loss in the high-frequency range. Cisplatin is a chemotherapeutic drug that is frequently used in the treatment of solid tumors. Its dose-limiting side effects include ototoxicity, neurotoxicity, and nephrotoxicity (reviewed in [100, 101]). The ototoxicity of cisplatin is evident from high-frequency hearing loss in ~50% of the patients who are treated with this drug [102]. Cisplatindamaged inner ears are characterized by the degeneration of hair cells as well as spiral ganglion neurons in the basal turn of the hearing organ (reviewed in [101]). Currently, no effective treatment for cisplatin-induced hearing loss is available.

172

Genetic deletion of Nox3 protects mice from cisplatininduced hearing loss. In heterologous expression systems, cisplatin leads to an increase in the amount of superoxide that is produced by the NOX3-containing enzyme complex [15]. This effect is not specific to NOX3; cisplatin also increases the amount of NOX5-generated superoxide [15]. Although the underlying molecular mechanism remains unclear, the demonstration that cisplatin enhances superoxide production motivated careful analyses of Nox3 expression in the cochlea. This led to the identification of a number of Nox3-expressing cell types, including spiral ganglion neurons, inner hair cells, outer hair cells, Deiters’ cells, Claudius’ cells, and root cells [15, 27, 103]. Within each of these cell types, Nox3 expression is highly variable [27]. In some, NOX3 is likely active and damaging during cisplatin treatment because genetic deletion of Nox3 protects mice from cisplatin-induced ototoxicity [27]. Also, siRNAmediated knock-down of cochlear Nox3 expression protects rats from cisplatin-induced hearing loss [104, 105]. Thus, a NOX3-containing enzyme complex mediates some of the ototoxic effects of cisplatin in the cochlea, though its subunit composition remains to be defined. Together, the studies summarized here raise the intriguing possibility that cisplatin-induced ototoxicity can be prevented by co-administering NOX3-specific inhibitors with cisplatin. In mice, genetic inactivation of p22phox and NOX3 delays age-related hearing loss (AHL). AHL occurs frequently in the human population, affecting 1 in 3 adults over the age of 65 (reviewed in [106]). The etiology of AHL is complex, with both genetic and environmental factors contributing (reviewed in [107]). AHL incidence is also frequent in inbred strains of mice [108], and the genetic background is an excellent predictor of AHL onset [109]. In the popular A/J strain, hearing loss starts 3–4 weeks after birth and is caused by hypomorphic mutations in the cadherin-23 and citrate synthase genes (Cdh23ahl/ahl; Csahl4/ ahl4 ) [109–111]. A recent study demonstrated that the hearing loss of A/J mice can be delayed by genetic inactivation of p22phox [103]. This effect is likely related to the NOX3stabilizing function of p22phox because genetic inactivation of NOX3 also delays AHL in C57BL/6 mice, which are homozygous for the same Cdh23ahl allele as the A/J mice [27]. In both the A/J and C57BL/6 strains, AHL is associated with progressive loss of outer hair cells as well as degeneration of synapses between inner hair cells and spiral ganglion neurons [110, 112]. Both of these pathologies are delayed by the inactivation of p22phox and NOX3 [27, 103]. Thus, p22phox and NOX3 accelerate AHL in mice. Further studies are required to determine whether this effect of p22phox and NOX3 is specific to mice that are homozygous for the Cdh23ahl allele. NOX3 and p22phox deficiencies protect mice from noise-induced hearing loss (NIHL). Three research groups

Y. Nakano and B. Bánfi

have evaluated the effect of NOX3 on NIHL in mice [27, 113, 114]. They exposed NOX3-deficient and wildtype mice to broadband noise and tested the ability of the noise-exposed mice to hear pure tones of various frequencies. Two of the three groups reported that genetic inactivation of NOX3 protects mice from NIHL in the high-frequency range [27, 114], whereas the third reported that genetic inactivation of NOX3 sensitizes mice to NIHL in the low-frequency range [113]. To resolve this controversy, one of the three groups also analyzed the p22phox-deficient Cyba-/- mice for the severity of NIHL [114]. Their data indicate that noise induces less severe high-frequency hearing loss in Cyba-/- mice than in wild-type littermates. Based on these data, NOX3 inhibition has been proposed as a possible approach to preventing NIHL [114].

4.2

Pathological Effects of NOX3 Outside of the Inner Ear

NOX3 has been suggested to play roles in many pathological processes that are unrelated to hearing loss. Most of the pathological effects were proposed based on associations of single-nucleotide polymorphisms (SNPs) in NOX3 with various diseases, upregulation of NOX3 expression in certain tissues under pathological conditions, or NOX3-dependent alterations in metabolism and signaling in cell culture systems [115–130]. These effects are not described in detail here because they have not been validated in vivo using NOX3-deficient or over-expressor animals. Our description of the pathological effects of NOX3 is based on studies that used Nox3 mutant mice to identify damaging effects of this enzyme. NOX3 contributes to the inflammatory effects of hyperoxia in mice. Hyperoxia causes upregulation of Nox3 expression in lung endothelial cells and a massive inflammatory reaction in the lungs [131]. The inflammatory reaction is probably a consequence of the increase in Nox3 levels because genetic knock-out (Nox3-/-) is associated with abnormally low inflammation in the lungs of hyperoxiaexposed mice [131]. Nevertheless, much of the hyperoxiacaused tissue damage is likely independent of NOX3 activity because Nox3-/- and wild-type mice die at approximately the same time following exposure to 100% oxygen [131]. Thus, NOX3-dependent lung inflammation might be one of several mechanisms by which hyperoxia damages the lungs. A novel variant of Nox3 is associated with histological alterations in the cerebellum. A recent study reported that homozygosity for a chemically induced missense mutation in Nox3 (Nox3eqlb) is associated with abnormally high production of superoxide in cerebellar neurons, excessive proliferation of cerebellar granule cell precursors, and defects in

11

Physiological Functions and Pathological Significance of NADPH Oxidase 3

motor coordination in mice [132]. These data suggest that Nox3eqlb is a gain-of-function mutation. However, the enzymatic source of cerebellar superoxide in Nox3eqlb/eqlb mice has not been identified. Also, the superoxide-producing activity of the Nox3eqlb-encoded protein (i.e., NOX3N64Y) has not been evaluated in transfected cells. Therefore, we suggest that further studies are needed to clarify whether it is abnormally high, or abnormally low, NOX3 activity that causes developmental defects in Nox3eqlb/eqlb mice.

5

Conclusions

NOX3 is a superoxide-producing enzyme that plays a constructive role in the development of the inner ear. This role is demonstrated by the lack of otoconia in NOX3deficient mouse lines. NOX3 functions in a multi-subunit complex whose transmembrane ‘core’ is the NOX3-p22phox dimer. Additionally, cytosolic subunit(s) are necessary for robust superoxide production by NOX3. In the endolymphatic sac and duct, this subunit is NOXO1. In other organs (e.g., the cochlea), the cytosolic subunits of the NOX3containing complex have not been identified. Although NOX3 function may not be limited to the genesis of otoconia, to date this is the only physiological role that has been demonstrated using NOX3-deficient animals. Analysis of the ionic composition of the endolymph will be a crucial step towards defining the role of NOX3 in the genesis of otoconia. The precise role of the NOX3-p22phoxNOXO1 complex in the genesis of otoconia is unknown. The proposed effects of this complex on the pH of the endolymph and the conformation of otoconial proteins remain to be tested. We suggest that a critical step towards defining the role of superoxide in otoconial genesis will be to analyze the ionic composition of the endolymph (e.g., pH, as well as Ca2+ and HCO3- concentrations) at the time that otoconia are nucleated. Although the anatomical locations and small sizes of the developing utricle and saccule make such an analysis challenging, it is expected to achieve two important goals. One is to reveal potential differences in the endolymphatic pH between Nox3-/- and wild-type mice. The second is to define baseline ionic conditions for in vitro reconstitution of otoconial crystal formation, an assay that will likely be crucial to determining whether superoxide regulates the nucleation of otoconia by modifying otoconial proteins. vOCR tests are suitable for evaluating the functions of the utricle and saccule. In mice and rats, p22phoxinactivating mutations disrupt otoconial genesis and impair balance. In contrast, in humans p22phox-inactivating mutations are not associated with any symptom of a vestibular dysfunction. Thus, it is possible that in humans p22phox is not needed for the genesis of otoconia, or that otoconial agenesis is asymptomatic in humans. It is currently difficult

173

to distinguish between these two possibilities because the symptoms of otoconial agenesis of any etiology have not been defined. To evaluate specifically the utricular and saccular function, vOCR tests have been developed. These tests are effective in diagnosing idiopathic loss of function of the utricle and saccule in humans, as well as otoconial agenesisassociated vestibular dysfunction in mice. Therefore, we suggest that vOCR testing holds the greatest promise for detecting possible dysfunction of the utricle and saccule in subjects who are homozygous for inactivating mutations in genes related to the genesis of otoconia. NOX3 is a drug target in cisplatin ototoxicity. Both systemic and local administration of ROS scavengers has been reported to reduce the ototoxicity of cisplatin in rodents (reviewed in [101]). Given that genetic deletion of Nox3 prevents many of the ototoxic effects of cisplatin in mice, most of the damaging ROS are likely produced by NOX3 in the inner ear. Currently, no NOX3-specific inhibitors are available, so it is not possible to take advantage of these findings. However, NOX inhibitors are being developed at a rapid pace [133]. Some are already being tested in phase II clinical trials, though not for the prevention of cisplatin ototoxicity (see Chap. 21 by C.M. Dustin, E. CifuentesPagano, and P.J. Pagano). What does the future hold for NOX3 research? When the NOX3 protein sequence was described ~20 years ago [16], nothing was known about the inner ear expression of this protein, its physiological function, its pathological significance, or the subunit requirements for its activity. Thus, much has been achieved in NOX3 research since then. We expect that the next 20 years will bring many new tools and discoveries to the field. These might include high-resolution structural information about the NOX3-p22phox dimer, NOX3-specific inhibitors, refinements of the model of otoconial formation and, perhaps, the identification of human cases of otoconial agenesis. Acknowledgments We thank Dr. Christine Blaumueller for critical review of the manuscript. This project was supported by a grant from the National Institute on Deafness and Other Communication Disorders (https://www.nih.gov/R01DC014953 to B Bánfi) and by resources of the Iowa City Department of Veterans Affairs Medical Center.

References 1. Cross AR, Segal AW (2004) The NADPH oxidase of professional phagocytes--prototype of the NOX electron transport chain systems. Biochim Biophys Acta 1657:1–22. https://doi.org/10. 1016/j.bbabio.2004.03.008 2. Magnani F, Nenci S, Millana Fananas E et al (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci U S A 114:6764–6769. https://doi.org/10.1073/pnas. 1702293114

174 3. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313. https://doi.org/10.1152/physrev.00044.2005 4. Lambeth JD, Neish AS (2014) Nox enzymes and new thinking on reactive oxygen: a double-edged sword revisited. Annu Rev Pathol 9:119–145. https://doi.org/10.1146/annurev-pathol012513-104651 5. Royer-Pokora B, Kunkel LM, Monaco AP et al (1986) Cloning the gene for an inherited human disorder chronic granulomatous disease on the basis of its chromosomal location. Nature 322:32–38. https://doi.org/10.1038/322032a0 6. Nauseef WM, Clark RA (2019) Intersecting stories of the phagocyte NADPH oxidase and chronic granulomatous disease. Methods Mol Biol 1982:3–16. https://doi.org/10.1007/978-1-4939-9424-3_ 1 7. Nauseef WM (2014) Myeloperoxidase in human neutrophil host defence. Cell Microbiol 16:1146–1155. https://doi.org/10.1111/ cmi.12312 8. Rada B, Lekstrom K, Damian S et al (2008) The pseudomonas toxin pyocyanin inhibits the dual oxidase-based antimicrobial system as it imposes oxidative stress on airway epithelial cells. J Immunol 181:4883–4893. https://doi.org/10.4049/jimmunol.181. 7.4883 9. Moskwa P, Lorentzen D, Excoffon KJ et al (2007) A novel host defense system of airways is defective in cystic fibrosis. Am J Respir Crit Care Med 175:174–183. https://doi.org/10.1164/rccm. 200607-1029OC 10. Geiszt M, Witta J, Baffi J et al (2003) Dual oxidases represent novel hydrogen peroxide sources supporting mucosal surface host defense. FASEB J 17:1502–1504. https://doi.org/10.1096/fj.021104fje 11. Conner GE, Wijkstrom-Frei C, Randell SH et al (2007) The lactoperoxidase system links anion transport to host defense in cystic fibrosis. FEBS Lett 581:271–278. https://doi.org/10.1016/j. febslet.2006.12.025 12. Gerson C, Sabater J, Scuri M et al (2000) The lactoperoxidase system functions in bacterial clearance of airways. Am J Respir Cell Mol Biol 22:665–671. https://doi.org/10.1165/ajrcmb.22.6. 3980 13. Lorentzen D, Durairaj L, Pezzulo AA et al (2011) Concentration of the antibacterial precursor thiocyanate in cystic fibrosis airway secretions. Free Radic Biol Med 50:1144–1150. https://doi.org/ 10.1016/j.freeradbiomed.2011.02.013 14. Vermot A, Petit-Härtlein I, Smith SME et al (2021) NADPH oxidases (NOX): an overview from discovery, molecular mechanisms to physiology and pathology. Antioxidants (Basel) 10. https://doi.org/10.3390/antiox10060890 15. Bánfi B, Malgrange B, Knisz J et al (2004) NOX3, a superoxidegenerating NADPH oxidase of the inner ear. J Biol Chem 279: 46065–46072. https://doi.org/10.1074/jbc.M403046200 16. Cheng G, Cao Z, Xu X et al (2001) Homologs of gp91phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269:131–140. https://doi.org/10.1016/s0378-1119(01)00449-8 17. Tabula Muris Consortium (2018) Single-cell transcriptomics of 20 mouse organs creates a Tabula Muris. Nature 562:367–372. https://doi.org/10.1038/s41586-018-0590-4 18. Bánfi B, Molnár G, Maturana A et al (2001) A Ca(2+)-activated NADPH oxidase in testis, spleen, and lymph nodes. J Biol Chem 276:37594–37601. https://doi.org/10.1074/jbc.M103034200 19. Bánfi B, Tirone F, Durussel I et al (2004) Mechanism of Ca2+ activation of the NADPH oxidase 5 (NOX5). J Biol Chem 279: 18583–18591. https://doi.org/10.1074/jbc.M310268200 20. Kawahara T, Jackson HM, Smith SM et al (2011) Nox5 forms a functional oligomer mediated by self-association of its dehydrogenase domain. Biochemistry 50:2013–2025. https://doi.org/10. 1021/bi1020088

Y. Nakano and B. Bánfi 21. Dinauer MC, Pierce EA, Bruns GA et al (1990) Human neutrophil cytochrome b light chain (p22-phox). Gene structure, chromosomal location, and mutations in cytochrome-negative autosomal recessive chronic granulomatous disease. J Clin Invest 86:1729– 1737. https://doi.org/10.1172/JCI114898 22. Parkos CA, Allen RA, Cochrane CG et al (1987) Purified cytochrome b from human granulocyte plasma membrane is comprised of two polypeptides with relative molecular weights of 91,000 and 22,000. J Clin Invest 80:732–742. https://doi.org/10.1172/ JCI113128 23. Nakano Y, Banfi B, Jesaitis AJ et al (2007) Critical roles for p22phox in the structural maturation and subcellular targeting of Nox3. Biochem J 403:97–108. https://doi.org/10.1042/bj20060819 24. Ueno N, Takeya R, Miyano K et al (2005) The NADPH oxidase Nox3 constitutively produces superoxide in a p22phox-dependent manner: its regulation by oxidase organizers and activators. J Biol Chem 280:23328–23339. https://doi.org/10.1074/jbc. M414548200 25. Ueyama T, Geiszt M, Leto TL (2006) Involvement of Rac1 in activation of multicomponent Nox1- and Nox3-based NADPH oxidases. Mol Cell Biol 26:2160–2174. https://doi.org/10.1128/ mcb.26.6.2160-2174.2006 26. Nakano Y, Longo-Guess CM, Bergstrom DE et al (2008) Mutation of the Cyba gene encoding p22phox causes vestibular and immune defects in mice. J Clin Invest 118:1176–1185. https://doi.org/10. 1172/jci33835 27. Mohri H, Ninoyu Y, Sakaguchi H et al (2021) Nox3-derived superoxide in cochleae induces sensorineural hearing loss. J Neurosci 41:4716–4731. https://doi.org/10.1523/jneurosci. 2672-20.2021 28. Honda K, Kim SH, Kelly MC et al (2017) Molecular architecture underlying fluid absorption by the developing inner ear. Elife 6. https://doi.org/10.7554/eLife.26851 29. Cheng G, Ritsick D, Lambeth JD (2004) Nox3 regulation by NOXO1, p47phox, and p67phox. J Biol Chem 279:34250– 34255. https://doi.org/10.1074/jbc.M400660200 30. Kiss PJ, Knisz J, Zhang Y et al (2006) Inactivation of NADPH oxidase organizer 1 results in severe imbalance. Curr Biol 16:208– 213. https://doi.org/10.1016/j.cub.2005.12.025 31. Bánfi B, Clark RA, Steger K et al (2003) Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 278:3510–3513. https://doi.org/10.1074/jbc. C200613200 32. Geiszt M, Lekstrom K, Witta J et al (2003) Proteins homologous to p47phox and p67phox support superoxide production by NAD (P)H oxidase 1 in colon epithelial cells. J Biol Chem 278:20006– 20012. https://doi.org/10.1074/jbc.M301289200 33. Takeya R, Ueno N, Kami K et al (2003) Novel human homologues of p47phox and p67phox participate in activation of superoxideproducing NADPH oxidases. J Biol Chem 278:25234–25246. https://doi.org/10.1074/jbc.M212856200 34. Volpp BD, Nauseef WM, Donelson JE et al (1989) Cloning of the cDNA and functional expression of the 47-kilodalton cytosolic component of human neutrophil respiratory burst oxidase. Proc Natl Acad Sci U S A 86:7195–7199. https://doi.org/10.1073/ pnas.86.18.7195 35. Sumimoto H, Kage Y, Nunoi H et al (1994) Role of Src homology 3 domains in assembly and activation of the phagocyte NADPH oxidase. Proc Natl Acad Sci U S A 91:5345–5349. https://doi.org/ 10.1073/pnas.91.12.5345 36. de Mendez I, Homayounpour N, Leto TL (1997) Specificity of p47phox SH3 domain interactions in NADPH oxidase assembly and activation. Mol Cell Biol 17:2177–2185. https://doi.org/10. 1128/MCB.17.4.2177 37. Davis NY, McPhail LC, Horita DA (2012) The NOXO1β PX domain preferentially targets PtdIns(4,5)P2 and PtdIns(3,4,5)P3. J Mol Biol 417:440–453. https://doi.org/10.1016/j.jmb.2012.01.058

11

Physiological Functions and Pathological Significance of NADPH Oxidase 3

38. Kanai F, Liu H, Field SJ et al (2001) The PX domains of p47phox and p40phox bind to lipid products of PI(3)K. Nat Cell Biol 3:675– 678. https://doi.org/10.1038/35083070 39. Sumimoto H, Minakami R, Miyano K (2019) Soluble regulatory proteins for activation of NOX family NADPH oxidases. Methods Mol Biol 1982:121–137. https://doi.org/10.1007/978-1-49399424-3_8 40. Cheng G, Diebold BA, Hughes Y et al (2006) Nox1-dependent reactive oxygen generation is regulated by Rac1. J Biol Chem 281: 17718–17726. https://doi.org/10.1074/jbc.M512751200 41. Leto TL, Adams AG, de Mendez I (1994) Assembly of the phagocyte NADPH oxidase: binding of Src homology 3 domains to proline-rich targets. Proc Natl Acad Sci U S A 91:10650–10654. https://doi.org/10.1073/pnas.91.22.10650 42. Pick E (2014) Role of the Rho GTPase Rac in the activation of the phagocyte NADPH oxidase: outsourcing a key task. Small GTPases 5:e27952. https://doi.org/10.4161/sgtp.27952 43. Bokoch GM, Diebold BA (2002) Current molecular models for NADPH oxidase regulation by Rac GTPase. Blood 100:2692– 2696. https://doi.org/10.1182/blood-2002-04-1149 44. Bechor E, Zahavi A, Berdichevsky Y et al (2021) The molecular basis of Rac-GTP action-promoting binding of p67(phox) to Nox2 by disengaging the β hairpin from downstream residues. J Leukoc Biol 110:219–237. https://doi.org/10.1002/JLB.4HI1220-855RR 45. Sarfstein R, Gorzalczany Y, Mizrahi A et al (2004) Dual role of Rac in the assembly of NADPH oxidase, tethering to the membrane and activation of p67phox: a study based on mutagenesis of p67phox-Rac1 chimeras. J Biol Chem 279:16007–16016. https:// doi.org/10.1074/jbc.M312394200 46. Miyano K, Ueno N, Takeya R et al (2006) Direct involvement of the small GTPase Rac in activation of the superoxide-producing NADPH oxidase Nox1. J Biol Chem 281:21857–21868. https:// doi.org/10.1074/jbc.M513665200 47. Miyano K, Sumimoto H (2007) Role of the small GTPase Rac in p22phox-dependent NADPH oxidases. Biochimie 89:1133–1144. https://doi.org/10.1016/j.biochi.2007.05.003 48. Paffenholz R, Bergstrom RA, Pasutto F et al (2004) Vestibular defects in head-tilt mice result from mutations in Nox3, encoding an NADPH oxidase. Genes Dev 18:486–491. https://doi.org/10. 1101/gad.1172504 49. Flaherty JP, Spruce CA, Fairfield HE et al (2010) Generation of a conditional null allele of NADPH oxidase activator 1 (NOXA1). Genesis 48:568–575. https://doi.org/10.1002/dvg.20655 50. Jackson SH, Gallin JI, Holland SM (1995) The p47phox mouse knock-out model of chronic granulomatous disease. J Exp Med 182:751–758. https://doi.org/10.1084/jem.182.3.751 51. Jacob CO, Yu N, Yoo DG et al (2017) Haploinsufficiency of NADPH oxidase subunit neutrophil cytosolic factor 2 is sufficient to accelerate full-blown lupus in NZM 2328 mice. Arthritis Rheumatol 69:1647–1660. https://doi.org/10.1002/art.40141 52. Thalmann R, Ignatova E, Kachar B et al (2001) Development and maintenance of otoconia: biochemical considerations. Ann N Y Acad Sci 942:162–178. https://doi.org/10.1111/j.1749-6632.2001. tb03743.x 53. Lins U, Farina M, Kurc M et al (2000) The otoconia of the Guinea pig utricle: internal structure, surface exposure, and interactions with the filament matrix. J Struct Biol 131:67–78. https://doi.org/ 10.1006/jsbi.2000.4260 54. Lundberg YW, Xu Y, Thiessen KD et al (2015) Mechanisms of otoconia and otolith development. Dev Dyn 244:239–253. https:// doi.org/10.1002/dvdy.24195 55. Mori M, Li G, Hashimoto M et al (2009) Pivotal advance: eosinophilia in the MES rat strain is caused by a loss-of-function mutation in the gene for cytochrome b(-245), alpha polypeptide (Cyba). J Leukoc Biol 86:473–478. https://doi.org/10.1189/jlb.1108715 56. Jones TA, Jones SM, Vijayakumar S et al (2011) The adequate stimulus for mammalian linear vestibular evoked potentials

175

(VsEPs). Hear Res 280:133–140. https://doi.org/10.1016/j.heares. 2011.05.005 57. Nazareth AM, Jones TA (1998) Central and peripheral components of short latency vestibular responses in the chicken. J Vestib Res 8: 233–252 58. Jones SM, Erway LC, Bergstrom RA et al (1999) Vestibular responses to linear acceleration are absent in otoconia-deficient C57BL/6JEi-het mice. Hear Res 135:56–60. https://doi.org/10. 1016/s0378-5955(99)00090-8 59. Jones SM, Erway LC, Johnson KR et al (2004) Gravity receptor function in mice with graded otoconial deficiencies. Hear Res 191: 34–40. https://doi.org/10.1016/j.heares.2004.01.008 60. Stasia MJ (2016) CYBA encoding p22(phox), the cytochrome b558 alpha polypeptide: gene structure, expression, role and physiopathology. Gene 586:27–35. https://doi.org/10.1016/j.gene. 2016.03.050 61. Peterka RJ (2002) Sensorimotor integration in human postural control. J Neurophysiol 88:1097–1118. https://doi.org/10.1152/jn. 2002.88.3.1097 62. Horak FB, Shupert CL, Dietz V et al (1994) Vestibular and somatosensory contributions to responses to head and body displacements in stance. Exp Brain Res 100:93–106. https://doi. org/10.1007/BF00227282 63. Eppsteiner RW, Smith RJ (2011) Genetic disorders of the vestibular system. Curr Opin Otolaryngol Head Neck Surg 19:397–402. https://doi.org/10.1097/MOO.0b013e32834a9852 64. Hurle B, Ignatova E, Massironi SM et al (2003) Non-syndromic vestibular disorder with otoconial agenesis in tilted/mergulhador mice caused by mutations in otopetrin 1. Hum Mol Genet 12:777– 789. https://doi.org/10.1093/hmg/ddg087 65. Kozel PJ, Friedman RA, Erway LC et al (1998) Balance and hearing deficits in mice with a null mutation in the gene encoding plasma membrane Ca2+-ATPase isoform 2. J Biol Chem 273: 18693–18696. https://doi.org/10.1074/jbc.273.30.18693 66. Mariño G, Fernández AF, Cabrera S et al (2010) Autophagy is essential for mouse sense of balance. J Clin Invest 120:2331–2344. https://doi.org/10.1172/JCI42601 67. Ornitz DM, Bohne BA, Thalmann I et al (1998) Otoconial agenesis in tilted mutant mice. Hear Res 122:60–70. https://doi.org/10.1016/ s0378-5955(98)00080-x 68. Flaherty JP, Fairfield HE, Spruce CA et al (2011) Molecular characterization of an allelic series of mutations in the mouse Nox3 gene. Mamm Genome 22:156–169. https://doi.org/10.1007/ s00335-010-9309-z 69. Tu YH, Cooper AJ, Teng B et al (2018) An evolutionarily conserved gene family encodes proton-selective ion channels. Science 359:1047–1050. https://doi.org/10.1126/science.aao3264 70. Harrod CG, Baker JF (2003) The vestibulo ocular reflex (VOR) in otoconia deficient head tilt (het) mutant mice versus wild type C57BL/6 mice. Brain Res 972:75–83. https://doi.org/10.1016/ s0006-8993(03)02505-8 71. Ward BK, Lee YH, Roberts DC et al (2018) Mouse magnetic-field nystagmus in strong static magnetic fields is dependent on the presence of Nox3. Otol Neurotol 39:e1150–e1159. https://doi. org/10.1097/mao.0000000000002024 72. Millar JL, Gimmon Y, Roberts D et al (2020) Improvement after vestibular rehabilitation not explained by improved passive VOR gain. Front Neurol 11:79. https://doi.org/10.3389/fneur.2020. 00079 73. Sadeghpour S, Fornasari F, Otero-Millan J et al (2021) Evaluation of the video ocular counter-roll (vOCR) as a new clinical test of otolith function in peripheral Vestibulopathy. JAMA Otolaryngol Head Neck Surg 147:518–525. https://doi.org/10.1001/jamaoto. 2021.0176 74. Holme RH, Kiernan BW, Brown SD et al (2002) Elongation of hair cell stereocilia is defective in the mouse mutant whirler. J Comp Neurol 450:94–102. https://doi.org/10.1002/cne.10301

176 75. Ebermann I, Scholl HP, Charbel Issa P et al (2007) A novel gene for Usher syndrome type 2: mutations in the long isoform of whirlin are associated with retinitis pigmentosa and sensorineural hearing loss. Hum Genet 121:203–211. https://doi.org/10.1007/ s00439-006-0304-0 76. Mburu P, Mustapha M, Varela A et al (2003) Defects in whirlin, a PDZ domain molecule involved in stereocilia elongation, cause deafness in the whirler mouse and families with DFNB31. Nat Genet 34:421–428. https://doi.org/10.1038/ng1208 77. Fasquelle L, Scott HS, Lenoir M et al (2011) Tmprss3, a transmembrane serine protease deficient in human DFNB8/10 deafness, is critical for cochlear hair cell survival at the onset of hearing. J Biol Chem 286:17383–17397. https://doi.org/10.1074/jbc.M110. 190652 78. Scott HS, Kudoh J, Wattenhofer M et al (2001) Insertion of betasatellite repeats identifies a transmembrane protease causing both congenital and childhood onset autosomal recessive deafness. Nat Genet 27:59–63. https://doi.org/10.1038/83768 79. Mitchem KL, Hibbard E, Beyer LA et al (2002) Mutation of the novel gene Tmie results in sensory cell defects in the inner ear of spinner, a mouse model of human hearing loss DFNB6. Hum Mol Genet 11:1887–1898. https://doi.org/10.1093/hmg/11.16.1887 80. Naz S, Giguere CM, Kohrman DC et al (2002) Mutations in a novel gene, TMIE, are associated with hearing loss linked to the DFNB6 locus. Am J Hum Genet 71:632–636. https://doi.org/10.1086/ 342193 81. Longo-Guess CM, Gagnon LH, Cook SA et al (2005) A missense mutation in the previously undescribed gene Tmhs underlies deafness in hurry-scurry (hscy) mice. Proc Natl Acad Sci U S A 102: 7894–7899. https://doi.org/10.1073/pnas.0500760102 82. Kalay E, Li Y, Uzumcu A et al (2006) Mutations in the lipoma HMGIC fusion partner-like 5 (LHFPL5) gene cause autosomal recessive nonsyndromic hearing loss. Hum Mutat 27:633–639. https://doi.org/10.1002/humu.20368 83. Shabbir MI, Ahmed ZM, Khan SY et al (2006) Mutations of human TMHS cause recessively inherited non-syndromic hearing loss. J Med Genet 43:634–640. https://doi.org/10.1136/jmg.2005. 039834 84. Ahmed ZM, Masmoudi S, Kalay E et al (2008) Mutations of LRTOMT, a fusion gene with alternative reading frames, cause nonsyndromic deafness in humans. Nat Genet 40:1335–1340. https://doi.org/10.1038/ng.245 85. Du X, Schwander M, Moresco EM et al (2008) A catechol-Omethyltransferase that is essential for auditory function in mice and humans. Proc Natl Acad Sci U S A 105:14609–14614. https://doi. org/10.1073/pnas.0807219105 86. Cryns K, van Alphen AM, van Spaendonck MP et al (2004) Circling behavior in the Ecl mouse is caused by lateral semicircular canal defects. J Comp Neurol 468:587–595. https://doi.org/10. 1002/cne.10975 87. Ohlemiller KK, Jones SM, Johnson KR (2016) Application of mouse models to research in hearing and balance. J Assoc Res Otolaryngol 17:493–523. https://doi.org/10.1007/s10162-0160589-1 88. Jones SM, Jones TA (2014) Genetics of peripheral vestibular dysfunction: lessons from mutant mouse strains. J Am Acad Audiol 25:289–301. https://doi.org/10.3766/jaaa.25.3.8 89. Baker M (2013) Neuroscience: through the eyes of a mouse. Nature 502:156–158. https://doi.org/10.1038/502156a 90. Brenowitz EA, Zakon HH (2015) Emerging from the bottleneck: benefits of the comparative approach to modern neuroscience. Trends Neurosci 38:273–278. https://doi.org/10.1016/j.tins.2015. 02.008 91. Kniep R, Zahn D, Wulfes J et al (2017) The sense of balance in humans: structural features of otoconia and their response to linear

Y. Nakano and B. Bánfi acceleration. PLoS One 12:e0175769. https://doi.org/10.1371/ journal.pone.0175769 92. Lundberg YW, Zhao X, Yamoah EN (2006) Assembly of the otoconia complex to the macular sensory epithelium of the vestibule. Brain Res 1091:47–57. https://doi.org/10.1016/j.brainres. 2006.02.083 93. Kniep R (2015) Otoconia: mimicking a calcite-based functional material of the human body. From basic research to medical aspects. Pure Appl Chem 87:719–736 94. Anniko M (1980) Development of otoconia. Am J Otolaryngol 1: 400–410. https://doi.org/10.1016/s0196-0709(80)80021-4 95. Lim DJ (1973) Formation and fate of the otoconia. Scanning and transmission electron microscopy. Ann Otol Rhinol Laryngol 82: 23–35. https://doi.org/10.1177/000348947308200109 96. Kawamata S, Igarashi Y (1995) Growth and turnover of rat otoconia as revealed by labeling with tetracycline. Anat Rec 242: 259–266. https://doi.org/10.1002/ar.1092420216 97. Accetta R, Damiano S, Morano A et al (2016) Reactive oxygen species derived from NOX3 and NOX5 drive differentiation of human oligodendrocytes. Front Cell Neurosci 10:146. https://doi. org/10.3389/fncel.2016.00146 98. Morimoto H, Kanatsu-Shinohara M, Shinohara T (2015) ROS-generating oxidase Nox3 regulates the self-renewal of mouse spermatogonial stem cells. Biol Reprod 92:147. https:// doi.org/10.1095/biolreprod.114.127647 99. Feng YY, Tang M, Suzuki M et al (2019) Essential role of NADPH oxidase-dependent production of reactive oxygen species in maintenance of sustained B cell receptor signaling and B cell proliferation. J Immunol 202:2546–2557. https://doi.org/10.4049/ jimmunol.1800443 100. Hazlitt RA, Min J, Zuo J (2018) Progress in the development of preventative drugs for cisplatin-induced hearing loss. J Med Chem 61:5512–5524. https://doi.org/10.1021/acs.jmedchem.7b01653 101. Ramkumar V, Mukherjea D, Dhukhwa A et al (2021) Oxidative stress and inflammation caused by cisplatin ototoxicity. Antioxidants (Basel) 10. https://doi.org/10.3390/antiox10121919 102. Breglio AM, Rusheen AE, Shide ED et al (2017) Cisplatin is retained in the cochlea indefinitely following chemotherapy. Nat Commun 8:1654. https://doi.org/10.1038/s41467-017-01837-1 103. Rousset F, Nacher-Soler G, Coelho M et al (2020) Redox activation of excitatory pathways in auditory neurons as mechanism of age-related hearing loss. Redox Biol 30:101434. https://doi.org/10. 1016/j.redox.2020.101434 104. Mukherjea D, Jajoo S, Kaur T et al (2010) Transtympanic administration of short interfering (si)RNA for the NOX3 isoform of NADPH oxidase protects against cisplatin-induced hearing loss in the rat. Antioxid Redox Signal 13:589–598. https://doi.org/10. 1089/ars.2010.3110 105. Mukherjea D, Jajoo S, Sheehan K et al (2011) NOX3 NADPH oxidase couples transient receptor potential vanilloid 1 to signal transducer and activator of transcription 1-mediated inflammation and hearing loss. Antioxid Redox Signal 14:999–1010. https://doi. org/10.1089/ars.2010.3497 106. Gates GA, Mills JH (2005) Presbycusis. Lancet 366:1111–1120. https://doi.org/10.1016/S0140-6736(05)67423-5 107. Ruan Q, Ma C, Zhang R et al (2014) Current status of auditory aging and anti-aging research. Geriatr Gerontol Int 14:40–53. https://doi.org/10.1111/ggi.12124 108. Zheng QY, Johnson KR, Erway LC (1999) Assessment of hearing in 80 inbred strains of mice by ABR threshold analyses. Hear Res 130:94–107. https://doi.org/10.1016/s0378-5955(99)00003-9 109. Johnson KR, Gagnon LH, Longo-Guess C et al (2012) Association of a citrate synthase missense mutation with age-related hearing loss in A/J mice. Neurobiol Aging 33:1720–1729. https://doi.org/ 10.1016/j.neurobiolaging.2011.05.009

11

Physiological Functions and Pathological Significance of NADPH Oxidase 3

110. Zheng QY, Ding D, Yu H et al (2009) A locus on distal chromosome 10 (ahl4) affecting age-related hearing loss in A/J mice. Neurobiol Aging 30:1693–1705. https://doi.org/10.1016/j. neurobiolaging.2007.12.011 111. Noben-Trauth K, Zheng QY, Johnson KR (2003) Association of cadherin 23 with polygenic inheritance and genetic modification of sensorineural hearing loss. Nat Genet 35:21–23. https://doi.org/10. 1038/ng1226 112. Stamataki S, Francis HW, Lehar M et al (2006) Synaptic alterations at inner hair cells precede spiral ganglion cell loss in aging C57BL/ 6J mice. Hear Res 221:104–118. https://doi.org/10.1016/j.heares. 2006.07.014 113. Lavinsky J, Crow AL, Pan C et al (2015) Genome-wide association study identifies nox3 as a critical gene for susceptibility to noiseinduced hearing loss. PLoS Genet 11:e1005094. https://doi.org/10. 1371/journal.pgen.1005094 114. Rousset F, Nacher-Soler G, Kokje VBC et al (2022) NADPH oxidase 3 deficiency protects from noise-induced sensorineural hearing loss. Front Cell Dev Biol 10:832314. https://doi.org/10. 3389/fcell.2022.832314 115. Cantu E, Shah RJ, Lin W et al (2015) Oxidant stress regulatory genetic variation in recipients and donors contributes to risk of primary graft dysfunction after lung transplantation. J Thorac Cardiovasc Surg 149:596–602. https://doi.org/10.1016/j.jtcvs. 2014.09.077 116. Carnesecchi S, Carpentier JL, Foti M et al (2006) Insulin-induced vascular endothelial growth factor expression is mediated by the NADPH oxidase NOX3. Exp Cell Res 312:3413–3424. https://doi. org/10.1016/j.yexcr.2006.07.003 117. Chen G, Adeyemo AA, Zhou J et al (2007) A genome-wide search for linkage to renal function phenotypes in West Africans with type 2 diabetes. Am J Kidney Dis 49:394–400. https://doi.org/10.1053/ j.ajkd.2006.12.011 118. Choi JH, Oh J, Lee MJ et al (2021) Inhibition of lysophosphatidic acid receptor 1-3 deteriorates experimental autoimmune encephalomyelitis by inducing oxidative stress. J Neuroinflammation 18: 240. https://doi.org/10.1186/s12974-021-02278-w 119. Gao D, Nong S, Huang X et al (2010) The effects of palmitate on hepatic insulin resistance are mediated by NADPH oxidase 3-derived reactive oxygen species through JNK and p38MAPK pathways. J Biol Chem 285:29965–29973. https://doi.org/10.1074/ jbc.M110.128694 120. Gupta AP, Syed AA, Garg R et al (2019) Pancreastatin inhibitor PSTi8 attenuates hyperinsulinemia induced obesity and inflammation mediated insulin resistance via MAPK/NOX3-JNK pathway. Eur J Pharmacol 864:172723. https://doi.org/10.1016/j.ejphar. 2019.172723 121. Issa N, Lachance G, Bellmann K et al (2018) Cytokines promote lipolysis in 3T3-L1 adipocytes through induction of NADPH oxidase 3 expression and superoxide production. J Lipid Res 59: 2321–2328. https://doi.org/10.1194/jlr.M086504

177

122. Malik SA, Acharya JD, Mehendale NK et al (2019) Pterostilbene reverses palmitic acid mediated insulin resistance in HepG2 cells by reducing oxidative stress and triglyceride accumulation. Free Radic Res 53:815–827. https://doi.org/10.1080/10715762.2019. 1635252 123. Malleter M, Tauzin S, Bessede A et al (2013) CD95L cell surface cleavage triggers a prometastatic signaling pathway in triplenegative breast cancer. Cancer Res 73:6711–6721. https://doi.org/ 10.1158/0008-5472.Can-13-1794 124. Nakayama N, Nakamura T, Okada H et al (2011) Modulators of induction of plasminogen activator inhibitor type-1 in HepG2 cells by transforming growth factor-β. Coron Artery Dis 22:468–478. https://doi.org/10.1097/MCA.0b013e32834a3817 125. Plantinga TS, Arts P, Knarren GH et al (2017) Rare NOX3 variants confer susceptibility to agranulocytosis during thyrostatic treatment of Graves' disease. Clin Pharmacol Ther 102:1017–1024. https://doi.org/10.1002/cpt.733 126. Radkowski P, Wątor G, Skupien J et al (2016) Analysis of gene expression to predict dynamics of future hypertension incidence in type 2 diabetic patients. BMC Proc 10:113–117. https://doi.org/10. 1186/s12919-016-0015-z 127. Yasuoka H, Garrett SM, Nguyen XX et al (2019) NADPH oxidasemediated induction of reactive oxygen species and extracellular matrix deposition by insulin-like growth factor binding protein-5. Am J Physiol Lung Cell Mol Physiol 316:L644–l655. https://doi. org/10.1152/ajplung.00106.2018 128. Yin C, Li K, Yu Y et al (2018) Genome-wide association study identifies loci and candidate genes for non-idiopathic pulmonary hypertension in eastern Chinese Han population. BMC Pulm Med 18:158. https://doi.org/10.1186/s12890-018-0719-0 129. Li L, He Q, Huang X et al (2010) NOX3-derived reactive oxygen species promote TNF-alpha-induced reductions in hepatocyte glycogen levels via a JNK pathway. FEBS Lett 584:995–1000. https:// doi.org/10.1016/j.febslet.2010.01.044 130. Tyler AD, Milgrom R, Stempak JM et al (2013) The NOD2insC polymorphism is associated with worse outcome following ileal pouch-anal anastomosis for ulcerative colitis. Gut 62:1433–1439. https://doi.org/10.1136/gutjnl-2011-301957 131. Zhang Y, Shan P, Srivastava A et al (2016) An endothelial Hsp70TLR4 axis limits Nox3 expression and protects against oxidant injury in lungs. Antioxid Redox Signal 24:991–1012. https://doi. org/10.1089/ars.2015.6505 132. Mazzonetto PC, Ariza CB, Ocanha SG et al (2019) Mutation in NADPH oxidase 3 (NOX3) impairs SHH signaling and increases cerebellar neural stem/progenitor cell proliferation. Biochim Biophys Acta Mol basis Dis 1865:1502–1515. https://doi.org/10. 1016/j.bbadis.2019.02.022 133. Augsburger F, Filippova A, Rasti D et al (2019) Pharmacological characterization of the seven human NOX isoforms and their inhibitors. Redox Biol 26:101272. https://doi.org/10.1016/j.redox. 2019.101272

Nox4: From Discovery to Pathophysiology

12

Louise Hecker, Kosuke Kato, and Kathy K. Griendling

Abstract

Keywords

Nox4, originally termed Renox, was discovered in 2000. Since that time, our understanding of its structure and function has expanded enormously. Nox4 is ubiquitously expressed and is regulated at multiple levels, including epigenetic modification, transcriptional regulation, posttranslational modification and interaction with partner proteins. It is unique among Nox enzymes in its lack of required cytosolic regulatory subunits and a propensity to produce hydrogen peroxide (H2O2) rather than superoxide. Nox4 has multiple subcellular localizations depending on cell type and environmental context. It regulates a number of fundamental cellular processes, including cytoskeletal structure and dynamics, proliferation, differentiation, survival, apoptosis, senescence, oxygen sensing, and metabolic homeostasis. Nox4 is required for normal organ functioning and certain physiological signaling processes, but it also contributes to multiple disease processes, especially fibrosis and diabetic vascular disease. It has been linked to cancer, myocardial hypertrophy, pulmonary disease, kidney disease and neurological conditions. Much remains to be learned about Nox4 in human disease and controversies remain about its contribution to physiology and pathophysiology. Attention has shifted to the development of specific Nox4 inhibitors, which will not only serve as a much-needed experimental tool, but also will help to realize the potential clinical utility of therapies targeting Nox4.

Nox4 · Fibrosis · Cytoskeleton · Proliferation · Differentiation · Pathophysiology

L. Hecker Department of Medicine, Emory University School of Medicine, Atlanta, GA, USA Atlanta VA Healthcare System, Atlanta, GA, USA e-mail: [email protected] K. Kato · K. K. Griendling (✉) Department of Medicine, Emory University School of Medicine, Atlanta, GA, USA e-mail: [email protected]; [email protected]

1

Introduction

Nox4, originally termed “Renox”, was first identified in kidney cortex [1, 2] and nearly simultaneously in osteoclasts [3] as a novel superoxide (O2•-)-generating oxidase. Since that time, much has been learned about its structure, binding partners, tissue and subcellular distribution, role in the biology of the cell and contributions to normal physiological and pathophysiological processes. This chapter describes advances in each of these areas to provide a rounded view of the current knowledge of Nox4.

2

Structure and Expression

A 66.5 kDa protein, Nox4 is comprised of 578 amino acids and shares 39% identity with Nox2; especially conserved in membrane-spanning regions and binding sites for heme, flavin adenine dinucleotide (FAD), and reduced nicotinamide adenine dinucleotide phosphate (NADPH) [2]. Similar to Nox1, 2, and 3, Nox4 has six membrane-spanning helices, two of which contain conserved histidine residues that are responsible for binding two hemes which sequentially reduce oxygen to O2•- [4]. These helices create three extracellular loops (termed A, C, and E) along with two intracellular loops (B and D) and an extended C-terminal tail (Fig. 12.1). Nox4 forms a heterodimeric complex with the membrane protein p22phox that stabilizes it and enhances its reactive oxygen species (ROS)-generating activity [5–7], but unlike Nox1 and Nox2, it does not require traditional cytosolic subunits or Rac for activity [6]. Genetic ablation of p22phox significantly diminishes Nox4-dependent ROS generation [8]. Although p22phox is a common binding partner for all

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_12

179

180

L. Hecker et al.

Fig. 12.1 Protein structure and exon structures of Nox4 splice variants. (a) The structural domains of Nox4 and p22phox are schematized. (b) Linear representation of Nox4 splice variants compared to Nox4

prototype. The numbers in the boxes represent the exon number. Exons corresponding to structural domains in translated proteins are shown at the bottom. Figure made using BioRender

Nox isoforms except Nox5, mutation analysis revealed a unique binding surface for Nox4 [9]. In common with other Nox enzymes, Nox4 contains predicted glycosylation sites and a dehydrogenase domain that requires FAD and NADPH [7, 10]. Either mutating Pro437 in the canonical NADPH binding motif [11, 12] or deleting the NADPH binding domain [13] inactivates the enzyme. Nox4 activity is also inhibited by an R96E mutation in the cytosolic B loop that interacts with the NADPH binding site [12, 14]. Surprisingly, Nox4 produces much higher levels of H2O2 than O2•- [6, 12, 15, 16], potentially due to its unusually large extracellular E-loop. Mutation of His222, Cys270 or Cys226 in the E-loop specifically inhibits H2O2 production, suggesting that this region of Nox4 is at least in part responsible for the ability of Nox4 to preferentially release H2O2 [17]. Nox4 is encoded by a gene that contains 18 exons and is transcribed into splice variants, four of which (Nox4B, Nox4C, Nox4D, and Nox4E) have been shown to generate protein in human lung cell lines and lung tissues from healthy donors [18] (Fig. 12.1). Nox4B lacks the first NADPH binding domain, and Nox4C lacks all NADPH and FADH binding sites. Overexpression of Nox4B or Nox4C led to decreased ROS levels, suggesting these variants have dominant negative characteristics for ROS generation [18]. Both Nox4D and Nox4E lack transmembrane domains, making these variants non-membrane associated isoforms. Nox4D is a small molecular weight variant (28 kD) that contains all NADPH and FAD-binding domains (required for electron transfer activity) but lacks all putative transmembrane domains. It is the only known functionally active Nox4 splice variant (capable of generating ROS) [18, 19], and has been shown to generate ROS at the same rate as Nox4 [18]. While Nox4 is significantly upregulated in ischemic heart failure, Nox4D is downregulated [20]; this inverse relationship may explain how Nox4 and its variants could play unique roles in

redox signaling within subcellular compartments. Nox4D has been shown to localize in the nucleus of various cell types [19, 21]. Nuclear Nox4D has been implicated in both ROS-induced DNA damage in vascular cells [19] and ROS-mediated survival in acute myeloid leukemia [21]. Further studies are needed to delineate both the physiological and/or pathological roles of these Nox4 variants.

2.1

Tissue Distribution

Nox4 is generally thought to be the most widely distributed Nox; it is ubiquitously expressed in various tissue and cell types [22]. For example, in the lung alone, Nox4 is expressed in multiple cell types, including macrophages [23–25], pulmonary artery endothelium [26], pulmonary artery smooth muscle cells [27], epithelial cells [28], and myofibroblasts [29, 30]. Nox4 has also been shown to be expressed in fetal tissues [31], embryonic stem cells [32], hematopoietic stem cells [33], and various other adult tissues/cells, including osteoclasts [3, 34], keratinocytes [35], melanoma cells [36], neurons [37, 38], adipocytes [39], chondrocytes [40, 41], hepatic stellate cells [42], epithelial cells [28], podocytes [43], endothelial cells [44–47], and vascular smooth muscle cells (VSMCs) [48–51].

2.2

Subcellular Distribution

The sub-cellular localization of Nox4 includes the nucleus, focal adhesions, endoplasmic reticulum (ER), and plasma membrane (Table 12.1). While multiple carefully performed studies describe a mitochondrial localization of Nox4, more recent work suggests that Nox4 actually resides in ER-mitochondria contact sites, or mitochondrial-associated membranes (MAMs) [65], which is consistent with the

12

Nox4: From Discovery to Pathophysiology

181

Table 12.1 Subcellular localization of Nox4 Sub-cellular localization of Nox4 Nucleus

Focal adhesions ER

Plasma membrane Mitochondria and mitochondrial-associated membranes

Cell type(s) • Monocytes • Macrophages • Endothelial • VSMCs • Smooth muscle cells • Renal cortical cells • VSMCs • Monocytes • HEK • COS cells • Endothelial cells • VSMCs • Airway smooth muscle cells • Renal cortical cells • Cardiomyocytes • Lung epithelial cells • COS cells • Mesangial cells • Renal cortical cells • Cardiac myocytes • Cancer cells

References [25, 45, 52–55]

[52, 56] [5, 6, 12, 16, 47, 54, 55, 57–60]

[9, 12] [61–64] [65]

transmembrane nature of the protein. The location of Nox4 contributes to its function; for example, during the early phase of energy deprivation, Nox4 activity and protein levels are increased in the ER (but not in mitochondria) of cardiomyocytes, a critical event that activates autophagy [57]. Mitochondrial Nox4-dependent ROS have been implicated in fibrosis [61], while cytoskeletal Nox4 contributes to focal adhesion turnover [66]. Nox4 splice variants have been reported to reside in the ER and nucleus [18]. Nox4 has also been shown to transition from one intracellular compartment to another [12, 67]; thus, Nox4 localization within compartments may also be transient. The relative abundance/expression of Nox4 variants within different compartments, in different cell types, and in response to various stimuli remains controversial due to technical limitations such as the ability to detect Nox4 vs. Nox4 variants, poor selective antibody availability, and/or staining approaches [68]. Overall, the precise sub-cellular localization and functional roles of Nox4 within localized compartments remain a topic for further study [69].

The biochemical and functional characteristics of Nox4 are distinct in comparison to other Nox isoforms. In particular, a unique feature of Nox4 activity is its capacity for generating H2O2 (vs. O2•-) [6, 9, 12, 16, 71]. Most Nox enzymes generate O2•-, but, as noted above, Nox4 catalyzes the reduction of molecular oxygen to H2O2, likely via a two-step process. In fact, although H2O2 is thought to be the major product of Nox4, Nox4 produces a small amount of O2•- which is difficult to detect due to its instability and intracellular compartmentalization [2, 72]. The consequences of Nox4-dependent H2O2 generation within specific cell types occur in a highly cell context-dependent fashion. High levels of Nox4-derived ROS cause cellular damage, senescence, and even cell death, while lower Nox4 H2O2 levels serve as second messengers to induce a panel of intracellular signaling pathways involved in cell proliferation, growth, and differentiation, among other physiological processes [73].

2.3

Nox4 has been proposed to be constitutively active; that is, its activity is regulated primarily by expression levels [16], although many studies have reported acute activation of Nox4 by external stimuli [13, 74–76] or chronic activation by expression of polymerase delta interacting protein-2(Poldip2) [67]. Nox4 expression and activity are increased by changes in oxygen tension [77–79] and high glucose [80] in endothelial cells in keeping with its proposed role in metabolic equilibrium (see below). Nox4 is also upregulated by shear stress [81, 82] and aldosterone [83] in this cell type.

H2O2 Production by Nox4

The only known function of Nox is the generation of ROS, including O2•- and H2O2. Other enzymes in the body are capable of producing ROS, including lipoxygenases, nitric oxide synthase, xanthine oxidase, cytochrome P450 oxidases, and mitochondrial electron transport chain [70]. However, a majority of these proteins produce ROS as a by-product of their catalytic activities, whereas Nox enzymes produce ROS as their primary and sole function.

3

Regulation of Expression and Activity

182

Hypertrophic agonists such as angiotensin II or phenylephrine upregulate Nox4 in cardiac myocytes [11], and transforming growth factor beta (TGF-β) is a potent inducer in heart [84], VSMCs [85], kidney [86, 87], lung [27, 88, 89], and hepatic stellate cells (HSCs) [90], among other tissues. Tumor necrosis factor alpha (TNF-α) upregulates Nox4 in VSMCs and human embryonic kidney (HEK293) cells [91], and lipopolysaccharide (LPS) induces Nox4 in the lung [44]. Negative regulators of Nox4 expression include cyclic strain [92], neuregulin [93], α-lipoic acid [94], and metformin [95]. Because in many cases Nox4 activity is tied to expression, considerable work has focused on the molecular mechanisms that control Nox4 expression. At the epigenetic level, pharmacological inhibition of bromodomain and extraterminal (BET) proteins, which serve as epigenetic readers that bind to acetylated lysine residues on histone and nonhistone proteins to regulate gene expression by recruiting transcriptional activators or repressors, reverses TGF-β-induced Nox4 expression in lung fibroblasts and myofibroblasts. Conversely, the BET proteins Brd3 and Brd4 bind to the Nox4 promoter and increase expression [96]. Brd4 appears to coordinate the association of the acetylated form of H4K16 and the acetyltransferase p300 with the Nox4 promoter [97]. Multiple transcription factors have been implicated in Nox4 gene regulation, including E2F [98], nuclear factor erythroid 2-related factor 2 (Nrf2) [78], myocardin-related transcription factor A (MRTF-A) [99], octamer-binding transcription factor 1 (Oct-1) [100], Kruppel Like Factor 5 (KLF5) [101], CCAAT-enhancer-binding proteins (C/EBP) [102], hypoxia inducible factor 1 alpha (HIF-1α) [103], Sp3 transcription factor (SP-3) [104], signal transducer and activator of transcription 1 and 3 (STAT1/3) [105], PU.1 [106], early growth response protein 1 (Egr1) [107], activating transcription factor 3 (ATF3) [108], GATA binding protein 4 (GATA4) [109], and nuclear factor kappa-lightchain-enhancer of activated B cells (NF-κB) [79], depending on the stimulus and the cell or tissue. While this list is daunting, transcriptional regulation of Nox4 is likely the major mechanism regulating Nox4 expression. In addition, there is good evidence for miRNA-mediated regulation of Nox4, although modulation by individual miRNAs seems to be tissue- and stimulus-specific (Table 12.2). Among the miRNAs that have been linked to Nox4, multiple reports support a role for the miR-92, miR-99, miR-100, and miR-146 families. In addition, it has been shown that human antigen R (HuR) is required for high glucose-induced Nox4 mRNA translation in mesangial cells by binding to AU-rich elements in the 3’untranslated region [139]. Nox4 also undergoes post-translational modification and interacts with other proteins that regulate its activity. Poldip2 can bind to p22phox and increase basal Nox4 activity in VSMCs without affecting mRNA levels [67]. Nox4 also

L. Hecker et al.

associates with protein disulfide isomerase (PDI) in VSMCs, enhancing its activity [49]. In the ER, Nox4 binding to calnexin is necessary for proper maturation, processing, and activity [140]. Finally, in LPS-stimulated HEK293 or endothelial cells, the COOH-terminal region of Nox4 can bind directly to the cytoplasmic tail of toll-like receptor 4 (TLR4), leading to ROS production and activation of inflammatory pathways [76, 141]. These studies highlight the distinct interaction of the key binding partners with Nox4, which can alter its ROS-generating capacity. Regulation of Nox4 at the protein level has been studied in multiple cell types. Nox4 ubiquitination and degradation have been reported in lung fibroblasts treated with TGF-β [142, 143]. Further work showed that treatment of human lung fibroblasts with the antibacterial macrolide azithromycin suppresses Nox4 by promoting proteasomal degradation [30]. Ubiquitination and degradation of Nox4 has also been reported in renal fibrosis [144] and in proximal tubular epithelial cells from spontaneously hypertensive rats (SHR) treated with the angiotensin II receptor type 1 (AT1) blocker losartan [145]. In a mouse model of acute lung injury, Nox4 ubiquitination was found to be deficient in senescent endothelial cells, resulting in sustained expression of Nox4 and enhanced H2O2 production [44]. Conversely, deubiquitination of Nox4 in HeLa cells and endothelial cells enhances its activity [146, 147], and in adventitial fibroblasts, Nox4 deubiquitination prolongs the upregulation of Nox4 by TGF-β [148]. A reduction of Nox4 ubiquitination in rat chondrocytes leads to elevated ROS production and contributes to the development of osteoarthritis [149]. Thus, ubiquitination/deubiquitination is a common mechanism of Nox4 regulation. There are very few reports of Nox4 phosphorylation, but in cardiomyocytes, Matsushima et al. [150] found that the proto-oncogene tyrosine-protein kinase (Src)-family member Fyn associates with the C-terminal domain of Nox4, directly phosphorylating it at Tyr-566 and reducing its activity. Conversely, phosphorylation of Nox4 on tyrosine 491 (Tyr-491) after insulin-like growth factor 1 (IGF-I) stimulation results in association with growth factor receptor-bound protein 2 (Grb2) and translocation to the plasma membrane scaffold Src homology 2 domain-containing protein tyrosine phosphatase substrate 1 (SHPS-1), leading to oxidation and activation of Src [151]. Further work is necessary to confirm the global importance of these two phosphorylation sites in acute regulation of Nox4 activity and its binding partners.

4

Physiological Functions

Nox4 has been implicated in numerous fundamental processes in multiple cell types (Fig. 12.2). While at times its many roles can seem directly opposed, the key to

12

Nox4: From Discovery to Pathophysiology

183

Table 12.2 Regulation of Nox4 expression by miRNAs miRNA miR-9-5p miR-17 miR-21a-3p miR-25

Regulation of Nox4 Downregulation Downregulation Downregulation Downregulation

Species Mouse Human Mouse Mouse Rat

miR30e miR-92b

Upregulation Downregulation Upregulation

Mouse Human

miR-92b-3p miR-99a

Downregulation Downregulation

Rat Mouse

miR-99a-5p miR99b-5p miR-100

Downregulation Downregulation Downregulation

miR-100-5p

Downregulation

Human Human Mouse Human Human

miR-137 miR-146 miR-146a

Downregulation Downregulation Downregulation

Human Mouse Human

miR-146a-5p miR-148b-3p miR-182 miR-182-5p miR-204-3p miR-322 miR-337-3p miR-363-3p miR-423-5p miR-590-3p

Downregulation Downregulation Downregulation Downregulation Downregulation Downregulation Downregulation Downregulation Downregulation Downregulation

Human Human Mouse Human Mouse Mouse Rat and human Mouse Mouse Human

Fig. 12.2 Role of Nox4 in fundamental cellular processes. Nox4 has been shown to regulate a number of cellular functions, including growth and differentiation, survival and senescence, oxygen sensing, mitochondrial function and metabolism, ER stress and autophagy, and cytoskeletal dynamics. Figure made using BioRender

Tissue Lung fibroblasts Microglial cells Endothelial tumor cells Kidney Schwann cells Heart Endothelial cells 293T cells Cardiomyocytes Endothelial cells Bone marrow mesenchymal cells Brown adipose tissue Lung adenocarcinoma TSSC1 cells Endothelial cells Cardiomyocytes Renal cell carcinoma Chondrocytes Endothelial cells PC-3 cells Cardiomyocytes HK-2 cells Aortic endothelial cells Mesenchymal progenitor cells Hepatic sinusoidal endothelial cells Sensory neurons Lens epithelial cells Brain Neural tube Tendon-derived stem cells Endothelial cells Podocytes Retinal microvascular endothelial cells

References [110] [111] [112] [113] [114] [115] [116] [117, 118] [119] [120] [121, 122] [123] [119] [118] [124] [125] [119] [126] [117] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138]

184

L. Hecker et al.

understanding the final biological response to Nox4 activation is to consider the subcellular location of activation (e.g., ER vs mitochondrial vs. cytoplasmic) as well as the degree of activation, since low Nox4 activity is involved in physiological signaling while higher levels can contribute to apoptosis, senescence, and ER stress. Because Nox4 regulates physiological processes such as oxygen sensing, cytoskeletal dynamics, and metabolism that are fundamental to many cellular responses, Nox4 has wide-ranging roles in cells and organisms.

4.1

Growth/Differentiation

Cell proliferation and differentiation are mutually exclusive events that are controlled by the fine-tuned balance of opposing cellular signaling. Numerous studies have documented the involvement of Nox4-dependent ROS in regulating both cell proliferation and differentiation. The role of Nox4 in proliferation has been reported in various cell types, including fibroblasts, endothelial cells, pulmonary VSMCs, hepatocytes, and cancer cells. Multiple types of cancer require Nox4 for progression through the cell cycle, including melanoma [152], glioblastoma [153], and non-small cell lung cancer [154]. In non-cancer cells, the role of Nox4 in endothelial cell proliferation is worth highlighting. The proliferation of endothelial cells is a crucial angiogenic process that is required for normal wound repair processes and other physiological as well as pathological processes, such as angiogenesis [155–157]. Nox4 is highly expressed in proliferating endothelial cells compared to quiescent cells [158] and mediates endothelial cell proliferation under basal conditions [159] and in response to various stimuli, including hypoxia [160], TGF-β1 [161], LPS [162], and vascular endothelial growth factor (VEGF) [163]. Generally, low levels of Nox4-derived H2O2 (in response to various stimuli) function as signaling molecules to mediate endothelial cell proliferation, which likely contributes to angiogenesis in vivo [164]. Nox4 has also been shown to play a role in pulmonary, as opposed to systemic, smooth muscle proliferation. In pulmonary VSMCs, TGF-β-induced Nox4 promotes proliferation under both normoxic [27] and hypoxic conditions [165]. In support of these findings, Green et al. [160] showed that pharmacological inhibition of Nox4 via GKT137831 (a putative Nox1/4 inhibitor) or genetic downregulation by siRNA attenuates hypoxia-induced ROS generation and proliferation in these cells. More usually, Nox4 has been shown to mediate differentiation of various cell types. Nox4-dependent ROS is required for maintenance of the differentiated systemic VSMC phenotype in the homeostatic state [72] and in the context of disease

(e.g., diabetic atherosclerosis) [166]. In fact, H2O2 derived from Nox4 is required for smooth muscle cell (SMC) differentiation from embryonic stem cells [167] and for differentiation of embryoid bodies into cells of all three germ layers [168]. Similarly, cardiomyocyte differentiation from pluripotent embryonal carcinoma cells is dependent upon redox activation of the proto-oncogene c-Jun by Nox4 [169]. TGF-β-induced upregulation of Nox4 mediates the expression of smooth muscle-specific pro-differentiation genes (e.g., smooth muscle α-actin) through the activation of p38 mitogen-activated protein kinase/serum response factor (MAPK/SRF) signaling [170]. Nox4 is also involved in TGF-β-induced myofibroblast differentiation via mothers against decapentaplegic homolog 3 (Smad3) signaling [88]. Moreover, siNox4 inhibits differentiation of pre-adipocytes to adipocytes by controlling the expression of MAP kinase phosphatase-1 (MKP-1), which inhibits extracellular receptor kinase 1/2 (ERK1/2) activation induced by insulin [171]. While the signaling pathways differ, a role for Nox4 in cell differentiation is clear.

4.2

Senescence

Senescence is broadly defined by irreversible cell-cycle arrest accompanied by profound phenotypic alterations, including a senescence-associated secretory phenotype (SASP)—a key component of which is elevated levels of extracellular ROS [172]. Nox4 has been implicated in mediating senescence of numerous different cell types, usually when highly expressed or overexpressed. It is upregulated in senescent endothelial cells [44, 173] and its overexpression leads to increased ROS production and induction of cellular senescence in these cells [174] and in fibroblasts [1]. Nox4 knockdown inhibits the induction of senescence (independent of telomere attrition) in endothelial cells [175, 176] and proximal tubular cells [177]. In human thyrocytes, oncogenic H-Ras upregulates Nox4 expression, which leads to DNA damage and senescence [178]. In oncogene Ras-induced fibroblast or embryonic senescence, upregulation of Nox4 plays a critical role [179, 180]. Mitochondrial dysfunction has also been implicated in Nox4-mediated senescence via targeting complex I of the electron transport chain [181]. In transgenic mice with high mitochondrial Nox4 expression (Nox4TG618; 4.9fold higher), VSMCs exhibit features of vascular aging including increased premature and replicative senescence, DNA damage, apoptosis, pro-inflammatory protein expression, and decreased respiration [182]. However, the role of Nox4 in senescence remains somewhat controversial, as some studies have also demonstrated that reduced expression of Nox4 leads to senescence. In human VSMCs,

12

Nox4: From Discovery to Pathophysiology

downregulation of Nox4 accelerates senescence, which correlates with increased secretion of pro-inflammatory cytokines and activation of HIF-1α [183]. Further, Nox4 knockout mice showed no difference in lifespan when compared to their wild type littermates [184]. Overall, however, the bulk of the evidence favors the interpretation that high levels of Nox4 activity induce senescence, in contrast to lower levels which mediate signaling leading to cell differentiation or proliferation as discussed above. Nox4-mediated senescence has been implicated in the pathogenesis of various diseases. For example, Nox4dependent senescence of endothelial cells impairs barrier function [44]. Although Nox4 is already upregulated in aged senescent endothelial cells, insults such as LPS can further induce Nox4 expression/activity in these cells, resulting in persistent elevation of Nox4 and excessive ROS generation [44]. In contrast, young endothelial cells exhibit only a transient increase in Nox4 expression, which rapidly returns to baseline, significantly lowering ROS levels [44]. Targeting sustained activation of Nox4 in senescent cells leads to significantly greater protection from ROS-mediated alterations in permeability [44], making Nox4 a viable therapeutic approach for age-associated acute respiratory distress syndrome (ARDS). As another example, senescence of nucleus pulposus (NP) cells plays a vital role in the pathogenesis of intervertebral disc degeneration. Downregulation of the epigenetic senescence regulator enhancer of zeste homolog 2 (EZH2) and upregulation of Nox4 and cyclin-dependent kinase inhibitor 2A (p16) were observed in degenerative discs of aging rats [185]. Overexpression of Nox4 inhibited the expression of EZH2, further inducing senescence of NP cells [185]. Another study demonstrated that Nox4 induces senescence in NP cells under oxidative stress through the MAPK and NF-κB pathways [186]. Nox4 expression is also elevated in senescent/idiopathic pulmonary fibrosis (IPF)-derived lung myofibroblasts [29, 187]. These senescent myofibroblasts exhibit an impaired capacity to induce Nrf2 antioxidant responses and apoptosis-resistance, and inhibition of Nox4-mediated ROS restores apoptosis susceptibility [29]. Genetic knockdown and pharmacological targeting of Nox4 in aged mice with established fibrosis leads to reduced myofibroblast senescence (presumably via increased apoptosis), reversal of age-associated persistent fibrosis, and improved survival [29]. Other studies have also demonstrated that pharmacological targeting of Nox4 leads to diminished ROS production and senescence markers in IPF fibroblasts, which interferes with critical events that regulate p53-dependent apoptosis and replicative senescence [188, 189]. These studies suggest that targeting Nox4 may have senolytic effects in lung fibrosis.

185

4.3

Apoptosis/Survival

A large number of studies have described both pro- and antiapoptotic responses resulting from Nox4 activation [73, 190]. These apparent contradictory functions may be due to the magnitude/duration of Nox4-dependent ROS signaling or differences in subcellular localization and cell type. It is known that ROS can trigger apoptosis either indirectly, through damage to DNA, lipids, and proteins, or more directly by ROS-mediated activation of signaling molecules. Nox4 has been shown to mediate apoptosis in human VSMCs [51], pancreatic cancer cells [191], leukemia cells [192], lung epithelial cells [28], astrocytoma cells [193], hepatocytes [90, 194, 195], cerebral vascular endothelial cells [196], liver cancer cells [197], human umbilical vein endothelial cells (HUVECs) [198], and cardiac myocytes [11], and is the primary source of TGF-β- and TNF–αinduced oxidative stress leading to apoptosis [194, 196, 197, 199– 201]. Studies suggest that Nox4-derived ROS in mitochondria could lead to mitochondrial dysfunction by opening the permeability transition pore, leading to cytochrome c release and apoptosome formation [61, 202, 203]. Conversely, Nox4 has also been shown to mediate apoptosis resistance. In a murine model of pulmonary fibrosis, Nox4 knockout mice or mice harboring a conditional deletion of Nox4 in monocyte-derived macrophages exhibited increased macrophage apoptosis and were protected from pulmonary fibrosis and resolution during established fibrosis [204]. Genetic and pharmacological targeting of Nox4 in aged mice with established fibrosis attenuates the antiapoptotic myofibroblast phenotype and leads to a reversal of age-associated persistent fibrosis [29]. A recent study suggests that senescence may promote profibrotic effects via impaired myofibroblast dedifferentiation, leading to apoptosis resistance that ultimately contributes to the accumulation of myofibroblasts in persistent fibrosis [205]. The capacity for myofibroblast dedifferentiation and subsequent apoptosis may be critical for normal physiologic responses to tissue injury, whereas restricted dedifferentiation and apoptosis resistance in senescent cells may underlie the progressive nature of age-associated human fibrotic disorders.

4.4

Cytoskeletal Rearrangement

Nox4 has also been implicated in cytoskeletal dynamics, mostly in VSMCs, thereby regulating cell migration. In VSMCs stimulated with TGF-β, Nox4 is required for focal adhesion formation, and impacts not only focal adhesion strength, but also platelet-derived growth factor (PDGF)induced migration [143]. Moreover, overexpression of Poldip2, a Nox4 regulator, attenuates focal adhesion turnover

186

L. Hecker et al.

by activating RhoA and focal adhesion kinase (FAK), an effect that is dependent upon Nox4. Conversely, knockdown of either Poldip2 or Nox4 results in a clear loss of focal adhesions [66, 67]. In addition to affecting Rho activity, Poldip2 and Nox4 are also responsible for actin oxidation during VSMC adhesion, which regulates the interaction of actin with vinculin, an important step in focal adhesion maturation [206]. Of interest, silencing PDI, another Nox4 regulator, has similar effects on the actin cytoskeleton and focal adhesions [207]. In contrast to systemic VSMCs, Nox4 activation in cerebrovascular VSMCs contributes to actin depolymerization after hypoxia, leading to impaired myogenic tone [208]. A link between Nox4 and the cytoskeleton and consequently cell migration has also been reported in other cell types. In monocytes, Nox4 localizes to focal adhesions and controls actin S-glutathionylation and turnover of actin, resulting in enhanced monocyte adhesion and chemotaxis when cells are under metabolic stress [56]. Nox4 knockout in HeLa cells markedly inhibits cell migration and cell invasion through Matrigel in response to epidermal growth factor (EGF). In these cells, the number and size of focal adhesions are reduced in the absence of Nox4 [209]. In human hepatoma (HepG2) and neuroblastoma (SH-SY5Y) cells derived from liver and brain tumors, respectively, H2O2 derived from Nox4 in the ER is necessary for maintenance of actin stress fibers, which are essential to cell migration and effectively disappear following Nox4 knockdown [210]. Similarly, in trabecular meshwork cells exposed to TGF-β2, incubation with GKT137831, a purported Nox1-Nox4 dual enzyme inhibitor (see below), inhibits actin stress fiber formation [211].

4.5

Oxygen Sensing

Almost since its original discovery, Nox4 has been proposed to be an oxygen sensor. Early on, Lee et al. [212] showed that the sensitivity of the K+ channel potassium two pore domain channel subfamily K member 3 (TASK-1) to hypoxia was abolished upon knockdown of Nox4. This group further defined the heme- and FAD-binding domains as important for oxygen sensing [213]. Sun et al. [214] found that Nox4 generates ROS in proportion to pO2 in skeletal muscle, which in turn regulates the redox state of the ryanodine receptorCa2+-release channels and the contractile response. Biochemical proof of an oxygen-sensing role for Nox4 came from the finding that Nox4 has an unusually high Km for oxygen (~18%), similar to the values of known oxygen-sensing enzymes [4], which allows it to generate ROS as a function of oxygen concentration and to respond rapidly to changes in pO2.

4.6

ER Stress and Autophagy

ER stress occurs when the ability of the ER to fold proteins becomes saturated, which has long been linked to oxidative stress, and more recently to Nox4. In an elegant study using HUVECs, Wu et al. [60] showed that siNox4 blocked ER H2O2 production in response to ER stressors and prevented the subsequent unfolded protein response. They found that Nox4 activates kirsten rat sarcoma virus (K-Ras) to induce autophagy and prevent cell death, consistent with a subsequent report implicating Nox4 in 7-ketocholesterolinduced autophagy in vascular smooth muscle [215]. Further work from this group showed that Nox4 oxidizes the sarco-/ ER Ca2+ ATPase (SERCA) to activate ER resident guanine nucleotide exchange factors [216]. Sciarretta et al. [217] provided further insight into the mechanism by which Nox4 promotes autophagy in cardiac myocytes subjected to glucose deprivation. They showed that Nox4 increases ROS in the ER to activate the protein kinase R-like ER kinase (PERK) signaling pathway, leading to increased autophagy and ultimately survival. Alternatively, in VSMCs from SHR, Nox4 regulates the inositol requiring enzyme 1 (IRE1)-XBox Binding Protein 1 (XBP1) pathway of the ER stress response [218]. It is noteworthy that in vessels from aged mice, Nox4-mediated IRE1α cysteine-based oxidation is thought to contribute to vascular dysfunction [174]. While transient activation of autophagy is protective, prolonged autophagy can lead to cell death, so here again, the magnitude of Nox4 activation is critical.

4.7

Mitochondrial Function and Metabolism

As noted above, multiple groups have reported Nox4 localization to mitochondria or mitochondrial-associated membranes (MAMs). In this location in cardiomyocytes, Nox4 inhibits stress-induced calcium transfer by enhancing Akt-dependent phosphorylation of the inositol trisphosphate receptor, thereby inhibiting calcium flux and mitochondrial permeability transition-dependent necrosis [65]. Others have shown that localization of Nox4 to mitochondria in cardiomyocytes results in increased O2•- production and oxidation of aconitase and NADH dehydrogenases, reducing their enzymatic activity [11]. Cardiomyocyte-specific deletion of Nox4 abrogates this mitochondrial dysfunction [61]. A more recent report clearly showed that Nox4 redirects glucose metabolism away from oxidation by activating the hexosamine biosynthetic pathway, which enhances fatty acid utilization [219]. This ultimately maintains cardiac energetics during stress. In other tissues, Nox4 has also been linked to mitochondrial function. Sun et al. [106] reported that Nox4 contributes to both oxidative stress and mitochondrial dysfunction in

12

Nox4: From Discovery to Pathophysiology

amyloid beta-induced degeneration of retinal pigment epithelium. Similarly, Block et al. [62] provided strong evidence for association of Nox4 with mitochondria in cultured mesangial cells and kidney cortex, where it contributes to glucose-induced mitochondrial O2•- generation. Nox4mediated generation of ROS in the mitochondria has also been implicated in tumorigenicity in breast cancer cells [63] as well as angiotensin II signaling in neurons [220]. An intriguing study by Shanmugasundaram et al. [64] demonstrated that Nox4-derived ROS inhibits acetylation and degradation of pyruvate kinase-M2 and may contribute to drug resistance in renal carcinoma cells. They showed that ATP can bind directly to Nox4 and negatively regulate its activity, providing a link between Nox4 and metabolism. Other events reported to be mediated by mitochondrial Nox4 include TGF-β-induced mitochondrial membrane depolarization in podocytes [221], cancer invasiveness [222], and age-associated aortic stiffness and mitochondrial respiration [182].

187

5

Role in (Patho)physiology

While the role of Nox4 in many aspects of normal physiology is clear (Fig. 12.3), an enormous amount of work has gone into understanding its role in the pathogenesis of disease. The literature is often confusing, with high quality manuscripts reporting contrasting results. Here, we consider studies that claim Nox4 is protective against disease, as well as those that point to a role of Nox4 in disease development as outlined in Fig. 12.4.

5.1

Myocardial Development and Hypertrophy

In normal physiology, Nox4 has been proposed to have a role in cardiac development. Treatment of mouse embryonic stem cells with siNox4 suppresses the appearance of spontaneously beating cardiac cells within embryoid bodies, an effect that can be reversed with low levels of exogenous H2O2

Fig. 12.3 Functions of Nox4 in different cell types. Nox4 has been implicated in multiple cellular responses that vary according to cell type. Common themes include growth, differentiation, apoptosis promotion or resistance, senescence and migration. Figure made using BioRender

188

L. Hecker et al.

Fig. 12.4 Pathophysiological roles of Nox4 in different tissues. While Nox4 has important physiological roles, it has also been implicated in multiple diseases of the brain, heart, kidney, lung and liver, as well as various cancers. Figure made using BioRender

[223]. Moreover, as noted above, Nox4 mediates differentiation of pluripotent embryonal carcinoma cells to cardiomyocytes [169]. In another study, Nadworny et al. [224] found that upregulation of Nox2 and Nox4 influences the differentiation of the receptor tyrosine kinase c-kit(+) cardiac progenitor cells into mature cardiomyocytes. Overexpression of Nox4 in postnatal cardiomyocytes, which normally do not divide, increases their ability to proliferate by upregulating cyclin D2, resulting in larger hearts [225]. In some of the earliest studies on the role of Nox enzymes in pathological cardiac hypertrophy, Ajay Shah’s group reported an increase in Nox4 expression and NADPH oxidase activity in the left ventricle of animals exposed to aortic banding to induce cardiac hypertrophy, in contrast to hypertrophy induced by angiotensin II infusion, which was largely dependent upon Nox2 [226]. A major characteristic of cardiac hypertrophy and heart failure is fibrosis, due in part to conversion of fibroblasts to myofibroblasts. Cucoranu et al. [84] showed that silencing Nox4 prevented this conversion, suggesting that Nox4 is critically involved in heart failure. The creation of mice with cardiac-specific overexpression of Nox4, compared with mice expressing a catalytically inactive Nox4 in the heart, revealed a role for Nox4 in agingassociated fibrosis and apoptosis along with cardiomyocyte hypertrophy and mitochondrial dysfunction [11]. In agreement with this study, in whole body- or cardiac specific Nox4 knockout mice exposed to pressure overload, cardiac

hypertrophy, fibrosis, apoptosis, and mitochondrial dysfunction were attenuated [61, 227]. These findings rapidly became controversial when another group published contrary results using a different Nox4-/- mouse, a controversy that remains unexplained. In this second study, using perhaps a less severe pressure overload model, loss of Nox4 actually exaggerated contractile dysfunction, hypertrophy, and cardiac dilatation, in part due to loss of Nox4-dependent preservation of myocardial capillary density after the intervention [157]. Similar results were found when Nox4 was deleted only in endothelial cells [228]. Subsequent work showed that Nox4 activates Nrf2-dependent expression of antioxidant genes, upregulates glutathione biosynthetic and recycling enzymes, and limits mitochondrial damage [229, 230]. Nox4 also regulates cardiac fatty acid oxidation, preferentially oxidizing fatty acids for energy, which improves myocardial energetics under stresses such as pressure-overload hypertrophy [231]. These latter findings uncover an important role for Nox4 in coupling redox state to metabolism. Comparable to the study by Kuroda et al. [61] using pressure overload, Nox4-/- mice exposed to volume overload due to creation of an aortocaval fistula showed less left ventricular hypertrophy and less eccentric left ventricular remodeling than wild type mice, due in part to impaired activation of Akt [232]. Conversely, endothelial-specific Nox4 transgenic mice had enlarged hearts. After angiotensin II infusion, Nox4 transgenic mice had significantly less

12

Nox4: From Discovery to Pathophysiology

fibrosis and inflammatory cell infiltration, potentially due to reduced endothelial activation, but similar cardiac hypertrophy and contractile function compared to wild type mice [233]. Taken together, these studies highlight a complex role of Nox4 in cardiac hypertrophy that may depend on the age of the mice, the cell types affected, the degree of injury, and the genetic makeup of the modified mouse itself.

5.2

Myocardial Ischemia

A few investigators have addressed a potential role for Nox4 in myocardial ischemia. Yu et al. [234] showed that cardiacspecific overexpression of Nox4 increases the NADP+/ NADPH ratio and decreases the glutathione/glutathione disulfide ratio, leading to impaired cardiac energetics and contractile function after ischemia-reperfusion. In contrast, another group found that mice overexpressing Nox4 in cardiomyocytes had higher survival, better contractile function and decreased cardiac remodeling compared with wild type mice [235]. The discrepancy between these two studies may be related to the level of overexpression of Nox4. In Nox4-/- mice, the infarct area resulting from permanent left anterior descending coronary artery ligation was smaller than in wild type mice, capillary density was higher, and macrophage infiltration was lower, with less cell death [236]. These data imply that Nox4 plays a role in ischemic heart failure, in agreement with Yu et al.

5.3

Vascular Disease

Nox4 is highly expressed in all layers of the vessel wall, including endothelium, smooth muscle, and adventitial fibroblasts [159, 237, 238]. Of note, a functionally active 28 kDa splice variant of Nox4 was first reported in the nucleus of VSMCs [19]. As mentioned above, in endothelial cells Nox4 mediates proliferation [159], while in VSMCs, Nox4 is required for maintenance of the differentiated phenotype [72]. In human atherosclerotic arteries, Nox4 correlates with α-smooth muscle actin content [237] and is localized to SMCs that maintain a contractile phenotype [239], and in cultured VSMCs, it is responsible for basal levels of H2O2 production [15]. There are also some reports of Nox4 expression in macrophages, where it localizes to focal adhesions and regulates migration, an important step in the formation of atherosclerotic plaques [56]. The role of Nox4 in vascular lesion development, such as occurs in atherosclerosis or narrowing of arteries following angioplasty, has been studied in several mouse models with conflicting results. Szocs et al. [240] reported that Nox4 is upregulated during the late phase of lesion development following balloon injury of the carotid artery when

189

neointimal VSMCs are beginning to re-differentiate. Xu et al. [241] confirmed these results, but instead showed a correlation of Nox4 expression with VSMC senescence and apoptosis. Schurmann et al. [242] found that deletion of Nox4 in apolipoprotein E (ApoE)-/- mice, which have increased circulating cholesterol and an increased susceptibility to vascular lesions, resulted in increased atherosclerotic plaque formation after high fat diet, suggesting that Nox4 normally protects against lesion formation. Similarly, in Nox4-/- mice crossed with LDL receptor knockout mice (a comparable model to ApoE-/- mice) and fed a high fat diet, atherosclerotic burden was higher and endothelial function was impaired [243]. Consistent with a protective effect of Nox4, transgenic overexpression of Nox4 in the endothelium reduced lesion development in ApoE-/- mice fed a high fat diet, in part by promoting a T-cell distribution that favors repair over inflammation [244]. However, not all studies support a protective role for Nox4 in atherosclerosis. Mice transgenic for a human Nox4 P437H dominant negative mutation expressed in smooth muscle have reduced inflammation and lesion formation [245, 246]. In agreement with this study, mice with double knockout of ApoE and Nox4 showed a mitigation of plaque development in response to bacterial activation of the TLR5 pathway [247]. Some of the discrepancies between these studies might be attributable to the genetic makeup of the mice, as mentioned above for myocardial hypertrophy, but also may be related to the age of the mice at the time endpoints were analyzed. Runge’s group showed that aged mice have higher expression of Nox4 in the vasculature and that Nox4 expression correlates with age and atherosclerotic severity [248]. Combining a diabetic model (streptozotocin injection) with Nox4 deletion on an ApoE-/- background uncovered a worsening of aortic pathology, with increased plaques and an increase in pro-inflammatory genes at 20 weeks [249]. These mice also exhibited increased growth factors and matrix proteins along with a de-differentiated vascular smooth muscle phenotype [166]. In contrast, the same group showed that at 10 weeks, deletion of Nox4 attenuated plaque formation and reduced activated T-cells [250]. Treatment with GKT137831, a putative Nox1/4 inhibitor, beginning at 10 weeks suppressed proinflammatory and profibrotic events [251]. Other studies in diabetic animals also showed conflicting results. In the obese Zucker rat, downregulation of Nox4 with short hairpin RNA reduced injury-induced neointimal formation [252]. However, endothelial-specific expression of dominant negative P437H Nox4 to reduce Nox4 activity in the endothelium worsened type I diabetes-induced atherosclerotic lesion formation [253] and exacerbated lesions in ApoE-/- mice [254], in both cases by upregulating soluble epoxide hydrolase. Because the efficacy of each of these approaches is likely less than that of full genetic knockout,

190

L. Hecker et al.

these studies suggest that basal Nox4 activity, as well as potentially different roles of Nox4 in different cell types, may also influence outcomes. In addition to atherosclerosis, recent work has implicated Nox4 in aneurysm development and aortic root dilation, which are characterized by weakening of the blood vessel wall. For example, in a mouse model of Marfan’s syndrome, a genetic condition predisposed to aneurysms that is caused by a mutation in the matrix protein fibrillin 1, Nox4 deletion reduced fragmentation of elastin fibers and ameliorated aortic root dilation [255]. Similarly, in mice expressing a dominant negative Nox4 in smooth muscle, aortic collateral aneurysms were reduced [245]. Moreover, in hyperphenylalaninemia (hph)-1 mice infused with angiotensin II that are genetically deficient in Nox4, the incidence of abdominal aortic aneurysms was dramatically reduced from 76.5% to zero [256]. It is notable that two human Nox4 mutations were identified in aneurysm patients, N129S and T555S, both of which produce more H2O2 than wild type Nox4 [256]. Overall, there are many conflicting reports on the role of Nox4 in vascular disease. A slight majority of studies show that Nox4 is protective, but equally compelling work demonstrates that under some circumstances, Nox4 contributes to disease development, more so in aged animals. Definitive studies in humans remain elusive.

5.4

Hypertension

Because some of the earliest studies of novel Nox2 homologues involved angiotensin II stimulation of NADPH oxidase activity [257], and angiotensin II is intimately involved in hypertension, there has been much interest in a role for the various homologues in the development of hypertension. With respect to Nox4, early work showed that Nox4 expression is higher in aorta of SHR compared with Wistar Kyoto rats [258]. Further work demonstrated that Nox4 in brain, blood vessels and kidney not only has distinct mechanisms of regulation, but also makes distinct contributions to the hypertensive phenotype. Peterson et al. [259] showed that both Nox2 and Nox4 in the subfornical organ of the brain are required for vasopressor effects of angiotensin II. Moreover, aldosterone-induced hypertension is mitigated by injection of siNox4 adenovirus into the paraventricular nucleus of mice [260]. With the development of an inducible Nox4 knockout mouse, the role of Nox4 in angiotensin II-induced hypertension was studied in detail. Blood pressure in these animals was similar to controls and their aortas showed enhancement of the prototypical inflammation, medial hypertrophy, and endothelial dysfunction that results from angiotensin II infusion, indicating that Nox4 is normally protective against the hypertensive phenotype [156]. Conversely, Bouabout et al. [261] reported that

Nox4-/- mice infused with a higher dose of angiotensin II showed reduced arterial and pulse pressure increases. Mice with endothelium-specific overexpression of Nox4 exhibit enhanced vasodilatory responses to acetylcholine or histamine as a result of increased H2O2 production, leading to lower systemic blood pressure [262]. Of importance, Nox4derived H2O2 appears to be an important endotheliumdependent vasodilator in coronary arteries as well [263]. With regard to the kidney, in hypertensive mice with 5/6 nephrectomy, deletion of Nox4 had no effect on the development of hypertension or albuminuria [264]. In contrast, Cowley et al. [265] deleted Nox4 in Dahl salt-sensitive (SS) rats and reported a significant reduction of salt-induced hypertension compared with wild type SS rats and later showed that this reduction is in part due to inhibition of Nox4-mediated attenuation of sodium reabsorption in the distal nephron [266, 267]. In these Nox4-knockout Dahl SS rats, mitigation of renal injury and blood pressure elevation is associated with normalization of renal mitochondrial bioenergetics and decreased mitochondrial ROS production [265]. Thus, there seems to be a clear relationship between Nox4 and salt, a concept supported by a study in mice showing that in Nox4-/- mice, a low sodium diet, which activates the renin-angiotensin system as well as sympathetic tone to increase blood pressure in wild type mice, was less effective at increasing blood pressure [268]. More recently, in humans, a large genome-wide association study identified Nox4 as a blood pressure-associated gene [269]. Clearly, a more thorough understanding of how Nox4 contributes to hypertension is warranted.

5.5

Acute Lung Injury (ALI) and Acute Respiratory Distress Syndrome (ARDS)

The critical pathological features of ARDS (loss of lung endothelial barrier integrity and inflammatory injury) are strongly associated with oxidative stress. Numerous studies have demonstrated that excessive ROS production plays a major role in endothelial cell barrier dysfunction, which leads to increased vascular permeability and pulmonary edema. Nox4 is the most abundantly expressed Nox isoform and the primary enzymatic source of ROS in endothelial cells [47, 53, 173, 270–272]. A number of stimuli (e.g., LPS and P. aeruginosa, the causative agent of pneumonia) have been shown to induce Nox4-dependent ROS, which in turn mediate increased permeability of various types of endothelial cells [76, 273–277]. In vivo, TGF-β can activate Nox4, which in turn exacerbates pulmonary edema in ARDS [278]. Lung endothelial cells challenged with LPS increase Nox4-dependent ROS production and permeability [44], and data from various groups support a Nox4-specific role in mediating endothelial cell permeability (ruling out the role

12

Nox4: From Discovery to Pathophysiology

of other Nox isoforms) [273, 275]. In senescent endothelial cells, defective ubiquitination leads to excessive Nox4 levels, and pharmacologic inhibition of Nox4/ROS restores barrierregulatory responses in these cells [44]. In a two-hit pre-clinical animal model of ARDS, persistent Nox4 expression in the lung and elevated ROS levels in bronchioalveolar lavage (BAL) fluid are associated with increased severity of ALI in aged mice [44]. In contrast, young mice with mild injury exhibit only a modest increase in Nox4 and significantly lower BAL ROS levels than aged injured mice [44]. Collectively, these studies suggest an age-dependent divergent induction of Nox4 in the lungs following ALI, and that targeting Nox4 in age-dependent severe ARDS is likely to offer greater therapeutic protection than targeting Nox4 in young animals. Given the abundant studies supporting a role for Nox4 in mediating endothelial barrier function, it is not surprising that several groups have investigated genetic/pharmacologic strategies to target Nox4 in ALI in vivo. Previous work has demonstrated that targeting Nox4 using non-selective Nox inhibitors and/or inhibitors that indirectly target Nox4 leads to protection from ALI using various animal models [275, 279–283]. Genetic studies also support a Nox4-specific role in ALI. Nox4 knockdown (via an siRNA approach) attenuates P. aeruginosa-induced ALI in vivo [276] and protects against ALI in a cecal ligation puncture model [273]. In addition, global Nox4-/- mice are protected from bleomycin-induced ALI [278]. Chronic alcohol abuse, a comorbidity of ARDS that increases oxidative stress, impairs alveolar macrophage function, and ultimately increases the incidence and severity of ARDS, does so through elevated Nox4 [284]. Finally, a recent study by Hong et al. [285] evaluated Nox4 levels (presumably in extracellular vesicles) in blood plasma from mechanically ventilated ARDS patients, and reported that elevated Nox4 levels in plasma were associated with extubation failure and increased mortality. These data suggest that Nox4 may also be a predictive biomarker of ARDS patient outcomes.

5.6

Pulmonary Hypertension (PH)

PH is a progressive disease that is defined by abnormally high pressure in the blood vessels between the lungs and the heart. PH is a common complication of IPF and is strongly linked to mortality [286]. Vascular remodeling is an important pathological feature of PH, which is characterized by the increased proliferation of VSMCs and/or endothelial cells. ROS are important regulators of pulmonary vascular remodeling, and accumulating evidence supports a prominent role for Nox4 in the pathogenesis of PH. Nox4 is the major Nox isoform expressed in human pulmonary artery SMCs [27] and is significantly increased in the vessel media of pulmonary

191

arteries and in pulmonary artery VSMCs isolated from patients with idiopathic PH [77, 287] compared to corresponding healthy control tissues/cells. Nox4 expression is also increased in thickened pulmonary arteries of IPF patients compared to healthy subjects [288]. These findings suggest a role for Nox4 in vascular remodeling associated with the development of PH. Persistent hypoxia can trigger remodeling of the pulmonary arteries with SMC hyperproliferation, which can ultimately lead to the development of PH. Nox4 expression is elevated in hypoxic environments in pulmonary artery SMCs compared to normoxic environments [77, 103, 165]. Further, Nox4 plays a key role in promoting the proliferation of pulmonary artery SMCs [27, 165, 289] and pulmonary artery adventitial fibroblasts [290] under hypoxic environments. In mouse lungs, Nox4 is highly expressed in the media of small pulmonary arteries, and expression is further increased by hypoxia [77]. In animal models of PH, numerous studies have also demonstrated that pulmonary Nox4 expression is upregulated by chronic hypoxia [77, 291–295]. Pharmacological inhibition of Nox4 (via GKT137831) [160] or suppression of Nox4 upregulation (via a peroxisome proliferatoractivated receptor gamma (PPARγ) ligand, rosiglitazone) [296] during chronic hypoxia attenuates the development of PH in vivo. However, some studies have reported seemingly contradictory results. One study reported that genetic ablation of Nox4 in mice had no effect on vascular remodeling [287]. Another study reported that deletion of Nox4 had no effect on chronic hypoxia-induced pulmonary artery pressure or vascular remodeling [297]. Note that both of these latter studies were performed in the same Nox4 knockout line that showed a protective effect of Nox4 in vascular and cardiac disease and should be confirmed with other methods of Nox4 ablation.

5.7

Chronic Obstructive Pulmonary Disease (COPD)

COPD is characterized by remodeling of small airways and destruction of the lung parenchyma. Numerous studies have implicated persistent oxidative stress in the lung following long-term exposure to cigarette smoke as a key etiologic factor contributing to the development and progression of this disease [298]. The lungs of patients with COPD exhibit elevated markers of oxidative stress including nitrotyrosine [299] and lipid peroxidation products such as 8-isoprostane [300], 4-hydroxy-2-nonenal [301], and malondialdehyde (MDA) [302]. Dysregulation of Nox4 expression has been implicated in the pathogenesis of COPD. Immunohistochemical assessment of the lungs of COPD patients demonstrated increased Nox4 expression compared to the non-COPD cohort [303–

192

L. Hecker et al.

306]. In a murine model of cigarette smoke-induced emphysema, Nox4 expression is upregulated following exposure to cigarette smoke in the lung compared to controls [305], and Nox4 appears to mediate oxidant damage in primary human bronchial epithelial cells [307] and mouse lung fibroblasts [308] exposed to cigarette smoke. Nox4 expression is associated with smooth muscle hypertrophy of small airways, a hallmark of remodeling [304, 305], which in turn correlates with impaired pulmonary function and disease severity [304]. In airway smooth muscle cells isolated from COPD patients, pharmacological inhibition of Nox4 via GKT137831 significantly reduces ROS generation [303]. These studies indicate that dysregulated Nox4 expression and resultant redox imbalance could be a key mechanism contributing to the remodeling of smooth muscle in the small airways in the context of COPD.

5.8

Fibrotic Disease

Fibrotic disorders represent a major health problem, affecting various organ systems including the liver, skin, kidney, heart, and lung. It has been suggested that core pathways that mediate fibrosis in multiple organ systems may serve as promising targets for anti-fibrotic drug development [309]; redox imbalance in the context of aging has been suggested to represent one of these core pathways [310]. Overproduction of ROS by Nox4, particularly late in life, results in oxidative stress that damages tissues over time and may contribute to the development of fibrotic diseases. The role of Nox4 in cardiac and kidney fibrosis are discussed separately; here we will focus on Nox4 in lung and liver fibrosis.

5.8.1 Lung Fibrosis Myofibroblasts are the key effector cell type and a central mediator of fibrosis in diverse fibrotic disorders. TGF-β1, a cytokine known to be overexpressed in fibrotic disease [311], leads to the induction of Nox4-dependent ROS production, which promotes fibroblast migration [312] and mediates the transition to pro-fibrotic myofibroblast phenotypes, including differentiation, contraction, and extracellular matrix deposition [88]. A seminal study by Hecker and colleagues was the first to demonstrate a critical role for the Nox4 isoform in mediating tissue fibrosis [88]. In two murine models of pulmonary fibrosis, suppression of Nox4 by siRNA or pharmacologic targeting of endogenous Nox by diphenylene iodonium (DPI) (a well-validated FAD-containing enzyme inhibitor) attenuated the development of lung fibrosis [88]. Subsequent studies have further validated these findings using Nox4 knockout mice [28, 313] and/or pharmacological targeting of Nox4 [187]. Another study demonstrated that treatment with metformin, an anti-diabetic drug, attenuates TGF-β1–induced Nox4 expression as well as subsequent

ROS generation and myofibroblast differentiation in lung fibroblasts in vitro [314]. Further, intraperitoneal administration of metformin in mice was shown to attenuate bleomycininduced lung fibrosis [314]. Studies also suggest a role for Nox4 in mediating age-dependent persistent lung fibrosis. In an aging animal model of persistent fibrosis, therapeutic targeting of Nox4 led to a reversal of age-associated established fibrosis [29]. In vivo knockdown of Nox4 in aged mice led to decreased oxidative stress, diminished myofibroblast senescence, and increased susceptibility to apoptosis [29]. As noted above, Nox4 expression is controlled by the epigenetic protein BET, and aged mice with established and persistent lung fibrosis recover capacity for fibrosis resolution with OTX015 (BET inhibitor) treatment [97]. Similar to bleomycin-induced lung fibrosis, intraperitoneal administration of metformin in mice reverses age-associated established lung fibrosis as well [315]. Nox4 is upregulated in myofibroblasts from the lungs of IPF patients [88, 312], which is associated with elevated ROS levels [29, 88]. Increased Nox4 expression has also been reported in atypical hyperplastic surfactant-secreting AT-2 cells in the IPF lung [312]. In mice, the genetic ablation of Nox4 protected against bleomycin-induced pulmonary fibrosis with a concomitant reduction of alveolar epithelial cell apoptosis compared to wild type littermates [28]. Consistent with these studies, genetic ablation or pharmacological targeting of Nox4 with GKT136901 in murine primary alveolar epithelial cells significantly reduces TGF-β1-mediated ROS generation and mitigates apoptosis [28]. Another study demonstrated that Nox4 mediates the profibrotic polarization of lung macrophages, where deletion of Nox4 in macrophages led to decreased extracellular matrix deposition and protection from asbestos-induced pulmonary fibrosis [23]. Together, these studies suggest that Nox4 is expressed and activated in different cell types in the lung and contributes in multiple ways to fibrogenic responses.

5.8.2 Hepatic Fibrosis Upon liver injury, quiescent HSCs become activated, and there is increasing evidence that Nox-derived ROS play a key role in the activation of HSCs and liver fibrosis. Treatment of HSCs with TGF-β increases Nox4 expression associated with transdifferentiation to myofibroblasts [90, 316, 317]. Genetic knockdown or pharmacological targeting of Nox4 attenuates ROS production and the expression of fibrogenic markers in HSCs [90, 316, 318, 319], and also decreases hepatocyte apoptosis [316]. These studies suggest that Nox4 downstream of TGF-β is necessary for HSC activation, which plays a key role in hepatic myofibroblast activation/maintenance. In vivo, Nox4 has been shown to be upregulated in several animal models of liver fibrosis [90]. Nox4-/- mice exhibit

12

Nox4: From Discovery to Pathophysiology

decreased oxidative stress, inflammation, injury, and fibrosis in the liver after carbon tetrachloride (CCl4) treatment as compared to wild type mice [319]. Targeting Nox4 using GKT137831 attenuates CCl4 or bile duct ligation-induced ROS production and hepatic fibrosis in mice [316, 318, 319]. Importantly, Nox4 protein expression levels are significantly elevated in livers of non-alcoholic steatohepatitis (NASH) patients [320], patients with hepatitis C viral infection [90], and patients with cirrhosis [319] as compared with normal controls.

5.9

Kidney Disease

The levels and functions of Nox4 have been widely studied in various types of kidney diseases including nephropathies such as diabetic nephropathy and hypertensive nephropathy, acute kidney injury, and renal carcinoma. Although Nox4deficient mice do not show any detectable abnormality in their kidneys under homeostatic conditions [38, 157], the role of Nox4 in kidney disease remains complex and controversial. This is evidenced by the fact that studies have reported Nox4 to play protective [264, 321–324] and detrimental [64, 221, 265, 325–338] roles in kidney disease. Strong evidence from animal models indicates a role for Nox4 in development of diabetic nephropathy. Early studies showed an increase in Nox4 in the kidneys of rats in which type 1 diabetes was induced by streptozotocin injection that was proposed to be responsible for diabetic nephropathy [339]. Subsequent work demonstrated that Nox4 antisense attenuates renal hypertrophy and fibronectin expression in this model [326]. Treatment of SHR with streptozotocin similarly increases Nox4 gene expression in the kidney [340], while treatment of Dahl SS rats null for Nox4 with streptozotocin leads to improved albuminuria and less glomerular injury compared to Dahl SS rats replete with Nox4 [341]. In ApoE-/- mice injected with streptozotocin, deletion of Nox4 attenuates albuminuria, reduces glomerular accumulation of extracellular matrix proteins, and inhibits glomerular macrophage infiltration, results that were replicated using a putative Nox1/4 inhibitor, GKT137831 [330]. Similar results were found in Nox4-/- mice [328], in OVE26 mice (a chronic model of type 1 diabetes) treated with GKT13783 [342], in podocyte-specific Nox4-/- mice treated with streptozotocin [329], and in Akita mice exposed to GKT137831 [343]. This latter study linked Nox4 to fumarate hydratase and fumarate levels, which in turn mediates several pathways leading to glomerular dysfunction. Another putative Nox4 inhibitor, APX-115, demonstrated pre-clinical efficacy in animal models of diabetic nephropathy [344, 345]. In contrast to these observations, other studies suggest that Nox4 is protective in the context of renal injury. In a model of chronic renal injury, Nox4-deficient mice developed more

193

severe interstitial fibrosis and tubular apoptosis after obstruction when compared with wild type mice [324]. A study by Babelova et al. [264] evaluated the role of Nox4 in chronic kidney disease using three different injury models (streptozotocin diabetes, unilateral ureteral ligation, and 5/6 nephrectomy) in either Nox4-inducible or constitutive knockout mice. Although diabetes, hydronephrosis, and chronic renal failure are all models of increased renal deposition of matrix proteins such as fibronectin, genetic deletion of Nox4 did not slow the fibrotic process in any of these disease models. These authors concluded that Nox4 does not promote renal disease development but may rather have a small, limited protective effect. In acute kidney injury induced by ischemia-reperfusion, Nox4-deficient animals displayed increased apoptosis and lower renal function than wild type animals [322]. The absence of Nox4 may aggravate the kidney phenotype in conditions of acute and chronic tubular injury by downregulating major cytoprotective pathways active at baseline, rendering tubular cells more sensitive to injury [346]. There are a number of possible explanations for these conflicting reports. First, Nox4 has been shown to play different roles in tubular, interstitial cells, podocytes, and glomerular cells [347]. Diabetes induces systemic hyperglycemia, which affects organs differently; this may not be comparable to in vitro experimental conditions where an agonist is added to one renal cell type (mesangial, tubular epithelial, or podocyte cell) with subsequent measurement of a biological output [347]. Additionally, the severity of the renal injury seen in diabetic nephropathy differs considerably among different animal models. Thus, it is conceivable that the role of Nox4 varies during the stages and types of kidney disease, such as early vs. advanced diabetes as well as the severity of the disease at these different phases of diabetic complications. While the reason for these discrepancies remains unclear, the preponderance of evidence suggests that Nox4 plays a pathological role in diabetic kidney disease.

5.10 Diabetes/Insulin signaling Nox4 is involved in other consequences of type 1 diabetes as well. Nox4 antisense oligonucleotides inhibit molecular markers of cardiac hypertrophy and myofibrosis induced by streptozotocin injection of rats [348]. Adipocyte-specific deletion of Nox4 in mice fed a high fat, high sucrose diet results in reduced adipose tissue inflammation and a delayed onset of insulin resistance [349]. In another study, deletion of Nox4 in animals fed a high fat diet had no effect on the metabolic profile, but increased glucose intolerance in female but not male mice [261].

194

L. Hecker et al.

Nox4 has also been implicated in type 2 diabetes. Nox4 expression is increased in aortas from type 2 diabetic db/db mice [350], as well as in kidney cortex where it mediates fibrotic processes [80]. Of interest, in hepatocytes, ablation of Nox4 recapitulates insulin resistance seen in livers of db/db mice [351]. In vivo, treatment of db/db mice with GKT136901 has no effect on plasma glucose, but protects the kidney from albuminuria and excess matrix production, potentially through reduced oxidative damage [352]. Nox4 is also involved in diabetic retinopathy in this model. Nox4 expression is increased in the retina and siNox4 reduces retinal permeability in db/db mice [353]. In support of these pre-clinical observations, a multi-cohort study in humans uncovered an association of several single nucleotide polymorphisms of Nox4 with severe diabetic retinopathy [354]. The idea that Nox4 may be involved in insulin signaling came from a study by Mahadev et al. [13] showing that expression of dominant negative Nox4 or knockdown of Nox4 using siRNA inhibited insulin-stimulated H2O2 generation, insulin receptor tyrosine phosphorylation, activation of downstream serine kinases, and glucose uptake. Subsequent work showed a similar role for Nox4 in IGF-1 signaling [355]. In cardiomyocytes stimulated with insulin, Nox4derived H2O2 blocks β-adrenergic-induced increases in contractility [356]. Conversely, high glucose, or glucose in the presence of hypoxia, upregulates Nox4 in retinal epithelial cells leading to VEGF expression [353]; in retinal endothelial cells, where it promotes cell death [357]; in microvascular endothelial cells, where it controls VEGF expression [358] and mediates monocyte adhesion [359]; and in glomerular mesangial cells, where it promotes endothelial nitric oxide synthase (eNOS) uncoupling [327]. Moreover, glucosestimulated insulin secretion in pancreatic islets depends upon Nox4-derived H2O2, such that Nox4 knockout mice exhibit impaired glucose tolerance and peripheral insulin resistance [360]. All of these studies support a close relationship between Nox4 and diabetes, which is perhaps not surprising given the mechanistic link between Nox4 and glucose metabolism [219].

Several gain- and loss of function in vivo studies suggest that Nox4 contributes to angiogenesis, a process of new blood vessel formation involved in a variety of physiological and pathological conditions. Endothelial cell-specific Nox4 overexpression in mice increases angiogenesis and enhances blood flow recovery from hindlimb ischemia [155, 364]. Cardiomyocyte-targeted overexpression of Nox4 in mice facilitates preservation of myocardial capillaries by increasing angiogenic activity following pressure overload [157]. Conversely, genetic deletion of Nox4 in mice reduces angiogenesis in response to various stimuli including femoral artery ligation [156], pressure overload [157], TGF-β1 [365, 366], and prostacyclin receptor agonist (cicaprost) [367]. In Nox4 knockout mice, exercise-induced capillary growth is also inhibited, while developmental angiogenesis in the retina is unchanged [368]. In stark contrast to the above, two in vivo studies showed no effect of genetic ablation of Nox4 on chronic hypoxia-induced vascular remodeling [287, 297]. Compelling evidence suggests that angiogenic activity in endothelial cells depends upon Nox4-redox signaling. In cultured endothelial cells, Nox4 promotes upregulation of a potent proangiogenic cytokine, VEGF [358, 369, 370]. Nox4 also mediates endothelial angiogenic responses stimulated by VEGF [371, 372], TGF-β1 [365, 366], insulin [358], and hypoxia [53]. Angiogenesis is necessary for cancer progression and, therefore an important target for cancer therapy. In a murine model of fibrosarcomas, Nox4 knockout mice exhibit a significant reduction in tumor vascularization [373]. Similarly, Nox4 knockdown in glioblastoma cells reduces tumorassociated angiogenesis [374]. Overall, substantial evidence supports a critical role for Nox4 in angiogenesis.

5.11 Bone Remodeling

5.13 Cancer

One of the earliest reports on Nox4 showed that it mediates O2•- generation as well as resorption pit formation by osteoclasts [3]. Subsequent work showed that bone morphogenetic protein-2 requires Nox4-derived ROS to stimulate osteoblast differentiation [361], suggesting that Nox4 plays a role in both osteoblastogenesis and osteoclast function. Using Nox4 knockout mice, Goettsch et al. [362] demonstrated that the overarching role of Nox4 is to promote bone loss. They found that loss of Nox4 increased bone

In addition to angiogenesis, multiple studies implicate Nox4 in cancer. In general, upregulation of Nox4 expression occurs in carcinomas that originate from epithelial cells exposed to environmental stresses, including those of the head and neck, pancreas, lungs, and gastrointestinal tract [375, 376], and has been implicated in proliferation, apoptosis, cell migration, invasion, and tumorigenicity. Nox4 is also involved in epithelial to mesenchymal cell transition (EMT), an early event in cancer metastasis, yet, in some cancers Nox4 is anti-

density by attenuating osteoclastogenesis, work confirmed by Sun et al. [363]. Increased Nox4 in human samples correlates with increased osteoclast activity, suggesting that Nox4 may contribute to osteoporosis [362].

5.12 Angiogenesis

12

Nox4: From Discovery to Pathophysiology

195

Table 12.3 Nox4 and cancer Cancer type Skin Malignant melanoma Gastrointestinal Pancreatic adenocarcinoma cells Hepatocellular carcinoma Colorectal carcinoma

Mechanism of action

Human

a

Human Human

#Apoptosis "Senescence #Proliferation "Proliferation, #Apoptosis, "Cell migration "Proliferation

Human

Human

"Proliferation via G2/M

References [35, 152] [193, 378] [379] [166] [380]

Gastric cancer Brain Glioblastoma

Human

"Proliferation #Apoptosis "Angiogenesis

[153] [382]

Breast Breast cancer

Human

"Tumorigenicity "Cell migration

[62, 383]

Genitourinary system Urothelial carcinogenesis Renal cell carcinoma

Human/mouse Human

"Cell cycle progression "Cytokines "Cell invasion "Tumorigenesis

[384] [339] [342]

Human

"Cell invasion

[385]

Human Human

"Cell invasion "Proliferation #Apoptosis

[154] [386]

Mouse

"Angiogenesis

[387]

Reproductive system Ovarian cancer Lung Non-small cell lung cancer Malignant pleural mesothelioma Bone Fibrosarcoma a

Species

[381]

"=increase; #=decrease

tumorigenic. Shortly after its discovery, Nox4 was linked to proliferation of malignant melanoma cells [36] and was subsequently shown to regulate G2-M cell cycle progression in these cells [152]. It was also demonstrated to protect pancreatic cancer cells from apoptosis, thus enhancing their survival [191, 377]. Furthermore, Nox4 expression is elevated in glioblastomas and Nox4 siRNA inhibits growth and enhances apoptosis in cultured glioma cell lines [153]. Nox4 expression is also elevated in breast cancer and shows increased tumorigenicity [63]. Table 12.3 summarizes the various tumors linked to Nox4. The above studies focus on the role of Nox4 in enhancing proliferation and reducing apoptosis, which both serve to promote tumor growth and survival, but Nox4 has also been shown to affect other stages of cancer progression. Nox4 promotes EMT in breast and pancreatic cancer [388, 389] and has been implicated in metastasis and cell invasion [337, 390], and in particular in invadopodia formation, which correlates with invasiveness. In Src-3T3 cells, HeLa cells, and C8161.9 human melanoma cells, knockdown of Nox4 reduces invadopodia formation [209, 391]. More recent work has shown that upregulation of endothelial Nox4

in primary invasive breast tumors is associated with lymphangiogenesis and lymph node metastasis [392]. Some of the apparent contradictions with the role of Nox4 in cancer shown in Table 12.3 may be explained by other mutations. For example, Ma et al. [393] found a mutation in TP53 that can switch Nox4 from being pro-survival to being linked to poor survival; that is, in tumors with the TP53 mutations, Nox4 was pro-proliferative and anti-apoptotic, while the opposite was true in tumors with wild type TP53. In addition to contributing to cancer progression, Nox4 expression has been proposed to predict cancer outcomes with varying success. For example, Lin et al. [394] showed a strong correlation of Nox4 levels with overall survival in colorectal cancer. Eun et al. [395] found similar results in HCC, but later showed that nuclear Nox4 was associated with poor prognosis [396], a result confirmed by others in renal cell carcinoma [397]. Conversely, high Nox4 expression was found to be associated with tumor size and worse prognosis in patients with gastric cancers [398] and with poor overall survival in endometrial cancer [387] and esophageal squamous cell carcinoma [382], while low Nox4 was significantly associated with survival of oral tongue squamous cell

196

carcinoma patients [399]. These conflicting reports underline the necessity of considering the context of particular cancers and stages of cancer development when investigating the relationship between Nox4 and cancer.

5.14 Neurodegenerative Disorders Alzheimer’s disease (AD) is the most common neurodegenerative disorder in individuals aged >65 years worldwide. Progressive memory deficits and cognitive decline are the major clinical feature of AD. The abnormal deposition of amyloid β (Aβ) peptide and intracellular accumulation of neurofibrillary tangles of tau protein are thought to be key etiologic factors contributing to neurodegeneration. Numerous studies have implicated oxidative stress in the pathogenesis of AD. Oxidized protein levels are markedly increased in the most severely affected brain regions of AD patients [400]. In brain samples from AD patients, Nox4 expression is highly elevated compared with subjects without AD [378, 379, 401]. In mice, Nox4 expression and ROS levels are significantly higher in the micro-vessels of the cortex and hippocampus (one of the earliest and most affected regions in AD) compared to other regions of the brain [402]; this may explain the vulnerability of the cortex and hippocampus to oxidative stress in AD. Several in vitro studies suggest that Aβ mediates upregulation of Nox4/ROS in various brain cell types, including human astrocytes [379], human neuroblastoma cells [403–405], murine neuroblastoma cells [385], and murine cortical neurons [406]. In human neuroblastoma cells, Aβ activates Nox4-dependent ROS via TLR4-mediated signaling, which promotes neuroinflammation [403]. Genetic suppression of Nox4 attenuates ROS generation and cytotoxicity in mouse neuron cultures [407] and a differentiated rat pheochromocytoma cell line [380]. The in vivo role of Nox4 in AD was evaluated using a humanized knock-in model of Aβ pathology (APP/PS1 mice), which exhibit impaired cognitive performance and decreased key synaptic proteins in an age-dependent manner [408]. APP/PS1 mice exhibit significantly increased Nox4 expression/activity in the frontal cortex as compared to littermate controls [408]. Further, genetic suppression of Nox4 (via miR-204-3p, a negative regulator of Nox4 transcription) and pharmacological targeting of Nox4 (via GLX351322) in APP/PS1 mice leads to reduced oxidative stress and amyloid load in the hippocampus and improved cognitive function [133]. In a different humanized murine model of AD (induced by tau protein delivery to the brain), neuronaltargeted genetic knockdown of Nox4 reduced neurotoxicity and prevented cognitive decline [378]. Overall, these studies suggest the involvement of Nox4-dependent ROS in the pathogenesis of neurodegenerative disorders.

L. Hecker et al.

5.15 Stroke The role of Nox4 in cerebral ischemia and stroke is complex due to the multiple cell types involved. Nox4 is induced in neurons, astrocytes, and endothelial cells after middle cerebral artery occlusion (MCAO) to induce stroke in mice [409] and in pericytes subjected to hypoxia [410]. Importantly, ischemia induced by MCAO markedly increases cortical Nox4 expression for up to 30 days post-injury, first likely in neurons and later in newly formed vessels [37]. Nox4-/- mice are largely protected from oxidative stress, blood-brain barrier disruption, and neuronal apoptosis after transient or permanent MCAO [38]. Conversely, transgenic expression of Nox4 in pericytes increases infarct volume and blood-brain barrier disruption after MCAO [381]. Cell type-specific knockout of Nox4 showed that endothelial Nox4 disrupts the blood-brain barrier, while neuronal Nox4 causes neuronal neurotoxicity [411]. The recent identification of a single nucleotide polymorphism in humans that is associated with decreased risk for, and a better short-term recovery from, ischemic stroke suggests that Nox4 may be clinically important as well [412]. Indeed, serum Nox4, presumably found in microvesicles, has been suggested as an independent predictor for delayed cerebral ischemia and poor outcome after aneurysmal subarachnoid hemorrhage [413].

6

Inhibitors

As noted in the previous section, accumulating data support the concept that Nox4 plays critical roles in the pathogenesis of various diseases, which has led to considerable efforts focused on the development of Nox4 inhibitors. It is important to note that Nox4 knockout mice are viable with no appreciable phenotype in the unstressed state [28], which is reassuring from a drug-targeting standpoint. Although highthroughput screening approaches have been used to identify putative Nox4 small-molecule inhibitors, these typically utilize ROS detection-based screening assays that have limited specificity. Thus, it is difficult to discern whether a putative inhibitor is acting directly on Nox4 or is rather inhibiting a signaling pathway leading to Nox4 induction/activation. One study reported that of >350 Nox inhibitors described, a majority of them did not directly block enzymatic activity, but rather interfered with upstream signaling pathways or were ROS scavengers [383]. The ideal Nox4 inhibitor needs to fulfill several criteria: limited antioxidant activity, lack of effect on other sources of ROS upstream of Nox4 signaling, demonstrated direct inhibition of Nox4 activity, and high selectivity for Nox4. Below, we have summarized current knowledge of the most promising and/or most widely used putative Nox4 inhibitors (Fig. 12.5).

12

Nox4: From Discovery to Pathophysiology

197

Fig. 12.5 Pipeline of Nox4 inhibitors. Numerous putative Nox4 inhibitors have been developed and are in various phases of testing

6.1

APX-115

APX-115, 3-phenyl-1-(pyridin-2-yl)-4-propyl-1-5hydroxypyrazol HCl, (also previously referred to as Ewha18278) [384], is a first-in-class pan Nox inhibitor currently being developed by Aptabio Therapeutics Inc. [414]. APX-115 has been shown to inhibit Nox1, Nox2, and Nox4 in the micromolar range [384]. APX-115 is orally available and has demonstrated pre-clinical efficacy in animal models of ovariectomy-induced osteoporosis [384] and diabetic nephropathy [344, 345]. Treatment with APX-115 significantly improved insulin resistance in a type 2 diabetic mouse model [344, 345]. In a mouse model of streptozotocin-induced diabetic kidney injury, APX-115 inhibited renal oxidative stress and prevented albuminuria, glomerular hypertrophy, tubular injury, podocyte injury, fibrosis, and inflammation as effectively as losartan alone (the current standard of treatment for kidney injury in diabetic patients) [345]. A Phase I clinical trial to assess the safety, tolerability, pharmacokinetics, and pharmacodynamics of single and multiple ascending doses of APX-115 in 88 healthy patients was completed in 2019 (NCT03694041). In 2020, a phase 2 clinical trial with APX-115 for Type 2 diabetic nephropathy patients was initiated (NCT04534439) [414], but results are not yet available. Of interest, a Phase 2 clinical trial was initiated in 2021 to evaluate the safety/efficacy of APX-115 in hospitalized patients with mild to moderate COVID-19 (NCT04880109) and is currently recruiting patients.

6.2

GLX

GLX351322 was initially identified via a high-throughput approach using an inducible Nox4 overexpressing cell line and an Amplex Red assay-based screen [386], and data suggests GLX351322 selectively targets Nox4 over Nox2 [386]. Other Nox-targeting GLX compounds have also been reported, including GLX7013114 [415–417] and GLX481304 [418]. Pre-clinical studies have evaluated GLX351322 as a therapeutic strategy for type 2 diabetes. GLX351322 treatment inhibited high-glucose-induced ROS production and cell death in human islet cells [386]. In a murine model of high-fat diet-induced glucose intolerance, treatment with GLX351322 counteracted non-fasting hyperglycemia and impaired glucose tolerance, [386]. Another study demonstrated that treatment with GLX7013114 led to protection from stress-induced human islet cell death [415]. These studies suggest that selective Nox4-inhibition may be a therapeutic strategy in type 2 diabetes. Recent pre-clinical studies have also evaluated the therapeutic utility of GLX candidates for other disease indications. In an in vitro model using rat lens epithelial explants, Nox4 inhibition by GLX7013114 was shown to reduce TGF-β-induced EMT and profibrotic gene expression [416]. In addition, GLX481304 inhibited ROS production and improved contractility in isolated cardiomyocytes and contraction of whole hearts retrogradely perfused following global ischemia [418]. GLX candidates have not yet reached clinical trials;

198

L. Hecker et al.

however, they are currently being developed by Glucox Biotech (Sweden).

6.3

Setanaxib

Setanaxib (formally GKT137831 or GKT136901), initially developed by Genkyotex (Geneva, Switzerland) [419], remains marketed as a Nox1/4 inhibitor. Numerous pre-clinical studies have provided evidence supporting their Nox inhibition and therapeutic utility for pulmonary fibrosis [29], liver disease [420, 421], cardiac remodeling/fibrosis [422, 423], ischemic retinal disease [357, 424, 425], diabetes [426], and cancer [427]. Although these compounds have been evaluated in clinical trials [428, 429], their specificity as bona fide Nox4 inhibitors has recently been disproven. A rigorous in-depth evaluation of putative Nox inhibitors demonstrated that these candidates are, in fact, inactive as Nox inhibitors but rather interfere with peroxidase-dependent assays [430]. In particular, these compounds non-specifically inhibit horseradish peroxidase, a component of the assay for H2O2, as they show similar inhibitory activity in the presence of H2O2 alone using a cell-free assay system [430], suggesting they act as general reducing agents. Thus, the observed protective effects of these candidates reported in pre-clinical in vivo studies do not stem from direct inhibition of Nox, but rather from a non-specific redox mechanism and/or downstream effects on Nox4 activity.

7

Conclusion

In summary, our knowledge of Nox4 has exploded since its discovery twenty years ago. Expression patterns have been established, regulatory mechanisms identified, structureactivity relationships explored, physiologic effects identified, knockout models developed, pathophysiology investigated and inhibitor development initiated. Nox4 has a number of unique attributes such as a lack of required cytosolic regulatory subunits and a propensity to produce mostly H2O2 rather than O2•-. It plays a key role in numerous fundamental cellular processes, including cytoskeletal structure and dynamics, proliferation, differentiation, survival and metabolic homeostasis. In many ways, Nox4 is required for normal organ functioning, but it also contributes to multiple disease processes, especially fibrosis and diabetic vascular disease. Due to conflicting literature using genetically modified animal models and a rudimentary understanding of genetic associations and polymorphisms of Nox4 in human disease, much remains to be learned about Nox4 and its contribution to physiology and pathophysiology. Development of specific Nox4 inhibitors will not only spur

the field forward, but will help to realize the potential clinical utility of therapies targeting Nox4 in numerous diseases. Acknowledgments LH received support from the Department of Veterans Affairs (BX006003), the Office of the Assistance Secretary of Defense for Health Affairs through the Peer Reviewed Medical Research Program under award no. W81XWH-17-1-0443, and National Institutes of Health grants 1R21AG054766-01 and 1R41HL151043-01A1. KKG is supported by NIH grants R21AI163427 and R56HL152167.

References 1. Geiszt M, Kopp JB, Varnai P, Leto TL (2000) Identification of renox, an NAD(P)H oxidase in kidney. Proc Natl Acad Sci U S A 97(14):8010–8014 2. Shiose A, Kuroda J, Tsuruya K, Hirai M, Hirakata H, Naito S, Hattori M, Sakaki Y, Sumimoto H (2001) A novel superoxideproducing NAD(P)H oxidase in kidney. J Biol Chem 276(2): 1417–1423 3. Yang S, Madyastha P, Bingel S, Ries W, Key L (2001) A new superoxide-generating oxidase in murine osteoclasts. J Biol Chem 276(8):5452–5458 4. Nisimoto Y, Diebold BA, Cosentino-Gomes D, Lambeth JD (2014) Nox4: a hydrogen peroxide-generating oxygen sensor. Biochemistry 53(31):5111–5120. https://doi.org/10.1021/bi500331y 5. Ambasta RK, Kumar P, Griendling KK, Schmidt HH, Busse R, Brandes RP (2004) Direct interaction of the novel Nox proteins with p22phox is required for the formation of a functionally active NADPH oxidase. J Biol Chem 279(44):45935–45941 6. Martyn KD, Frederick LM, von Loehneysen K, Dinauer MC, Knaus UG (2006) Functional analysis of Nox4 reveals unique characteristics compared to other NADPH oxidases. Cell Signal 18(1):69–82 7. Prior KK, Leisegang MS, Josipovic I, Lowe O, Shah AM, Weissmann N, Schroder K, Brandes RP (2016) CRISPR/Cas9mediated knockout of p22phox leads to loss of Nox1 and Nox4, but not Nox5 activity. Redox Biol 9:287–295. https://doi.org/10. 1016/j.redox.2016.08.013 8. Kawahara T, Ritsick D, Cheng G, Lambeth JD (2005) Point mutations in the proline-rich region of p22phox are dominant inhibitors of Nox1- and Nox2-dependent reactive oxygen generation. J Biol Chem 280(36):31859–31869 9. von Lohneysen K, Noack D, Jesaitis AJ, Dinauer MC, Knaus UG (2008) Mutational analysis reveals distinct features of the Nox4p22 phox complex. J Biol Chem 283(50):35273–35282. https:// doi.org/10.1074/jbc.M804200200 10. Nisimoto Y, Jackson HM, Ogawa H, Kawahara T, Lambeth JD (2010) Constitutive NADPH-dependent electron transferase activity of the Nox4 dehydrogenase domain. Biochemistry 49(11): 2433–2442. https://doi.org/10.1021/bi9022285 11. Ago T, Kuroda J, Pain J, Fu C, Li H, Sadoshima J (2010) Upregulation of Nox4 by hypertrophic stimuli promotes apoptosis and mitochondrial dysfunction in cardiac myocytes. Circ Res 106(7):1253–1264. https://doi.org/10.1161/CIRCRESAHA.109. 213116 12. von Lohneysen K, Noack D, Wood MR, Friedman JS, Knaus UG (2010) Structural insights into Nox4 and Nox2: motifs involved in function and cellular localization. Mol Cell Biol 30(4):961–975. https://doi.org/10.1128/MCB.01393-09 13. Mahadev K, Motoshima H, Wu X, Ruddy JM, Arnold RS, Cheng G, Lambeth JD, Goldstein BJ (2004) The NAD(P)H oxidase homolog Nox4 modulates insulin-stimulated generation of

12

Nox4: From Discovery to Pathophysiology

H2O2 and plays an integral role in insulin signal transduction. Mol Cell Biol 24(5):1844–1854 14. Jackson HM, Kawahara T, Nisimoto Y, Smith SM, Lambeth JD (2010) Nox4 B-loop creates an interface between the transmembrane and dehydrogenase domains. J Biol Chem 285(14): 10281–10290. https://doi.org/10.1074/jbc.M109.084939 15. Dikalov SI, Dikalova AE, Bikineyeva AT, Schmidt HH, Harrison DG, Griendling KK (2008) Distinct roles of Nox1 and Nox4 in basal and angiotensin II-stimulated superoxide and hydrogen peroxide production. Free Radic Biol Med. https://doi.org/10.1016/j. freeradbiomed.2008.08.013 16. Serrander L, Cartier L, Bedard K, Banfi B, Lardy B, Plastre O, Sienkiewicz A, Forro L, Schlegel W, Krause KH (2007) NOX4 activity is determined by mRNA levels and reveals a unique pattern of ROS generation. Biochem J 406(1):105–114 17. Takac I, Schroder K, Zhang L, Lardy B, Anilkumar N, Lambeth JD, Shah AM, Morel F, Brandes RP (2011) The E-loop is involved in hydrogen peroxide formation by the NADPH oxidase Nox4. J Biol Chem 286(15):13304–13313. https://doi.org/10.1074/jbc. M110.192138 18. Goyal P, Weissmann N, Rose F, Grimminger F, Schafers HJ, Seeger W, Hanze J (2005) Identification of novel Nox4 splice variants with impact on ROS levels in A549 cells. Biochem Biophys Res Commun 329(1):32–39. https://doi.org/10.1016/j. bbrc.2005.01.089 19. Anilkumar N, San Jose G, Sawyer I, Santos CX, Sand C, Brewer AC, Warren D, Shah AM (2013) A 28-kDa splice variant of NADPH oxidase-4 is nuclear-localized and involved in redox signaling in vascular cells. Arterioscler Thromb Vasc Biol 33(4): e104–e112. https://doi.org/10.1161/ATVBAHA.112.300956 20. Varga ZV, Pipicz M, Baan JA, Baranyai T, Koncsos G, Leszek P, Kusmierczyk M, Sanchez-Cabo F, Garcia-Pavia P, Brenner GJ, Giricz Z, Csont T, Mendler L, Lara-Pezzi E, Pacher P, Ferdinandy P (2017) Alternative splicing of NOX4 in the failing human heart. Front Physiol 8:935. https://doi.org/10.3389/fphys.2017.00935 21. Moloney JN, Jayavelu AK, Stanicka J, Roche SL, O'Brien RL, Scholl S, Bohmer FD, Cotter TG (2017) Nuclear membranelocalised NOX4D generates pro-survival ROS in FLT3-ITDexpressing AML. Oncotarget 8(62):105440–105457. https://doi. org/10.18632/oncotarget.22241 22. Krause KH (2004) Tissue distribution and putative physiological function of NOX family NADPH oxidases. Jpn J Infect Dis 57(5): S28–S29 23. He C, Larson-Casey JL, Davis D, Hanumanthu VS, Longhini ALF, Thannickal VJ, Gu L, Carter AB (2019) NOX4 modulates macrophage phenotype and mitochondrial biogenesis in asbestosis. JCI. Insight 4(16). https://doi.org/10.1172/jci.insight.126551 24. Moon JS, Nakahira K, Chung KP, DeNicola GM, Koo MJ, Pabon MA, Rooney KT, Yoon JH, Ryter SW, Stout-Delgado H, Choi AM (2016) NOX4-dependent fatty acid oxidation promotes NLRP3 inflammasome activation in macrophages. Nat Med 22(9): 1002–1012. https://doi.org/10.1038/nm.4153 25. Lee CF, Qiao M, Schroder K, Zhao Q, Asmis R (2010) Nox4 is a novel inducible source of reactive oxygen species in monocytes and macrophages and mediates oxidized low density lipoproteininduced macrophage death. Circ Res 106(9):1489–1497. https:// doi.org/10.1161/CIRCRESAHA.109.215392 26. Zelko IN, Folz RJ (2015) Regulation of oxidative stress in pulmonary artery endothelium. modulation of extracellular superoxide dismutase and NOX4 expression using histone deacetylase class I inhibitors. Am J Respir Cell Mol Biol 53(4):513–524. https://doi. org/10.1165/rcmb.2014-0260OC 27. Sturrock A, Cahill B, Norman K, Huecksteadt TP, Hill K, Sanders K, Karwande SV, Stringham JC, Bull DA, Gleich M, Kennedy TP, Hoidal JR (2006) Transforming growth factor-beta1

199 induces Nox4 NAD(P)H oxidase and reactive oxygen speciesdependent proliferation in human pulmonary artery smooth muscle cells. Am J Physiol Lung Cell Mol Physiol 290(4):L661–L673 28. Carnesecchi S, Deffert C, Donati Y, Basset O, Hinz B, PreynatSeauve O, Guichard C, Arbiser JL, Banfi B, Pache JC, BarazzoneArgiroffo C, Krause KH (2011) A key role for NOX4 in epithelial cell death during development of lung fibrosis. Antioxid Redox Signal 15(3):607–619. https://doi.org/10.1089/ars.2010.3829 29. Hecker L, Logsdon NJ, Kurundkar D, Kurundkar A, Bernard K, Hock T, Meldrum E, Sanders YY, Thannickal VJ (2014) Reversal of persistent fibrosis in aging by targeting Nox4-Nrf2 redox imbalance. Sci Transl Med 6(231):231ra247. https://doi.org/10.1126/ scitranslmed.3008182 30. Tsubouchi K, Araya J, Minagawa S, Hara H, Ichikawa A, Saito N, Kadota T, Sato N, Yoshida M, Kurita Y, Kobayashi K, Ito S, Fujita Y, Utsumi H, Yanagisawa H, Hashimoto M, Wakui H, Yoshii Y, Ishikawa T, Numata T, Kaneko Y, Asano H, Yamashita M, Odaka M, Morikawa T, Nakayama K, Nakanishi Y, Kuwano K (2017) Azithromycin attenuates myofibroblast differentiation and lung fibrosis development through proteasomal degradation of NOX4. Autophagy 13(8): 1420–1434. https://doi.org/10.1080/15548627.2017.1328348 31. Cheng G, Cao Z, Xu X, van Meir EG, Lambeth JD (2001) Homologs of gp91phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269(1–2):131–140 32. Bartsch C, Bekhite MM, Wolheim A, Richter M, Ruhe C, Wissuwa B, Marciniak A, Muller J, Heller R, Figulla HR, Sauer H, Wartenberg M (2011) NADPH oxidase and eNOS control cardiomyogenesis in mouse embryonic stem cells on ascorbic acid treatment. Free Radic Biol Med 51(2):432–443. https://doi.org/10. 1016/j.freeradbiomed.2011.04.029 33. Piccoli C, Ria R, Scrima R, Cela O, D'Aprile A, Boffoli D, Falzetti F, Tabilio A, Capitanio N (2005) Characterization of mitochondrial and extra-mitochondrial oxygen consuming reactions in human hematopoietic stem cells. Novel evidence of the occurrence of NAD(P)H oxidase activity. J Biol Chem 280(28): 26467–26476. https://doi.org/10.1074/jbc.M500047200 34. Yang S, Zhang Y, Ries W, Key L (2004) Expression of Nox4 in osteoclasts. J Cell Biochem 92(2):238–248 35. Chamulitrat W, Stremmel W, Kawahara T, Rokutan K, Fujii H, Wingler K, Schmidt HH, Schmidt R (2004) A constitutive NADPH oxidase-like system containing gp91phox homologs in human keratinocytes. J Invest Dermatol 122(4):1000–1009. https://doi. org/10.1111/j.0022-202X.2004.22410.x 36. Brar SS, Kennedy TP, Sturrock AB, Huecksteadt TP, Quinn MT, Whorton AR, Hoidal JR (2002) An NAD(P)H oxidase regulates growth and transcription in melanoma cells. Am J Physiol Cell Physiol 282(6):C1212–C1224 37. Vallet P, Charnay Y, Steger K, Ogier-Denis E, Kovari E, Herrmann F, Michel JP, Szanto I (2005) Neuronal expression of the NADPH oxidase NOX4, and its regulation in mouse experimental brain ischemia. Neuroscience 132(2):233–238. https://doi. org/10.1016/j.neuroscience.2004.12.038 38. Kleinschnitz C, Grund H, Wingler K, Armitage ME, Jones E, Mittal M, Barit D, Schwarz T, Geis C, Kraft P, Barthel K, Schuhmann MK, Herrmann AM, Meuth SG, Stoll G, Meurer S, Schrewe A, Becker L, Gailus-Durner V, Fuchs H, Klopstock T, de Angelis MH, Jandeleit-Dahm K, Shah AM, Weissmann N, Schmidt HH (2010) Post-stroke inhibition of induced NADPH oxidase type 4 prevents oxidative stress and neurodegeneration. PLoS Biol 8(9). https://doi.org/10.1371/journal.pbio.1000479 39. Kanda Y, Hinata T, Kang SW, Watanabe Y (2011) Reactive oxygen species mediate adipocyte differentiation in mesenchymal stem cells. Life Sci 89(7–8):250–258. https://doi.org/10.1016/j.lfs. 2011.06.007

200 40. Grange L, Nguyen MV, Lardy B, Derouazi M, Campion Y, Trocme C, Paclet MH, Gaudin P, Morel F (2006) NAD(P)H oxidase activity of Nox4 in chondrocytes is both inducible and involved in collagenase expression. Antioxid Redox Signal 8(9–10):1485–1496. https://doi.org/10.1089/ars.2006.8.1485 41. Kim KS, Choi HW, Yoon HE, Kim IY (2010) Reactive oxygen species generated by NADPH oxidase 2 and 4 are required for chondrogenic differentiation. J Biol Chem 285(51):40294–40302. https://doi.org/10.1074/jbc.M110.126821 42. Ikeda R, Ishii K, Hoshikawa Y, Azumi J, Arakaki Y, Yasui T, Matsuura S, Matsumi Y, Kono Y, Mizuta Y, Kurimasa A, Hisatome I, Friedman SL, Kawasaki H, Shiota G (2011) Reactive oxygen species and NADPH oxidase 4 induced by transforming growth factor beta1 are the therapeutic targets of polyenylphosphatidylcholine in the suppression of human hepatic stellate cell activation. Inflamm Res 60(6):597–604. https://doi.org/10.1007/ s00011-011-0309-6 43. Piwkowska A, Rogacka D, Audzeyenka I, Jankowski M, Angielski S (2011) High glucose concentration affects the oxidantantioxidant balance in cultured mouse podocytes. J Cell Biochem 112(6):1661–1672. https://doi.org/10.1002/jcb.23088 44. Palumbo S, Shin YJ, Ahmad K, Desai AA, Quijada H, Mohamed M, Knox A, Sammani S, Colson BA, Wang T, Garcia JG, Hecker L (2017) Dysregulated Nox4 ubiquitination contributes to redox imbalance and age-related severity of acute lung injury. Am J Physiol Lung Cell Mol Physiol 312(3):L297–L308. https:// doi.org/10.1152/ajplung.00305.2016 45. Ago T, Kitazono T, Kuroda J, Kumai Y, Kamouchi M, Ooboshi H, Wakisaka M, Kawahara T, Rokutan K, Ibayashi S, Iida M (2005) NAD(P)H oxidases in rat basilar arterial endothelial cells. Stroke 36(5):1040–1046. https://doi.org/10.1161/01.STR.0000163111. 05825.0b 46. Hu T, Ramachandrarao SP, Siva S, Valancius C, Zhu Y, Mahadev K, Toh I, Goldstein BJ, Woolkalis M, Sharma K (2005) Reactive oxygen species production via NADPH oxidase mediates TGF-beta-induced cytoskeletal alterations in endothelial cells. Am J Physiol Renal Physiol 289(4):F816–F825. https://doi.org/10. 1152/ajprenal.00024.2005 47. Van Buul JD, Fernandez-Borja M, Anthony EC, Hordijk PL (2005) Expression and localization of NOX2 and NOX4 in primary human endothelial cells. Antioxid Redox Signal 7(3–4):308–317. https://doi.org/10.1089/ars.2005.7.308 48. Ellmark SH, Dusting GJ, Fui MN, Guzzo-Pernell N, Drummond GR (2005) The contribution of Nox4 to NADPH oxidase activity in mouse vascular smooth muscle. Cardiovasc Res 65(2):495–504 49. Janiszewski M, Lopes LR, Carmo AO, Pedro MA, Brandes RP, Santos CX, Laurindo FR (2005) Regulation of NAD(P)H oxidase by associated protein disulfide isomerase in vascular smooth muscle cells. J Biol Chem 280(49):40813–40819 50. Laude K, Cai H, Fink B, Hoch N, Weber DS, McCann L, Kojda G, Fukai T, Schmidt HH, Dikalov S, Ramasamy S, Gamez G, Griendling KK, Harrison DG (2005) Hemodynamic and biochemical adaptations to vascular smooth muscle overexpression of p22phox in mice. Am J Physiol Heart Circ Physiol 288(1):H7–H12 51. Pedruzzi E, Guichard C, Ollivier V, Driss F, Fay M, Prunet C, Marie JC, Pouzet C, Samadi M, Elbim C, O’Dowd Y, Bens M, Vandewalle A, Gougerot-Pocidalo MA, Lizard G, Ogier-Denis E (2004) NAD(P)H oxidase Nox-4 mediates 7-ketocholesterolinduced endoplasmic reticulum stress and apoptosis in human aortic smooth muscle cells. Mol Cell Biol 24(24):10703–10717 52. Hilenski LL, Clempus RE, Quinn MT, Lambeth JD, Griendling KK (2004) Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 24(4):677–683 53. Pendyala S, Gorshkova IA, Usatyuk PV, He D, Pennathur A, Lambeth JD, Thannickal VJ, Natarajan V (2009) Role of Nox4

L. Hecker et al. and Nox2 in hyperoxia-induced reactive oxygen species generation and migration of human lung endothelial cells. Antioxid Redox Signal 11(4):747–764. https://doi.org/10.1089/ARS.2008.2203 54. Sturrock A, Huecksteadt TP, Norman K, Sanders K, Murphy TM, Chitano P, Wilson K, Hoidal JR, Kennedy TP (2007) Nox4 mediates TGF-beta1-induced retinoblastoma protein phosphorylation, proliferation, and hypertrophy in human airway smooth muscle cells. Am J Physiol Lung Cell Mol Physiol 292(6):L1543– L1555. https://doi.org/10.1152/ajplung.00430.2006 55. Pendergrass KD, Gwathmey TM, Michalek RD, Grayson JM, Chappell MC (2009) The angiotensin II-AT1 receptor stimulates reactive oxygen species within the cell nucleus. Biochem Biophys Res Commun 384(2):149–154. https://doi.org/10.1016/j.bbrc. 2009.04.126 56. Lee CF, Ullevig S, Kim HS, Asmis R (2013) Regulation of monocyte adhesion and migration by Nox4. PLoS One 8(6):e66964. https://doi.org/10.1371/journal.pone.0066964 57. Sciarretta S, Volpe M, Sadoshima J (2014) NOX4 regulates autophagy during energy deprivation. Autophagy 10(4):699–701. https://doi.org/10.4161/auto.27955 58. Chen K, Kirber MT, Xiao H, Yang Y, Keaney JF Jr (2008) Regulation of ROS signal transduction by NADPH oxidase 4 localization. J Cell Biol 181(7):1129–1139. https://doi.org/10.1083/jcb. 200709049 59. Helmcke I, Heumuller S, Tikkanen R, Schroder K, Brandes RP (2009) Identification of structural elements in Nox1 and Nox4 controlling localization and activity. Antioxid Redox Signal 11(6):1279–1287. https://doi.org/10.1089/ARS.2008.2383 60. Wu RF, Ma Z, Liu Z, Terada LS (2010) Nox4-derived H2O2 mediates endoplasmic reticulum signaling through local Ras activation. Mol Cell Biol 30(14):3553–3568. https://doi.org/10.1128/ MCB.01445-09 61. Kuroda J, Ago T, Matsushima S, Zhai P, Schneider MD, Sadoshima J (2010) NADPH oxidase 4 (Nox4) is a major source of oxidative stress in the failing heart. Proc Natl Acad Sci U S A 107(35):15565–15570. https://doi.org/10.1073/pnas.1002178107 62. Block K, Gorin Y, Abboud HE (2009) Subcellular localization of Nox4 and regulation in diabetes. Proc Natl Acad Sci U S A 106(34):14385–14390. https://doi.org/10.1073/pnas.0906805106 63. Graham KA, Kulawiec M, Owens KM, Li X, Desouki MM, Chandra D, Singh KK (2010) NADPH oxidase 4 is an oncoprotein localized to mitochondria. Cancer Biol Ther 10(3):223–231. https://doi.org/10.4161/cbt.10.3.12207 64. Shanmugasundaram K, Nayak BK, Friedrichs WE, Kaushik D, Rodriguez R, Block K (2017) NOX4 functions as a mitochondrial energetic sensor coupling cancer metabolic reprogramming to drug resistance. Nat Commun 8(1):997. https://doi.org/10.1038/s41467017-01106-1 65. Beretta M, Santos CX, Molenaar C, Hafstad AD, Miller CC, Revazian A, Betteridge K, Schroder K, Streckfuss-Bomeke K, Doroshow JH, Fleck RA, Su TP, Belousov VV, Parsons M, Shah AM (2020) Nox4 regulates InsP3 receptor-dependent Ca(2+) release into mitochondria to promote cell survival. EMBO J 39(19):e103530. https://doi.org/10.15252/embj.2019103530 66. Datla SR, McGrail DJ, Vukelic S, Huff LP, Lyle AN, Pounkova L, Lee M, Seidel-Rogol B, Khalil MK, Hilenski LL, Terada LS, Dawson MR, Lassegue B, Griendling KK (2014) Poldip2 controls vascular smooth muscle cell migration by regulating focal adhesion turnover and force polarization. Am J Physiol Heart Circ Physiol 307(7):H945–H957. https://doi.org/10.1152/ajpheart.00918.2013 67. Lyle AN, Deshpande NN, Taniyama Y, Seidel-Rogol B, Pounkova L, Du P, Papaharalambus C, Lassegue B, Griendling KK (2009) Poldip2, a novel regulator of Nox4 and cytoskeletal integrity in vascular smooth muscle cells. Circ Res 105(3): 249–259. https://doi.org/10.1161/CIRCRESAHA.109.193722

12

Nox4: From Discovery to Pathophysiology

68. Chen F, Haigh S, Barman S, Fulton DJ (2012) From form to function: the role of Nox4 in the cardiovascular system. Front Physiol 3:412. https://doi.org/10.3389/fphys.2012.00412 69. Dikalov S (2011) Cross talk between mitochondria and NADPH oxidases. Free Radic Biol Med 51(7):1289–1301. https://doi.org/ 10.1016/j.freeradbiomed.2011.06.033 70. Richter K, Konzack A, Pihlajaniemi T, Heljasvaara R, Kietzmann T (2015) Redox-fibrosis: impact of TGFbeta1 on ROS generators, mediators and functional consequences. Redox Biol 6:344–352. https://doi.org/10.1016/j.redox.2015.08.015 71. von Lohneysen K, Noack D, Hayes P, Friedman JS, Knaus UG (2012) Constitutive NADPH oxidase 4 activity resides in the composition of the B-loop and the penultimate C terminus. J Biol Chem 287(12):8737–8745. https://doi.org/10.1074/jbc.M111. 332494 72. Clempus RE, Sorescu D, Dikalova AE, Pounkova L, Jo P, Sorescu GP, Schmidt HH, Lassegue B, Griendling KK (2007) Nox4 is required for maintenance of the differentiated vascular smooth muscle cell phenotype. Arterioscler Thromb Vasc Biol 27(1): 42–48 73. Guo S, Chen X (2015) The human Nox4: gene, structure, physiological function and pathological significance. J Drug Target 23(10):888–896. https://doi.org/10.3109/1061186X.2015. 1036276 74. Anilkumar N, Weber R, Zhang M, Brewer A, Shah AM (2008) Nox4 and nox2 NADPH oxidases mediate distinct cellular redox signaling responses to agonist stimulation. Arterioscler Thromb Vasc Biol 28(7):1347–1354. https://doi.org/10.1161/ATVBAHA. 108.164277 75. Gorin Y, Kim NH, Feliers D, Bhandari B, Choudhury GG, Abboud HE (2001) Angiotensin II activates Akt/protein kinase B by an arachidonic acid/redox-dependent pathway and independent of phosphoinositide 3-kinase. Faseb J 15(11):1909–1920 76. Park HS, Chun JN, Jung HY, Choi C, Bae YS (2006) Role of NADPH oxidase 4 in lipopolysaccharide-induced proinflammatory responses by human aortic endothelial cells. Cardiovasc Res 72(3): 447–455 77. Mittal M, Roth M, Konig P, Hofmann S, Dony E, Goyal P, Selbitz AC, Schermuly RT, Ghofrani HA, Kwapiszewska G, Kummer W, Klepetko W, Hoda MA, Fink L, Hanze J, Seeger W, Grimminger F, Schmidt HH, Weissmann N (2007) Hypoxiadependent regulation of nonphagocytic NADPH oxidase subunit NOX4 in the pulmonary vasculature. Circ Res 101(3):258–267. https://doi.org/10.1161/CIRCRESAHA.107.148015 78. Pendyala S, Moitra J, Kalari S, Kleeberger SR, Zhao Y, Reddy SP, Garcia JG, Natarajan V (2011) Nrf2 regulates hyperoxia-induced Nox4 expression in human lung endothelium: identification of functional antioxidant response elements on the Nox4 promoter. Free Radic Biol Med 50(12):1749–1759. https://doi.org/10.1016/j. freeradbiomed.2011.03.022 79. Lu X, Murphy TC, Nanes MS, Hart CM (2010) PPAR{gamma} regulates hypoxia-induced Nox4 expression in human pulmonary artery smooth muscle cells through NF-{kappa}B. Am J Physiol Lung Cell Mol Physiol. https://doi.org/10.1152/ajplung.00090. 2010 80. Sedeek M, Callera G, Montezano A, Gutsol A, Heitz F, Szyndralewiez C, Page P, Kennedy CR, Burns KD, Touyz RM, Hebert RL (2010) Critical role of Nox4-based NADPH oxidase in glucose-induced oxidative stress in the kidney: implications in type 2 diabetic nephropathy. Am J Physiol Renal Physiol 299(6): F1348–F1358. https://doi.org/10.1152/ajprenal.00028.2010 81. Hwang J, Ing MH, Salazar A, Lassegue B, Griendling K, Navab M, Sevanian A, Hsiai TK (2003) Pulsatile versus oscillatory shear stress regulates NADPH oxidase subunit expression: implication for native LDL oxidation. Circ Res 93(12):1225–1232

201 82. Cunningham KS, Gotlieb AI (2005) The role of shear stress in the pathogenesis of atherosclerosis. Lab Invest 85(1):9–23. https://doi. org/10.1038/labinvest.3700215 83. Hashikabe Y, Suzuki K, Jojima T, Uchida K, Hattori Y (2006) Aldosterone impairs vascular endothelial cell function. J Cardiovasc Pharmacol 47(4):609–613. https://doi.org/10.1097/01. fjc.0000211738.63207.c3 84. Cucoranu I, Clempus R, Dikalova A, Phelan PJ, Ariyan S, Dikalov S, Sorescu D (2005) NAD(P)H oxidase 4 mediates transforming growth factor-beta1-induced differentiation of cardiac fibroblasts into myofibroblasts. Circ Res 97(9):900–907 85. Lozhkin A, Vendrov AE, Pan H, Wickline SA, Madamanchi NR, Runge MS (2017) NADPH oxidase 4 regulates vascular inflammation in aging and atherosclerosis. J Mol Cell Cardiol 102:10–21. https://doi.org/10.1016/j.yjmcc.2016.12.004 86. Manickam N, Patel M, Griendling KK, Gorin Y, Barnes JL (2014) RhoA/Rho kinase mediates TGF-beta1-induced kidney myofibroblast activation through Poldip2/Nox4-derived reactive oxygen species. Am J Physiol Renal Physiol 307(2):F159–F171. https://doi.org/10.1152/ajprenal.00546.2013 87. Bondi CD, Manickam N, Lee DY, Block K, Gorin Y, Abboud HE, Barnes JL (2010) NAD(P)H oxidase mediates TGF-beta1-induced activation of kidney myofibroblasts. J Am Soc Nephrol 21(1): 93–102. https://doi.org/10.1681/ASN.2009020146 88. Hecker L, Vittal R, Jones T, Jagirdar R, Luckhardt TR, Horowitz JC, Pennathur S, Martinez FJ, Thannickal VJ (2009) NADPH oxidase-4 mediates myofibroblast activation and fibrogenic responses to lung injury. Nat Med 15(9):1077–1081. https://doi. org/10.1038/nm.2005 89. Piera-Velazquez S, Makul A, Jimenez SA (2015) Increased expression of NAPDH oxidase 4 in systemic sclerosis dermal fibroblasts: regulation by transforming growth factor beta. Arthritis Rheumatol 67(10):2749–2758. https://doi.org/10.1002/art.39242 90. Sancho P, Mainez J, Crosas-Molist E, Roncero C, FernandezRodriguez CM, Pinedo F, Huber H, Eferl R, Mikulits W, Fabregat I (2012) NADPH oxidase NOX4 mediates stellate cell activation and hepatocyte cell death during liver fibrosis development. PLoS One 7(9):e45285. https://doi.org/10.1371/journal.pone.0045285 91. Moe KT, Aulia S, Jiang F, Chua YL, Koh TH, Wong MC, Dusting GJ (2006) Differential upregulation of Nox homologues of NADPH oxidase by tumor necrosis factor-alpha in human aortic smooth muscle and embryonic kidney cells. J Cell Mol Med 10(1): 231–239. https://doi.org/10.1111/j.1582-4934.2006.tb00304.x 92. Goettsch C, Goettsch W, Arsov A, Hofbauer LC, Bornstein SR, Morawietz H (2009) Long-term cyclic strain downregulates endothelial Nox4. Antioxid Redox Signal. https://doi.org/10.1089/ARS. 2009.2561 93. Wang F, Wang H, Liu X, Yu H, Huang X, Huang W, Wang G (2021) Neuregulin-1 alleviate oxidative stress and mitigate inflammation by suppressing NOX4 and NLRP3/caspase-1 in myocardial ischaemia-reperfusion injury. J Cell Mol Med 25(3):1783–1795. https://doi.org/10.1111/jcmm.16287 94. Elhadidy MG, Elmasry A, Elsayed HRH, El-Nablaway M, Hamed S, Elalfy MM, Rabei MR (2021) Modulation of COX-2 and NADPH oxidase-4 by alpha-lipoic acid ameliorates busulfaninduced pulmonary injury in rats. Heliyon 7(10):e08171. https:// doi.org/10.1016/j.heliyon.2021.e08171 95. Shi Y, Hou SA (2021) Protective effects of metformin against myocardial ischemiareperfusion injury via AMPKdependent suppression of NOX4. Mol Med Rep 24(4). https://doi.org/10.3892/ mmr.2021.12351 96. Stock CJW, Michaeloudes C, Leoni P, Durham AL, Mumby S, Wells AU, Chung KF, Adcock IM, Renzoni EA, Lindahl GE (2019) Bromodomain and extraterminal (BET) protein inhibition restores redox balance and inhibits myofibroblast activation.

202 Biomed Res Int 2019:1484736. https://doi.org/10.1155/2019/ 1484736 97. Sanders YY, Lyv X, Zhou QJ, Xiang Z, Stanford D, Bodduluri S, Rowe SM, Thannickal VJ (2020) Brd4-p300 inhibition downregulates Nox4 and accelerates lung fibrosis resolution in aged mice. JCI. Insight 5(14). https://doi.org/10.1172/jci.insight. 137127 98. Zhang L, Sheppard OR, Shah AM, Brewer AC (2008) Positive regulation of the NADPH oxidase NOX4 promoter in vascular smooth muscle cells by E2F. Free Radic Biol Med 45(5): 679–685. https://doi.org/10.1016/j.freeradbiomed.2008.05.019 99. Rozycki M, Bialik JF, Speight P, Dan Q, Knudsen TE, Szeto SG, Yuen DA, Szaszi K, Pedersen SF, Kapus A (2016) Myocardinrelated transcription factor regulates Nox4 protein expression: linking cytoskeletal organization to redox statE. J Biol Chem 291(1):227–243. https://doi.org/10.1074/jbc.M115.674606 100. Goettsch C, Goettsch W, Brux M, Haschke C, Brunssen C, Muller G, Bornstein SR, Duerrschmidt N, Wagner AH, Morawietz H (2011) Arterial flow reduces oxidative stress via an antioxidant response element and Oct-1 binding site within the NADPH oxidase 4 promoter in endothelial cells. Basic Res Cardiol 106(4): 551–561. https://doi.org/10.1007/s00395-011-0170-3 101. Kyriazis ID, Hoffman M, Gaignebet L, Lucchese AM, Markopoulou E, Palioura D, Wang C, Bannister TD, Christofidou-Solomidou M, Oka SI, Sadoshima J, Koch WJ, Goldberg IJ, Yang VW, Bialkowska AB, Kararigas G, Drosatos K (2021) KLF5 Is induced by FOXO1 and causes oxidative stress and diabetic cardiomyopathy. Circ Res 128(3):335–357. https:// doi.org/10.1161/CIRCRESAHA.120.316738 102. Manea SA, Todirita A, Raicu M, Manea A (2014) C/EBP transcription factors regulate NADPH oxidase in human aortic smooth muscle cells. J Cell Mol Med 18(7):1467–1477. https://doi.org/10. 1111/jcmm.12289 103. Diebold I, Petry A, Hess J, Gorlach A (2010) The NADPH oxidase subunit NOX4 is a new target gene of the hypoxia-inducible factor-1. Mol Biol Cell 21(12):2087–2096. https://doi.org/10. 1091/mbc.E09-12-1003 104. Katsuyama M, Hirai H, Iwata K, Ibi M, Matsuno K, Matsumoto M, Yabe-Nishimura C (2011) Sp3 transcription factor is crucial for transcriptional activation of the human NOX4 gene. FEBS J 278(6):964–972. https://doi.org/10.1111/j.1742-4658.2011. 08018.x 105. Manea A, Tanase LI, Raicu M, Simionescu M (2010) Jak/STAT signaling pathway regulates Nox1 and Nox4-based NADPH oxidase in human aortic smooth muscle cells. Arterioscler Thromb Vasc Biol 30(1):105–112. https://doi.org/10.1161/ATVBAHA. 109.193896 106. Sun J, Chen J, Li T, Huang P, Li J, Shen M, Gao M, Sun Y, Liang J, Li X, Wang Y, Xiao Y, Shi X, Hu Y, Feng J, Jia H, Liu T, Sun X (2020) ROS production and mitochondrial dysfunction driven by PU.1-regulated NOX4-p22(phox) activation in Aβ-induced retinal pigment epithelial cell injury. Theranostics 10(25):11637–11655. https://doi.org/10.7150/thno.48064 107. Hu F, Xue M, Li Y, Jia YJ, Zheng ZJ, Yang YL, Guan MP, Sun L, Xue YM (2018) Early growth response 1 (Egr1) is a transcriptional activator of NOX4 in oxidative stress of diabetic kidney disease. J Diabetes Res 2018:3405695. https://doi.org/10.1155/2018/ 3405695 108. Liu Z, Gu S, Lu T, Wu K, Li L, Dong C, Zhou Y (2020) IFI6 depletion inhibits esophageal squamous cell carcinoma progression through reactive oxygen species accumulation via mitochondrial dysfunction and endoplasmic reticulum stress. J Exp Clin Cancer Res 39(1):144. https://doi.org/10.1186/s13046-020-01646-3 109. Xu H, Wang Z, Sun Z, Ni Y, Zheng L (2018) GATA4 protects against hyperglycemiainduced endothelial dysfunction by

L. Hecker et al. regulating NOX4 transcription. Mol Med Rep 17(1):1485–1492. https://doi.org/10.3892/mmr.2017.8062 110. Fierro-Fernandez M, Busnadiego O, Sandoval P, Espinosa-Diez C, Blanco-Ruiz E, Rodriguez M, Pian H, Ramos R, Lopez-Cabrera M, Garcia-Bermejo ML, Lamas S (2015) miR-9-5p suppresses pro-fibrogenic transformation of fibroblasts and prevents organ fibrosis by targeting NOX4 and TGFBR2. EMBO Rep 16(10): 1358–1377. https://doi.org/10.15252/embr.201540750 111. Jadhav VS, Krause KH, Singh SK (2014) HIV-1 Tat C modulates NOX2 and NOX4 expressions through miR-17 in a human microglial cell line. J Neurochem 131(6):803–815. https://doi.org/ 10.1111/jnc.12933 112. Gordillo GM, Biswas A, Khanna S, Pan X, Sinha M, Roy S, Sen CK (2014) Dicer knockdown inhibits endothelial cell tumor growth via microRNA 21a-3p targeting of Nox-4. J Biol Chem 289(13):9027–9038. https://doi.org/10.1074/jbc.M113.519264 113. Li J, Chen J, Yang Y, Ding R, Wang M, Gu Z (2021) Ginkgolide A attenuates sepsis-associated kidney damage via upregulating microRNA-25 with NADPH oxidase 4 as the target. Int Immunopharmacol 95:107514. https://doi.org/10.1016/j.intimp. 2021.107514 114. Zhang Y, Song C, Liu J, Bi Y, Li H (2018) Inhibition of miR-25 aggravates diabetic peripheral neuropathy. Neuroreport 29(11): 945–953. https://doi.org/10.1097/WNR.0000000000001058 115. Varga ZV, Kupai K, Szucs G, Gaspar R, Paloczi J, Farago N, Zvara A, Puskas LG, Razga Z, Tiszlavicz L, Bencsik P, Gorbe A, Csonka C, Ferdinandy P, Csont T (2013) MicroRNA-25-dependent up-regulation of NADPH oxidase 4 (NOX4) mediates hypercholesterolemia-induced oxidative/nitrative stress and subsequent dysfunction in the heart. J Mol Cell Cardiol 62:111– 121. https://doi.org/10.1016/j.yjmcc.2013.05.009 116. Cheng Y, Zhou M, Zhou W (2019) MicroRNA-30e regulates TGFbeta-mediated NADPH oxidase 4-dependent oxidative stress by Snai1 in atherosclerosis. Int J Mol Med 43(4):1806–1816. https:// doi.org/10.3892/ijmm.2019.4102 117. Xiao L, Gu Y, Ren G, Chen L, Liu L, Wang X, Gao L (2021) miRNA-146a mimic inhibits NOX4/P38 signalling to ameliorate mouse myocardial ischaemia reperfusion (I/R) injury. Oxid Med Cell Longev 2021:6366254. https://doi.org/10.1155/2021/ 6366254 118. Smolka C, Schlosser D, Koentges C, Tarkhnishvili A, Gorka O, Pfeifer D, Bemtgen X, Asmussen A, Gross O, Diehl P, Moser M, Bode C, Bugger H, Grundmann S, Pankratz F (2021) Cardiomyocyte-specific miR-100 overexpression preserves heart function under pressure overload in mice and diminishes fatty acid uptake as well as ROS production by direct suppression of Nox4 and CD36. FASEB J 35(11):e21956. https://doi.org/10. 1096/fj.202100829RR 119. Kriegel AJ, Baker MA, Liu Y, Liu P, Cowley AW Jr, Liang M (2015) Endogenous microRNAs in human microvascular endothelial cells regulate mRNAs encoded by hypertension-related genes. Hypertension 66(4):793–799. https://doi.org/10.1161/ HYPERTENSIONAHA.115.05645 120. Zhang Y, Liu MW, He Y, Deng N, Chen Y, Huang J, Xie W (2020) Protective effect of resveratrol on estrogen deficiency-induced osteoporosis though attenuating NADPH oxidase 4/nuclear factor kappa B pathway by increasing miR-92b-3p expression. Int J Immunopathol Pharmacol 34:2058738420941762. https://doi.org/ 10.1177/2058738420941762 121. Li P, Fan C, Cai Y, Fang S, Zeng Y, Zhang Y, Lin X, Zhang H, Xue Y, Guan M (2020) Transplantation of brown adipose tissue up-regulates miR-99a to ameliorate liver metabolic disorders in diabetic mice by targeting NOX4. Adipocyte 9(1):57–67. https:// doi.org/10.1080/21623945.2020.1721970 122. Sun M, Hong S, Li W, Wang P, You J, Zhang X, Tang F, Wang P, Zhang C (2016) MiR-99a regulates ROS-mediated invasion and

12

Nox4: From Discovery to Pathophysiology

migration of lung adenocarcinoma cells by targeting NOX4. Oncol Rep 35(5):2755–2766. https://doi.org/10.3892/or.2016.4672 123. Shi Y, Bo Z, Pang G, Qu X, Bao W, Yang L, Ma Y (2017) MiR-99a-5p regulates proliferation, migration and invasion abilities of human oral carcinoma cells by targeting NOX4. Neoplasma 64(5):666–673. https://doi.org/10.4149/neo_2017_503 124. Liu X, Zhong L, Li P, Zhao P (2020) MicroRNA-100 enhances autophagy and suppresses migration and invasion of renal cell carcinoma cells via disruption of NOX4-dependent mTOR pathway. Clin Transl Sci. https://doi.org/10.1111/cts.12798 125. Li X, Wang Y, Cai Z, Zhou Q, Li L, Fu P (2021) Exosomes from human umbilical cord mesenchymal stem cells inhibit ROS production and cell apoptosis in human articular chondrocytes via the miR-100-5p/NOX4 axis. Cell Biol Int 45(10):2096–2106. https:// doi.org/10.1002/cbin.11657 126. Wu QQ, Zheng B, Weng GB, Yang HM, Ren Y, Weng XJ, Zhang SW, Zhu WZ (2019) Downregulated NOX4 underlies a novel inhibitory role of microRNA-137 in prostate cancer. J Cell Biochem 120(6):10215–10227. https://doi.org/10.1002/jcb.28306 127. Wan RJ, Li YH (2018) MicroRNA146a/NAPDH oxidase4 decreases reactive oxygen species generation and inflammation in a diabetic nephropathy model. Mol Med Rep 17(3):4759–4766. https://doi.org/10.3892/mmr.2018.8407 128. Wang HJ, Huang YL, Shih YY, Wu HY, Peng CT, Lo WY (2014) MicroRNA-146a decreases high glucose/thrombin-induced endothelial inflammation by inhibiting NAPDH oxidase 4 expression. Mediators Inflamm 2014:379537. https://doi.org/10.1155/2014/ 379537 129. Milano G, Biemmi V, Lazzarini E, Balbi C, Ciullo A, Bolis S, Ameri P, Di Silvestre D, Mauri P, Barile L, Vassalli G (2020) Intravenous administration of cardiac progenitor cell-derived exosomes protects against doxorubicin/trastuzumab-induced cardiac toxicity. Cardiovasc Res 116(2):383–392. https://doi.org/10. 1093/cvr/cvz108 130. Zhu Y, Ni T, Lin J, Zhang C, Zheng L, Luo M (2019) Long non-coding RNA H19, a negative regulator of microRNA-148b3p, participates in hypoxia stress in human hepatic sinusoidal endothelial cells via NOX4 and eNOS/NO signaling. Biochimie 163:128–136. https://doi.org/10.1016/j.biochi.2019.04.006 131. Wang Y, Zhao X, Wu X, Dai Y, Chen P, Xie L (2016) microRNA182 mediates Sirt1-induced diabetic corneal nerve regeneration. Diabetes 65(7):2020–2031. https://doi.org/10.2337/db15-1283 132. Li ZN, Ge MX, Yuan ZF (2020) MicroRNA-182-5p protects human lens epithelial cells against oxidative stress-induced apoptosis by inhibiting NOX4 and p38 MAPK signalling. BMC Ophthalmol 20(1):233. https://doi.org/10.1186/s12886-02001489-8 133. Tao W, Yu L, Shu S, Liu Y, Zhuang Z, Xu S, Bao X, Gu Y, Cai F, Song W, Xu Y, Zhu X (2021) miR-204-3p/Nox4 mediates memory deficits in a mouse model of Alzheimer’s disease. Mol Ther 29(1): 396–408. https://doi.org/10.1016/j.ymthe.2020.09.006 134. Liu YS, Gu H, Huang TC, Wei XW, Ma W, Liu D, He YW, Luo WT, Huang JT, Zhao D, Jia SS, Wang F, Zhang T, Bai YZ, Wang WL, Yuan ZW (2020) miR-322 treatment rescues cell apoptosis and neural tube defect formation through silencing NADPH oxidase 4. CNS Neurosci Ther 26(9):902–912. https://doi.org/10. 1111/cns.13383 135. Geng Y, Zhao X, Xu J, Zhang X, Hu G, Fu SC, Dai K, Chen X, Patrick YS, Zhang X (2020) Overexpression of mechanical sensitive miR-337-3p alleviates ectopic ossification in rat tendinopathy model via targeting IRS1 and Nox4 of tendon-derived stem cells. J Mol Cell Biol 12(4):305–317. https://doi.org/10.1093/jmcb/ mjz030 136. Zhou T, Li S, Yang L, Xiang D (2021) microRNA-363-3p reduces endothelial cell inflammatory responses in coronary heart disease via inactivation of the NOX4-dependent p38 MAPK axis. Aging

203 (Albany NY) 13(8):11061–11082. https://doi.org/10.18632/aging. 202721 137. Xu Y, Zhang J, Fan L, He X (2018) miR-423-5p suppresses highglucose-induced podocyte injury by targeting Nox4. Biochem Biophys Res Commun 505(2):339–345. https://doi.org/10.1016/j. bbrc.2018.09.067 138. Gu C, Draga D, Zhou C, Su T, Zou C, Gu Q, Lahm T, Zheng Z, Qiu Q (2019) miR-590-3p inhibits pyroptosis in diabetic retinopathy by targeting NLRP1 and inactivating the NOX4 signaling pathway. Invest Ophthalmol Vis Sci 60(13):4215–4223. https://doi.org/10. 1167/iovs.19-27825 139. Shi Q, Lee DY, Feliers D, Abboud HE, Bhat MA, Gorin Y (2020) Interplay between RNA-binding protein HuR and Nox4 as a novel therapeutic target in diabetic kidney disease. Mol Metab 36: 100968. https://doi.org/10.1016/j.molmet.2020.02.011 140. Prior KK, Wittig I, Leisegang MS, Groenendyk J, Weissmann N, Michalak M, Jansen-Durr P, Shah AM, Brandes RP (2016) The endoplasmic reticulum chaperone calnexin is a NADPH oxidase NOX4 interacting protein. J Biol Chem 291(13):7045–7059. https://doi.org/10.1074/jbc.M115.710772 141. Park HS, Jung HY, Park EY, Kim J, Lee WJ, Bae YS (2004) Cutting edge: direct interaction of TLR4 with NAD(P)H oxidase 4 isozyme is essential for lipopolysaccharide-induced production of reactive oxygen species and activation of NF-kappa B. J Immunol 173(6):3589–3593 142. Desai LP, Zhou Y, Estrada AV, Ding Q, Cheng G, Collawn JF, Thannickal VJ (2014) Negative regulation of NADPH oxidase 4 by hydrogen peroxide-inducible clone 5 (Hic-5) protein. J Biol Chem 289(26):18270–18278. https://doi.org/10.1074/jbc.M114.562249 143. Fernandez I, Martin-Garrido A, Zhou DW, Clempus RE, SeidelRogol B, Valdivia A, Lassegue B, Garcia AJ, Griendling KK, San Martin A (2015) Hic-5 mediates TGFbeta-induced adhesion in vascular smooth muscle cells by a Nox4-dependent mechanism. Arterioscler Thromb Vasc Biol 35(5):1198–1206. https://doi.org/ 10.1161/ATVBAHA.114.305185 144. Chen Z, Sun X, Chen Q, Lan T, Huang K, Xiao H, Lin Z, Yang Y, Liu P, Huang H (2020) Connexin32 ameliorates renal fibrosis in diabetic mice by promoting K48-linked NADPH oxidase 4 polyubiquitination and degradation. Br J Pharmacol 177(1): 145–160. https://doi.org/10.1111/bph.14853 145. Gil Lorenzo AF, Costantino VV, Appiolaza ML, Cacciamani V, Benardon ME, Bocanegra V, Valles PG (2015) Heat shock protein 70 and CHIP promote Nox4 ubiquitination and degradation within the Losartan antioxidative effect in proximal tubule cells. Cell Physiol Biochem 36(6):2183–2197. https://doi.org/10.1159/ 000430184 146. Kim HJ, Magesh V, Lee JJ, Kim S, Knaus UG, Lee KJ (2015) Ubiquitin C-terminal hydrolase-L1 increases cancer cell invasion by modulating hydrogen peroxide generated via NADPH oxidase 4. Oncotarget 6(18):16287–16303. https://doi.org/10.18632/ oncotarget.3843 147. Song IK, Kim HJ, Magesh V, Lee KJ (2018) Ubiquitin C-terminal hydrolase-L1 plays a key role in angiogenesis by regulating hydrogen peroxide generated by NADPH oxidase 4. Biochem Biophys Res Commun 495(1):1567–1572. https://doi.org/10.1016/j.bbrc. 2017.11.051 148. Yu B, Liu Z, Fu Y, Wang Y, Zhang L, Cai Z, Yu F, Wang X, Zhou J, Kong W (2017) CYLD deubiquitinates nicotinamide adenine dinucleotide phosphate oxidase 4 contributing to adventitial remodeling. Arterioscler Thromb Vasc Biol 37(9):1698–1709. https://doi.org/10.1161/ATVBAHA.117.309859 149. Liu G, Liu Q, Yan B, Zhu Z, Xu Y (2020) USP7 inhibition alleviates H2O2-Induced injury in chondrocytes via inhibiting NOX4/NLRP3 pathway. Front Pharmacol 11:617270. https://doi. org/10.3389/fphar.2020.617270

204 150. Matsushima S, Kuroda J, Zhai P, Liu T, Ikeda S, Nagarajan N, Oka S, Yokota T, Kinugawa S, Hsu CP, Li H, Tsutsui H, Sadoshima J (2016) Tyrosine kinase FYN negatively regulates NOX4 in cardiac remodeling. J Clin Invest 126(9):3403–3416. https://doi.org/10.1172/JCI85624 151. Xi G, Shen XC, Wai C, Clemmons DR (2013) Recruitment of Nox4 to a plasma membrane scaffold is required for localized reactive oxygen species generation and sustained Src activation in response to insulin-like growth factor-I. J Biol Chem 288(22): 15641–15653. https://doi.org/10.1074/jbc.M113.456046 152. Yamaura M, Mitsushita J, Furuta S, Kiniwa Y, Ashida A, Goto Y, Shang WH, Kubodera M, Kato M, Takata M, Saida T, Kamata T (2009) NADPH oxidase 4 contributes to transformation phenotype of melanoma cells by regulating G2-M cell cycle progression. Cancer Res 69(6):2647–2654. https://doi.org/10.1158/0008-5472. CAN-08-3745 153. Shono T, Yokoyama N, Uesaka T, Kuroda J, Takeya R, Yamasaki T, Amano T, Mizoguchi M, Suzuki SO, Niiro H, Miyamoto K, Akashi K, Iwaki T, Sumimoto H, Sasaki T (2008) Enhanced expression of NADPH oxidase Nox4 in human gliomas and its roles in cell proliferation and survival. Int J Cancer 123(4): 787–792. https://doi.org/10.1002/ijc.23569 154. Zhang C, Lan T, Hou J, Li J, Fang R, Yang Z, Zhang M, Liu J, Liu B (2014) NOX4 promotes non-small cell lung cancer cell proliferation and metastasis through positive feedback regulation of PI3K/ Akt signaling. Oncotarget 5(12):4392–4405. https://doi.org/10. 18632/oncotarget.2025 155. Craige SM, Chen K, Pei Y, Li C, Huang X, Chen C, Shibata R, Sato K, Walsh K, Keaney JF Jr (2011) NADPH oxidase 4 promotes endothelial angiogenesis through endothelial nitric oxide synthase activation. Circulation 124(6):731–740. https://doi.org/10.1161/ CIRCULATIONAHA.111.030775 156. Schroder K, Zhang M, Benkhoff S, Mieth A, Pliquett R, Kosowski J, Kruse C, Luedike P, Michaelis UR, Weissmann N, Dimmeler S, Shah AM, Brandes RP (2012) Nox4 is a protective reactive oxygen species generating vascular NADPH oxidase. Circ Res 110(9):1217–1225. https://doi.org/10.1161/CIRCRESAHA. 112.267054 157. Zhang M, Brewer AC, Schroder K, Santos CX, Grieve DJ, Wang M, Anilkumar N, Yu B, Dong X, Walker SJ, Brandes RP, Shah AM (2010) NADPH oxidase-4 mediates protection against chronic load-induced stress in mouse hearts by enhancing angiogenesis. Proc Natl Acad Sci U S A 107(42):18121–18126. https:// doi.org/10.1073/pnas.1009700107 158. Peshavariya H, Dusting GJ, Jiang F, Halmos LR, Sobey CG, Drummond GR, Selemidis S (2009) NADPH oxidase isoform selective regulation of endothelial cell proliferation and survival. Naunyn Schmiedebergs Arch Pharmacol 380(2):193–204. https:// doi.org/10.1007/s00210-009-0413-0 159. Petry A, Djordjevic T, Weitnauer M, Kietzmann T, Hess J, Gorlach A (2006) NOX2 and NOX4 mediate proliferative response in endothelial cells. Antioxid Redox Signal 8(9-10):1473–1484 160. Green DE, Murphy TC, Kang BY, Kleinhenz JM, Szyndralewiez C, Page P, Sutliff RL, Hart CM (2012) The Nox4 inhibitor GKT137831 attenuates hypoxia-induced pulmonary vascular cell proliferation. Am J Respir Cell Mol Biol 47(5):718–726. https://doi.org/10.1165/rcmb.2011-0418OC 161. Hakami NY, Wong H, Shah MH, Dusting GJ, Jiang F, Peshavariya HM (2015) Smad-independent pathway involved in transforming growth factor beta1-induced Nox4 expression and proliferation of endothelial cells. Naunyn Schmiedebergs Arch Pharmacol 388(3): 319–326. https://doi.org/10.1007/s00210-014-1070-5 162. Patel DN, Bailey SR, Gresham JK, Schuchman DB, Shelhamer JH, Goldstein BJ, Foxwell BM, Stemerman MB, Maranchie JK, Valente AJ, Mummidi S, Chandrasekar B (2006) TLR4-NOX4AP-1 signaling mediates lipopolysaccharide-induced CXCR6

L. Hecker et al. expression in human aortic smooth muscle cells. Biochem Biophys Res Commun 347(4):1113–1120. https://doi.org/10.1016/j.bbrc. 2006.07.015 163. Xu H, Goettsch C, Xia N, Horke S, Morawietz H, Forstermann U, Li H (2008) Differential roles of PKCalpha and PKCepsilon in controlling the gene expression of Nox4 in human endothelial cells. Free Radic Biol Med 44(8):1656–1667. https://doi.org/10. 1016/j.freeradbiomed.2008.01.023 164. Ushio-Fukai M (2006) Redox signaling in angiogenesis: role of NADPH oxidase. Cardiovasc Res 71(2):226–235. https://doi.org/ 10.1016/j.cardiores.2006.04.015 165. Ismail S, Sturrock A, Wu P, Cahill B, Norman K, Huecksteadt T, Sanders K, Kennedy T, Hoidal J (2009) NOX4 mediates hypoxiainduced proliferation of human pulmonary artery smooth muscle cells: the role of autocrine production of transforming growth factor-{beta}1 and insulin-like growth factor binding protein-3. Am J Physiol Lung Cell Mol Physiol 296(3):L489–L499. https:// doi.org/10.1152/ajplung.90488.2008 166. Di Marco E, Gray SP, Kennedy K, Szyndralewiez C, Lyle AN, Lassegue B, Griendling KK, Cooper ME, Schmidt H, JandeleitDahm KAM (2016) NOX4-derived reactive oxygen species limit fibrosis and inhibit proliferation of vascular smooth muscle cells in diabetic atherosclerosis. Free Radic Biol Med 97:556–567. https:// doi.org/10.1016/j.freeradbiomed.2016.07.013 167. Xiao Q, Luo Z, Pepe AE, Margariti A, Zeng L, Xu Q (2009) Embryonic stem cell differentiation into smooth muscle cells is mediated by Nox4-produced H2O2. Am J Physiol Cell Physiol 296(4):C711–C723. https://doi.org/10.1152/ajpcell.00442.2008 168. Kim J, Kim J, Lim HJ, Lee S, Bae YS, Kim J (2021) Nox4-IGF2 axis promotes differentiation of embryoid body cells into derivatives of the three embryonic germ layers. Stem Cell Rev Rep. https://doi.org/10.1007/s12015-021-10303-x 169. Murray TV, Smyrnias I, Shah AM, Brewer AC (2013) NADPH oxidase 4 regulates cardiomyocyte differentiation via redox activation of c-Jun protein and the cis-regulation of GATA-4 gene transcription. J Biol Chem 288(22):15745–15759. https://doi.org/ 10.1074/jbc.M112.439844 170. Martin-Garrido A, Brown DI, Lyle AN, Dikalova A, Seidel-RogolB, Lassegue B, San Martin A, Griendling KK (2011) NADPH oxidase 4 mediates TGF-beta-induced smooth muscle alpha-actin via p38MAPK and serum response factor. Free Radic Biol Med 50(2):354–362. https://doi.org/10.1016/j.freeradbiomed.2010. 11.007 171. Schroder K, Wandzioch K, Helmcke I, Brandes RP (2009) Nox4 acts as a switch between differentiation and proliferation in preadipocytes. Arterioscler Thromb Vasc Biol 29(2):239–245. https://doi.org/10.1161/ATVBAHA.108.174219 172. Lopez-Otin C, Blasco MA, Partridge L, Serrano M, Kroemer G (2013) The hallmarks of aging. Cell 153(6):1194–1217. https://doi. org/10.1016/j.cell.2013.05.039 173. Goy C, Czypiorski P, Altschmied J, Jakob S, Rabanter LL, Brewer AC, Ale-Agha N, Dyballa-Rukes N, Shah AM, Haendeler J (2014) The imbalanced redox status in senescent endothelial cells is due to dysregulated Thioredoxin-1 and NADPH oxidase 4. Exp Gerontol 56:45–52. https://doi.org/10.1016/j.exger.2014.03.005 174. Lee HY, Kim HK, Hoang TH, Yang S, Kim HR, Chae HJ (2020) The correlation of IRE1alpha oxidation with Nox4 activation in aging-associated vascular dysfunction. Redox Biol 37:101727. https://doi.org/10.1016/j.redox.2020.101727 175. Lener B, Koziel R, Pircher H, Hutter E, Greussing R, HerndlerBrandstetter D, Hermann M, Unterluggauer H, Jansen-Durr P (2009) The NADPH oxidase Nox4 restricts the replicative lifespan of human endothelial cells. Biochem J 423(3):363–374. https://doi. org/10.1042/BJ20090666 176. Li TB, Zhang JJ, Liu B, Liu WQ, Wu Y, Xiong XM, Luo XJ, Ma QL, Peng J (2016) Involvement of NADPH oxidases and

12

Nox4: From Discovery to Pathophysiology

non-muscle myosin light chain in senescence of endothelial progenitor cells in hyperlipidemia. Naunyn Schmiedebergs Arch Pharmacol 389(3):289–302. https://doi.org/10.1007/s00210-0151198-y 177. Cao JY, Ling LL, Ni WJ, Guo HL, Yang M (2021) Autophagosome protects proximal tubular cells from aldosteroneinduced senescence through improving oxidative stress. Ren Fail 43(1):556–565. https://doi.org/10.1080/0886022X.2021.1902821 178. Weyemi U, Lagente-Chevallier O, Boufraqech M, Prenois F, Courtin F, Caillou B, Talbot M, Dardalhon M, Al Ghuzlan A, Bidart JM, Schlumberger M, Dupuy C (2012) ROS-generating NADPH oxidase NOX4 is a critical mediator in oncogenic HRas-induced DNA damage and subsequent senescence. Oncogene 31(9):1117–1129. https://doi.org/10.1038/onc.2011.327 179. Kim YY, Jee HJ, Um JH, Kim YM, Bae SS, Yun J (2017) Cooperation between p21 and Akt is required for p53-dependent cellular senescence. Aging Cell 16(5):1094–1103. https://doi.org/ 10.1111/acel.12639 180. Kodama R, Kato M, Furuta S, Ueno S, Zhang Y, Matsuno K, Yabe-Nishimura C, Tanaka E, Kamata T (2013) ROS-generating oxidases Nox1 and Nox4 contribute to oncogenic Ras-induced premature senescence. Genes Cells 18(1):32–41. https://doi.org/ 10.1111/gtc.12015 181. Koziel R, Pircher H, Kratochwil M, Lener B, Hermann M, Dencher NA, Jansen-Durr P (2013) Mitochondrial respiratory chain complex I is inactivated by NADPH oxidase Nox4. Biochem J 452(2): 231–239. https://doi.org/10.1042/BJ20121778 182. Canugovi C, Stevenson MD, Vendrov AE, Hayami T, Robidoux J, Xiao H, Zhang YY, Eitzman DT, Runge MS, Madamanchi NR (2019) Increased mitochondrial NADPH oxidase 4 (NOX4) expression in aging is a causative factor in aortic stiffening. Redox Biol 26:101288. https://doi.org/10.1016/j.redox.2019. 101288 183. Przybylska D, Janiszewska D, Gozdzik A, Bielak-Zmijewska A, Sunderland P, Sikora E, Mosieniak G (2016) NOX4 downregulation leads to senescence of human vascular smooth muscle cells. Oncotarget 7(41):66429–66443. https://doi.org/10. 18632/oncotarget.12079 184. Rezende F, Schurmann C, Schutz S, Harenkamp S, Herrmann E, Seimetz M, Weissmann N, Schroder K (2017) Knock out of the NADPH oxidase Nox4 has no impact on life span in mice. Redox Biol 11:312–314. https://doi.org/10.1016/j.redox.2016.12.012 185. Liu C, Liu L, Yang M, Li B, Yi J, Ai X, Zhang Y, Huang B, Li C, Feng C, Zhou Y (2020) A positive feedback loop between EZH2 and NOX4 regulates nucleus pulposus cell senescence in age-related intervertebral disc degeneration. Cell Div 15:2. https://doi.org/10.1186/s13008-020-0060-x 186. Feng C, Zhang Y, Yang M, Lan M, Liu H, Huang B, Zhou Y (2017) Oxygen-sensing Nox4 generates genotoxic ROS to induce premature senescence of nucleus pulposus cells through MAPK and NF-kappaB pathways. Oxid Med Cell Longev 2017:7426458. https://doi.org/10.1155/2017/7426458 187. Jarman ER, Khambata VS, Cope C, Jones P, Roger J, Ye LY, Duggan N, Head D, Pearce A, Press NJ, Bellenie B, Sohal B, Jarai G (2014) An inhibitor of NADPH oxidase-4 attenuates established pulmonary fibrosis in a Rodent disease model. Am J Respir Cell Mol Biol 50(1):158–169. https://doi.org/10.1165/rcmb.20130174OC 188. Eid AA, Ford BM, Block K, Kasinath BS, Gorin Y, GhoshChoudhury G, Barnes JL, Abboud HE (2010) AMP-activated protein kinase (AMPK) negatively regulates Nox4-dependent activation of p53 and epithelial cell apoptosis in diabetes. J Biol Chem 285(48):37503–37512. https://doi.org/10.1074/jbc.M110.136796 189. Feng L, Hollstein M, Xu Y (2006) Ser46 phosphorylation regulates p53-dependent apoptosis and replicative senescence. Cell Cycle 5(23):2812–2819. https://doi.org/10.4161/cc.5.23.3526

205 190. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87(1):245–313. https://doi.org/10.1152/physrev.00044.2005 191. Vaquero EC, Edderkaoui M, Pandol SJ, Gukovsky I, Gukovskaya AS (2004) Reactive oxygen species produced by NAD(P)H oxidase inhibit apoptosis in pancreatic cancer cells. J Biol Chem 279(33):34643–34654 192. McKallip RJ, Jia W, Schlomer J, Warren JW, Nagarkatti PS, Nagarkatti M (2006) Cannabidiol-induced apoptosis in human leukemia cells: a novel role of cannabidiol in the regulation of p22phox and Nox4 expression. Mol Pharmacol 70(3):897–908. https://doi.org/10.1124/mol.106.023937 193. Choi K, Han YH, Choi C (2007) N-acetyl cysteine and caffeic acid phenethyl ester sensitize astrocytoma cells to Fas-mediated cell death in a redox-dependent manner. Cancer Lett 257(1):79–86. https://doi.org/10.1016/j.canlet.2007.07.006 194. Carmona-Cuenca I, Roncero C, Sancho P, Caja L, Fausto N, Fernandez M, Fabregat I (2008) Upregulation of the NADPH oxidase NOX4 by TGF-beta in hepatocytes is required for its pro-apoptotic activity. J Hepatol 49(6):965–976. https://doi.org/ 10.1016/j.jhep.2008.07.021 195. Reinehr R, Becker S, Eberle A, Grether-Beck S, Haussinger D (2005) Involvement of NADPH oxidase isoforms and Src family kinases in CD95-dependent hepatocyte apoptosis. J Biol Chem 280(29):27179–27194. https://doi.org/10.1074/jbc.M414361200 196. Basuroy S, Bhattacharya S, Leffler CW, Parfenova H (2009) Nox4 NADPH oxidase mediates oxidative stress and apoptosis caused by TNF-alpha in cerebral vascular endothelial cells. Am J Physiol Cell Physiol 296(3):C422–C432. https://doi.org/10.1152/ajpcell.00381. 2008 197. Caja L, Sancho P, Bertran E, Iglesias-Serret D, Gil J, Fabregat I (2009) Overactivation of the MEK/ERK pathway in liver tumor cells confers resistance to TGF-{beta}-induced cell death through impairing up-regulation of the NADPH oxidase NOX4. Cancer Res 69(19):7595–7602. https://doi.org/10.1158/0008-5472.CAN09-1482 198. Simon F, Fernandez R (2009) Early lipopolysaccharide-induced reactive oxygen species production evokes necrotic cell death in human umbilical vein endothelial cells. J Hypertens 27(6): 1202–1216. https://doi.org/10.1097/HJH.0b013e328329e31c 199. Vanden Berghe T, Declercq W, Vandenabeele P (2007) NADPH oxidases: new players in TNF-induced necrotic cell death. Mol Cell 26(6):769–771. https://doi.org/10.1016/j.molcel.2007.06.002 200. Morgan MJ, Liu ZG (2010) Reactive oxygen species in TNFalphainduced signaling and cell death. Mol Cells 30(1):1–12. https://doi. org/10.1007/s10059-010-0105-0 201. Morgan MJ, Kim YS, Liu ZG (2008) TNFalpha and reactive oxygen species in necrotic cell death. Cell Res 18(3):343–349. https://doi.org/10.1038/cr.2008.31 202. Xu Y, Ruan S, Wu X, Chen H, Zheng K, Fu B (2013) Autophagy and apoptosis in tubular cells following unilateral ureteral obstruction are associated with mitochondrial oxidative stress. Int J Mol Med 31(3):628–636. https://doi.org/10.3892/ijmm.2013.1232 203. Wang J, Yi J (2008) Cancer cell killing via ROS: to increase or decrease, that is the question. Cancer Biol Ther 7(12):1875–1884. https://doi.org/10.4161/cbt.7.12.7067 204. Larson-Casey JL, Gu L, Kang J, Dhyani A, Carter AB (2021) NOX4 regulates macrophage apoptosis resistance to induce fibrotic progression. J Biol Chem 297(1):100810. https://doi.org/10.1016/j. jbc.2021.100810 205. Kato K, Logsdon NJ, Shin YJ, Palumbo S, Knox A, Irish JD, Rounseville SP, Rummel SR, Mohamed M, Ahmad K, Trinh JM, Kurundkar D, Knox KS, Thannickal VJ, Hecker L (2020) Impaired myofibroblast dedifferentiation contributes to non-resolving fibrosis in aging. Am J Respir Cell Mol Biol. https://doi.org/10.1165/ rcmb.2019-0092OC

206 206. Vukelic S, Xu Q, Seidel-Rogol B, Faidley EA, Dikalova AE, Hilenski LL, Jorde U, Poole LB, Lassegue B, Zhang G, Griendling KK (2018) NOX4 (NADPH Oxidase 4) and Poldip2 (polymerase delta-interacting protein 2) induce filamentous actin oxidation and promote its interaction with vinculin during integrin-mediated cell adhesion. Arterioscler Thromb Vasc Biol 38(10):2423–2434. https://doi.org/10.1161/ATVBAHA.118.311668 207. Pescatore LA, Bonatto D, Forti FL, Sadok A, Kovacic H, Laurindo FR (2012) Protein disulfide isomerase is required for plateletderived growth factor-induced vascular smooth muscle cell migration, Nox1 NADPH oxidase expression, and RhoGTPase activation. J Biol Chem 287(35):29290–29300. https://doi.org/10.1074/ jbc.M112.394551 208. Coucha M, Abdelsaid M, Li W, Johnson MH, Orfi L, El-Remessy AB, Fagan SC, Ergul A (2016) Nox4 contributes to the hypoxiamediated regulation of actin cytoskeleton in cerebrovascular smooth muscle. Life Sci 163:46–54. https://doi.org/10.1016/j.lfs. 2016.08.018 209. Jafari N, Kim H, Park R, Li L, Jang M, Morris AJ, Park J, Huang C (2017) CRISPR-Cas9 mediated NOX4 knockout inhibits cell proliferation and invasion in HeLa cells. PLoS One 12(1):e0170327. https://doi.org/10.1371/journal.pone.0170327 210. Auer S, Rinnerthaler M, Bischof J, Streubel MK, BreitenbachKoller H, Geisberger R, Aigner E, Cadamuro J, Richter K, Sopjani M, Haschke-Becher E, Felder TK, Breitenbach M (2017) The human NADPH oxidase, Nox4, regulates cytoskeletal organization in two cancer cell lines, HepG2 and SH-SY5Y. Front Oncol 7:111. https://doi.org/10.3389/fonc.2017.00111 211. Rao VR, Stubbs EB Jr (2021) TGF-beta2 Promotes Oxidative Stress in Human Trabecular Meshwork Cells by Selectively Enhancing NADPH Oxidase 4 Expression. Invest Ophthalmol Vis Sci 62(4):4. https://doi.org/10.1167/iovs.62.4.4 212. Lee YM, Kim BJ, Chun YS, So I, Choi H, Kim MS, Park JW (2006) NOX4 as an oxygen sensor to regulate TASK-1 activity. Cell Signal 18(4):499–507 213. Park SJ, Chun YS, Park KS, Kim SJ, Choi SO, Kim HL, Park JW (2009) Identification of subdomains in NADPH oxidase-4 critical for the oxygen-dependent regulation of TASK-1 K+ channels. Am J Physiol Cell Physiol 297(4):C855–C864. https://doi.org/10.1152/ ajpcell.00463.2008 214. Sun QA, Hess DT, Nogueira L, Yong S, Bowles DE, Eu J, Laurita KR, Meissner G, Stamler JS (2011) Oxygen-coupled redox regulation of the skeletal muscle ryanodine receptor-Ca2+ release channel by NADPH oxidase 4. Proc Natl Acad Sci U S A 108(38): 16098–16103. https://doi.org/10.1073/pnas.1109546108 215. He C, Zhu H, Zhang W, Okon I, Wang Q, Li H, Le YZ, Xie Z (2013) 7-Ketocholesterol induces autophagy in vascular smooth muscle cells through Nox4 and Atg4B. Am J Pathol 183(2): 626–637. https://doi.org/10.1016/j.ajpath.2013.04.028 216. Wu RF, Liao C, Hatoum H, Fu G, Ochoa CD, Terada LS (2017) RasGRF couples Nox4-dependent endoplasmic reticulum signaling to Ras. Arterioscler Thromb Vasc Biol 37(1):98–107. https:// doi.org/10.1161/ATVBAHA.116.307922 217. Sciarretta S, Zhai P, Shao D, Zablocki D, Nagarajan N, Terada LS, Volpe M, Sadoshima J (2013) Activation of NADPH oxidase 4 in the endoplasmic reticulum promotes cardiomyocyte autophagy and survival during energy stress through the protein kinase RNAactivated-like endoplasmic reticulum kinase/eukaryotic initiation factor 2alpha/activating transcription factor 4 pathway. Circ Res 113(11):1253–1264. https://doi.org/10.1161/CIRCRESAHA.113. 301787 218. Camargo LL, Harvey AP, Rios FJ, Tsiropoulou S, Da Silva RNO, Cao Z, Graham D, McMaster C, Burchmore RJ, Hartley RC, Bulleid N, Montezano AC, Touyz RM (2018) Vascular Nox (NADPH Oxidase) compartmentalization, protein hyperoxidation, and endoplasmic reticulum stress response in hypertension.

L. Hecker et al. Hypertension 72(1):235–246. https://doi.org/10.1161/ HYPERTENSIONAHA.118.10824 219. Nabeebaccus AA, Zoccarato A, Hafstad AD, Santos CX, Aasum E, Brewer AC, Zhang M, Beretta M, Yin X, West JA, Schroder K, Griffin JL, Eykyn TR, Abel ED, Mayr M, Shah AM (2017) Nox4 reprograms cardiac substrate metabolism via protein O-GlcNAcylation to enhance stress adaptation. JCI Insight 2(24). https://doi.org/10.1172/jci.insight.96184 220. Case AJ, Li S, Basu U, Tian J, Zimmerman MC (2013) Mitochondrial-localized NADPH oxidase 4 is a source of superoxide in angiotensin II-stimulated neurons. Am J Physiol Heart Circ Physiol 305(1):H19–H28. https://doi.org/10.1152/ajpheart.00974. 2012 221. Das R, Xu S, Quan X, Nguyen TT, Kong ID, Chung CH, Lee EY, Cha SK, Park KS (2014) Upregulation of mitochondrial Nox4 mediates TGF-beta-induced apoptosis in cultured mouse podocytes. Am J Physiol Renal Physiol 306(2):F155–F167. https://doi.org/10.1152/ajprenal.00438.2013 222. Mori K, Uchida T, Yoshie T, Mizote Y, Ishikawa F, Katsuyama M, Shibanuma M (2019) A mitochondrial ROS pathway controls matrix metalloproteinase 9 levels and invasive properties in RAS-activated cancer cells. FEBS J 286(3):459–478. https://doi. org/10.1111/febs.14671 223. Li J, Stouffs M, Serrander L, Banfi B, Bettiol E, Charnay Y, Steger K, Krause KH, Jaconi ME (2006) The NADPH oxidase NOX4 drives cardiac differentiation: Role in regulating cardiac transcription factors and MAP kinase activation. Mol Biol Cell 17(9):3978–3988 224. Nadworny AS, Guruju MR, Poor D, Doran RM, Sharma RV, Kotlikoff MI, Davisson RL (2013) Nox2 and Nox4 influence neonatal c-kit(+) cardiac precursor cell status and differentiation. Am J Physiol Heart Circ Physiol 305(6):H829–H842. https://doi. org/10.1152/ajpheart.00761.2012 225. Murray TV, Smyrnias I, Schnelle M, Mistry RK, Zhang M, Beretta M, Martin D, Anilkumar N, de Silva SM, Shah AM, Brewer AC (2015) Redox regulation of cardiomyocyte cell cycling via an ERK1/2 and c-Myc-dependent activation of cyclin D2 transcription. J Mol Cell Cardiol 79:54–68. https://doi.org/10. 1016/j.yjmcc.2014.10.017 226. Byrne JA, Grieve DJ, Bendall JK, Li JM, Gove C, Lambeth JD, Cave AC, Shah AM (2003) Contrasting roles of NADPH oxidase isoforms in pressure-overload versus angiotensin II-induced cardiac hypertrophy. Circ Res 93(9):802–805 227. Matsushima S, Kuroda J, Ago T, Zhai P, Park JY, Xie LH, Tian B, Sadoshima J (2013) Increased oxidative stress in the nucleus caused by Nox4 mediates oxidation of HDAC4 and cardiac hypertrophy. Circ Res 112(4):651–663. https://doi.org/10.1161/ CIRCRESAHA.112.279760 228. Zhang M, Mongue-Din H, Martin D, Catibog N, Smyrnias I, Zhang X, Yu B, Wang M, Brandes RP, Schroder K, Shah AM (2018) Both cardiomyocyte and endothelial cell Nox4 mediate protection against hemodynamic overload-induced remodelling. Cardiovasc Res 114(3):401–408. https://doi.org/10.1093/cvr/ cvx204 229. Brewer AC, Murray TV, Arno M, Zhang M, Anilkumar NP, Mann GE, Shah AM (2011) Nox4 regulates Nrf2 and glutathione redox in cardiomyocytes in vivo. Free Radic Biol Med 51(1):205–215. https://doi.org/10.1016/j.freeradbiomed.2011.04.022 230. Smyrnias I, Zhang X, Zhang M, Murray TV, Brandes RP, Schroder K, Brewer AC, Shah AM (2015) Nicotinamide adenine dinucleotide phosphate oxidase-4-dependent upregulation of nuclear factor erythroid-derived 2-like 2 protects the heart during chronic pressure overload. Hypertension 65(3):547–553. https:// doi.org/10.1161/HYPERTENSIONAHA.114.04208 231. Nabeebaccus A, Hafstad A, Eykyn T, Yin X, Brewer A, Zhang M, Mayr M, Shah A (2015) Cardiac-targeted NADPH oxidase 4 in the

12

Nox4: From Discovery to Pathophysiology

adaptive cardiac remodelling of the murine heart. Lancet 385 (Suppl 1):S73. https://doi.org/10.1016/S0140-6736(15)60388-9 232. Schnelle M, Sawyer I, Anilkumar N, Mohamed BA, Richards DA, Toischer K, Zhang M, Catibog N, Sawyer G, Mongue-Din H, Schroder K, Hasenfuss G, Shah AM (2021) NADPH oxidase-4 promotes eccentric cardiac hypertrophy in response to volume overload. Cardiovasc Res 117(1):178–187. https://doi.org/10. 1093/cvr/cvz331 233. Wang M, Murdoch CE, Brewer AC, Ivetic A, Evans P, Shah AM, Zhang M (2021) Endothelial NADPH oxidase 4 protects against angiotensin II-induced cardiac fibrosis and inflammation. ESC Heart Fail 8(2):1427–1437. https://doi.org/10.1002/ehf2.13228 234. Yu Q, Lee CF, Wang W, Karamanlidis G, Kuroda J, Matsushima S, Sadoshima J, Tian R (2014) Elimination of NADPH oxidase activity promotes reductive stress and sensitizes the heart to ischemic injury. J Am Heart Assoc 3(1):e000555. https://doi.org/10.1161/JAHA.113.000555 235. Mongue-Din H, Patel AS, Looi YH, Grieve DJ, Anilkumar N, Sirker A, Dong X, Brewer AC, Zhang M, Smith A, Shah AM (2017) NADPH oxidase-4 driven cardiac macrophage polarization protects against myocardial infarction-induced remodeling. JACC Basic Transl Sci 2(6):688–698. https://doi.org/10.1016/j.jacbts. 2017.06.006 236. Stevenson MD, Canugovi C, Vendrov AE, Hayami T, Bowles DE, Krause KH, Madamanchi NR, Runge MS (2019) NADPH oxidase 4 regulates inflammation in ischemic heart failure: role of soluble epoxide hydrolase. Antioxid Redox Signal 31(1):39–58. https:// doi.org/10.1089/ars.2018.7548 237. Sorescu D, Weiss D, Lassègue B, Clempus RE, Szöcs K, Sorescu GP, Valppu L, Quinn MT, Lambeth JD, Vega JD, Taylor WR, Griendling KK (2002) Superoxide production and expression of nox family proteins in human atherosclerosis. Circulation 105(12): 1429–1435 238. Chamseddine AH, Miller FJ Jr (2003) gp91phox contributes to NADPH oxidase activity in aortic fibroblasts but not smooth muscle cells. Am J Physiol Heart Circ Physiol 285(6):H2284–H2289 239. Perrotta I, Sciangula A, Perrotta E, Donato G, Cassese M (2011) Ultrastructural analysis and electron microscopic localization of Nox4 in healthy and atherosclerotic human aorta. Ultrastruct Pathol 35(1):1–6. https://doi.org/10.3109/01913123.2010.510261 240. Szöcs K, Lassègue B, Sorescu D, Hilenski LL, Valppu L, Couse TL, Wilcox JN, Quinn MT, Lambeth JD, Griendling KK (2002) Upregulation of Nox-based NAD(P)H oxidases in restenosis after carotid injury. Arterioscler Thromb Vasc Biol 22(1):21–27 241. Xu S, Chamseddine AH, Carrell S, Miller FJ Jr (2014) Nox4 NADPH oxidase contributes to smooth muscle cell phenotypes associated with unstable atherosclerotic plaques. Redox Biol 2: 642–650. https://doi.org/10.1016/j.redox.2014.04.004 242. Schurmann C, Rezende F, Kruse C, Yasar Y, Lowe O, Fork C, van de Sluis B, Bremer R, Weissmann N, Shah AM, Jo H, Brandes RP, Schroder K (2015) The NADPH oxidase Nox4 has antiatherosclerotic functions. Eur Heart J 36(48):3447–3456. https:// doi.org/10.1093/eurheartj/ehv460 243. Langbein H, Brunssen C, Hofmann A, Cimalla P, Brux M, Bornstein SR, Deussen A, Koch E, Morawietz H (2016) NADPH oxidase 4 protects against development of endothelial dysfunction and atherosclerosis in LDL receptor deficient mice. Eur Heart J 37(22):1753–1761. https://doi.org/10.1093/eurheartj/ehv564 244. Craige SM, Kant S, Reif M, Chen K, Pei Y, Angoff R, Sugamura K, Fitzgibbons T, Keaney JF Jr (2015) Endothelial NADPH oxidase 4 protects ApoE-/- mice from atherosclerotic lesions. Free Radic Biol Med 89:1–7. https://doi.org/10.1016/j. freeradbiomed.2015.07.004 245. Tong X, Khandelwal AR, Wu X, Xu Z, Yu W, Chen C, Zhao W, Yang J, Qin Z, Weisbrod RM, Seta F, Ago T, Lee KS, Hammock BD, Sadoshima J, Cohen RA, Zeng C (2016) Pro-atherogenic role

207 of smooth muscle Nox4-based NADPH oxidase. J Mol Cell Cardiol 92:30–40. https://doi.org/10.1016/j.yjmcc.2016.01.020 246. Yu W, Xiao L, Que Y, Li S, Chen L, Hu P, Xiong R, Seta F, Chen H (1866) Tong X (2020) Smooth muscle NADPH oxidase 4 promotes angiotensin II-induced aortic aneurysm and atherosclerosis by regulating osteopontin. Biochim Biophys Acta Mol Basis Dis 12:165912. https://doi.org/10.1016/j.bbadis.2020.165912 247. Kim J, Seo M, Kim SK, Bae YS (2016) Flagellin-induced NADPH oxidase 4 activation is involved in atherosclerosis. Sci Rep 6: 25437. https://doi.org/10.1038/srep25437 248. Vendrov AE, Vendrov KC, Smith A, Yuan J, Sumida A, Robidoux J, Runge MS, Madamanchi NR (2015) NOX4 NADPH oxidase-dependent mitochondrial oxidative stress in agingassociated cardiovascular disease. Antioxid Redox Signal 23(18): 1389–1409. https://doi.org/10.1089/ars.2014.6221 249. Gray SP, Di Marco E, Kennedy K, Chew P, Okabe J, El-Osta A, Calkin AC, Biessen EA, Touyz RM, Cooper ME, Schmidt HH, Jandeleit-Dahm KA (2016) Reactive oxygen species can provide atheroprotection via NOX4-dependent inhibition of inflammation and vascular remodeling. Arterioscler Thromb Vasc Biol 36(2): 295–307. https://doi.org/10.1161/ATVBAHA.115.307012 250. Di Marco E, Gray SP, Chew P, Kennedy K, Cooper ME, Schmidt HH, Jandeleit-Dahm KA (2016) Differential effects of NOX4 and NOX1 on immune cell-mediated inflammation in the aortic sinus of diabetic ApoE-/- mice. Clin Sci 130(15):1363–1374. https://doi. org/10.1042/CS20160249 251. Gray SP, Jha JC, Kennedy K, van Bommel E, Chew P, Szyndralewiez C, Touyz RM, Schmidt H, Cooper ME, JandeleitDahm KAM (2017) Combined NOX1/4 inhibition with GKT137831 in mice provides dose-dependent reno- and atheroprotection even in established micro- and macrovascular disease. Diabetologia 60(5):927–937. https://doi.org/10.1007/ s00125-017-4215-5 252. Tong X, Hou X, Jourd'heuil D, Weisbrod RM, Cohen RA (2010) Upregulation of Nox4 by TGF{beta}1 oxidizes SERCA and inhibits NO in arterial smooth muscle of the prediabetic Zucker rat. Circ Res 107(8):975–983. https://doi.org/10.1161/ CIRCRESAHA.110.221242 253. Hu P, Wu X, Khandelwal AR, Yu W, Xu Z, Chen L, Yang J, Weisbrod RM, Lee KSS, Seta F, Hammock BD, Cohen RA, Zeng C, Tong X (2017) Endothelial Nox4-based NADPH oxidase regulates atherosclerosis via soluble epoxide hydrolase. Biochim Biophys Acta Mol Basis Dis 1863(6):1382–1391. https://doi.org/ 10.1016/j.bbadis.2017.02.004 254. Yu W, Li S, Wu H, Hu P, Chen L, Zeng C, Tong X (2021) Endothelial Nox4 dysfunction aggravates atherosclerosis by inducing endoplasmic reticulum stress and soluble epoxide hydrolase. Free Radic Biol Med 164:44–57. https://doi.org/10.1016/j. freeradbiomed.2020.12.450 255. Jimenez-Altayo F, Meirelles T, Crosas-Molist E, Sorolla MA, Del Blanco DG, Lopez-Luque J, Mas-Stachurska A, Siegert AM, Bonorino F, Barbera L, Garcia C, Condom E, Sitges M, Rodriguez-Pascual F, Laurindo F, Schroder K, Ros J, Fabregat I, Egea G (2018) Redox stress in Marfan syndrome: Dissecting the role of the NADPH oxidase NOX4 in aortic aneurysm. Free Radic Biol Med 118:44–58. https://doi.org/10.1016/j.freeradbiomed. 2018.02.023 256. Siu KL, Li Q, Zhang Y, Guo J, Youn JY, Du J, Cai H (2017) NOX isoforms in the development of abdominal aortic aneurysm. Redox Biol 11:118–125. https://doi.org/10.1016/j.redox.2016.11.002 257. Griendling KK, Minieri CA, Ollerenshaw JD, Alexander RW (1994) Angiotensin II stimulates NADH and NADPH oxidase activity in cultured vascular smooth muscle cells. Circ Res 74: 1141–1148 258. Akasaki T, Ohya Y, Kuroda J, Eto K, Abe I, Sumimoto H, Iida M (2006) Increased expression of gp91phox homologues of NAD

208 (P)H oxidase in the aortic media during chronic hypertension: involvement of the renin-angiotensin system. Hypertens Res 29(10):813–820. https://doi.org/10.1291/hypres.29.813 259. Peterson JR, Burmeister MA, Tian X, Zhou Y, Guruju MR, Stupinski JA, Sharma RV, Davisson RL (2009) Genetic silencing of Nox2 and Nox4 reveals differential roles of these NADPH oxidase homologues in the vasopressor and dipsogenic effects of brain angiotensin II. Hypertension 54(5):1106–1114. https://doi. org/10.1161/HYPERTENSIONAHA.109.140087 260. Xue B, Beltz TG, Johnson RF, Guo F, Hay M, Johnson AK (2012) PVN adenovirus-siRNA injections silencing either NOX2 or NOX4 attenuate aldosterone/NaCl-induced hypertension in mice. Am J Physiol Heart Circ Physiol 302(3):H733–H741. https://doi. org/10.1152/ajpheart.00873.2011 261. Bouabout G, Ayme-Dietrich E, Jacob H, Champy MF, Birling MC, Pavlovic G, Madeira L, Fertak LE, Petit-Demouliere B, Sorg T, Herault Y, Mudgett J, Monassier L (2018) Nox4 genetic inhibition in experimental hypertension and metabolic syndrome. Arch Cardiovasc Dis 111(1):41–52. https://doi.org/10.1016/j.acvd. 2017.03.011 262. Ray R, Murdoch CE, Wang M, Santos CX, Zhang M, AlomRuiz S, Anilkumar N, Ouattara A, Cave AC, Walker SJ, Grieve DJ, Charles RL, Eaton P, Brewer AC, Shah AM (2011) Endothelial Nox4 NADPH oxidase enhances vasodilatation and reduces blood pressure in vivo. Arterioscler Thromb Vasc Biol 31(6):1368–1376. https://doi.org/10.1161/ATVBAHA.110.219238 263. Liu Y, Bubolz AH, Mendoza S, Zhang DX, Gutterman DD (2011) H2O2 is the transferrable factor mediating flow-induced dilation in human coronary arterioles. Circ Res 108(5):566–573. https://doi. org/10.1161/CIRCRESAHA.110.237636 264. Babelova A, Avaniadi D, Jung O, Fork C, Beckmann J, Kosowski J, Weissmann N, Anilkumar N, Shah AM, Schaefer L, Schroder K, Brandes RP (2012) Role of Nox4 in murine models of kidney disease. Free Radic Biol Med 53(4):842–853. https://doi. org/10.1016/j.freeradbiomed.2012.06.027 265. Cowley AW Jr, Yang C, Zheleznova NN, Staruschenko A, Kurth T, Rein L, Kumar V, Sadovnikov K, Dayton A, Hoffman M, Ryan RP, Skelton MM, Salehpour F, Ranji M, Geurts A (2016) Evidence of the importance of Nox4 in production of hypertension in Dahl Salt-sensitive rats. Hypertension 67(2): 440–450. https://doi.org/10.1161/HYPERTENSIONAHA.115. 06280 266. Kumar V, Kurth T, Zheleznova NN, Yang C, Cowley AW Jr (2020) NOX4/H2O2/mTORC1 pathway in salt-induced hypertension and kidney injury. Hypertension 76(1):133–143. https://doi. org/10.1161/HYPERTENSIONAHA.120.15058 267. Pavlov TS, Palygin O, Isaeva E, Levchenko V, Khedr S, Blass G, Ilatovskaya DV, Cowley AW Jr, Staruschenko A (2020) NOX4dependent regulation of ENaC in hypertension and diabetic kidney disease. FASEB J 34(10):13396–13408. https://doi.org/10.1096/fj. 202000966RR 268. Rezende F, Malacarne PF, Muller N, Rathkolb B, Hrabe de Angelis M, Schroder K (2021) Nox4 maintains blood pressure during low sodium diet. Antioxidants (Basel) 10(7). https://doi. org/10.3390/antiox10071103 269. Kraja AT, Cook JP, Warren HR, Surendran P, Liu C, Evangelou E, Manning AK, Grarup N, Drenos F, Sim X, Smith AV, Amin N, Blakemore AIF, Bork-Jensen J, Brandslund I, Farmaki AE, Fava C, Ferreira T, Herzig KH, Giri A, Giulianini F, Grove ML, Guo X, Harris SE, Have CT, Havulinna AS, Zhang H, Jorgensen ME, Karajamaki A, Kooperberg C, Linneberg A, Little L, Liu Y, Bonnycastle LL, Lu Y, Magi R, Mahajan A, Malerba G, Marioni RE, Mei H, Menni C, Morrison AC, Padmanabhan S, Palmas W, Poveda A, Rauramaa R, Rayner NW, Riaz M, Rice K, Richard MA, Smith JA, Southam L, Stancakova A, Stirrups KE, Tragante V, Tuomi T, Tzoulaki I, Varga TV, Weiss S, Yiorkas

L. Hecker et al. AM, Young R, Zhang W, Barnes MR, Cabrera CP, Gao H, Boehnke M, Boerwinkle E, Chambers JC, Connell JM, Christensen CK, de Boer RA, Deary IJ, Dedoussis G, Deloukas P, Dominiczak AF, Dorr M, Joehanes R, Edwards TL, Esko T, Fornage M, Franceschini N, Franks PW, Gambaro G, Groop L, Hallmans G, Hansen T, Hayward C, Heikki O, Ingelsson E, Tuomilehto J, Jarvelin MR, Kardia SLR, Karpe F, Kooner JS, Lakka TA, Langenberg C, Lind L, Loos RJF, Laakso M, McCarthy MI, Melander O, Mohlke KL, Morris AP, Palmer CNA, Pedersen O, Polasek O, Poulter NR, Province MA, Psaty BM, Ridker PM, Rotter JI, Rudan I, Salomaa V, Samani NJ, Sever PJ, Skaaby T, Stafford JM, Starr JM, van der Harst P, van der Meer P, Understanding Society Scientific G, van Duijn CM, Vergnaud AC, Gudnason V, Wareham NJ, Wilson JG, Willer CJ, Witte DR, Zeggini E, Saleheen D, Butterworth AS, Danesh J, Asselbergs FW, Wain LV, Ehret GB, Chasman DI, Caulfield MJ, Elliott P, Lindgren CM, Levy D, Newton-Cheh C, Munroe PB, Howson JMM, CHARGE EXOME BP, CHD Exome, Exome BP, GoT2D:T2DGenes Consortia, The UK Biobank Cardio-Metabolic Traits Consortium Blood Pressure Working Group (2017) New blood pressure-associated loci identified in meta-analyses of 475000 individuals. Circ Cardiovasc Genet 10(5). https://doi.org/ 10.1161/CIRCGENETICS.117.001778 270. Li J, Fan LM, George VT, Brooks G (2007) Nox2 regulates endothelial cell cycle arrest and apoptosis via p21cip1 and p53. Free Radic Biol Med 43(6):976–986. PMID: 17697942 271. Ago T, Kitazono T, Ooboshi H, Iyama T, Han YH, Takada J, Wakisaka M, Ibayashi S, Utsumi H, Iida M (2004) Nox4 as the major catalytic component of an endothelial NAD(P)H oxidase. Circulation 109(2):227–233. https://doi.org/10.1161/01.CIR. 0000105680.92873.70 272. Bernard K, Hecker L, Luckhardt TR, Cheng G, Thannickal VJ (2014) NADPH oxidases in lung health and disease. Antioxid Redox Signal 20(17):2838–2853. https://doi.org/10.1089/ars. 2013.5608 273. Jiang J, Huang K, Xu S, Garcia JGN, Wang C, Cai H (2020) Targeting NOX4 alleviates sepsis-induced acute lung injury via attenuation of redox-sensitive activation of CaMKII/ERK1/2/ MLCK and endothelial cell barrier dysfunction. Redox Biol 36: 101638. https://doi.org/10.1016/j.redox.2020.101638 274. Al Ghouleh I, Magder S (2008) Nicotinamide adenine dinucleotide phosphate (reduced form) oxidase is important for LPS-induced endothelial cell activation. Shock 29(5):553–559. https://doi.org/ 10.1097/SHK.0b013e318157ebc8 275. Qi D, Wang D, Zhang C, Tang X, He J, Zhao Y, Deng W, Deng X (2017) Vaspin protects against LPSinduced ARDS by inhibiting inflammation, apoptosis and reactive oxygen species generation in pulmonary endothelial cells via the Akt/GSK3beta pathway. Int J Mol Med 40(6):1803–1817. https://doi.org/10.3892/ijmm.2017. 3176 276. Jedrychowski WA, Maugeri U, Adamczyk B (1986) Effect of smoking on serum immunoglobulins and cellular blood constituents in healthy individuals. G Ital Med Lav 8(2):53–56 277. Fu P, Mohan V, Mansoor S, Tiruppathi C, Sadikot RT, Natarajan V (2013) Role of nicotinamide adenine dinucleotide phosphatereduced oxidase proteins in Pseudomonas aeruginosa-induced lung inflammation and permeability. Am J Respir Cell Mol Biol 48(4):477–488. https://doi.org/10.1165/rcmb.2012-0242OC 278. Peters DM, Vadasz I, Wujak L, Wygrecka M, Olschewski A, Becker C, Herold S, Papp R, Mayer K, Rummel S, Brandes RP, Gunther A, Waldegger S, Eickelberg O, Seeger W, Morty RE (2014) TGF-beta directs trafficking of the epithelial sodium channel ENaC which has implications for ion and fluid transport in acute lung injury. Proc Natl Acad Sci U S A 111(3):E374–E383. https://doi.org/10.1073/pnas.1306798111

12

Nox4: From Discovery to Pathophysiology

279. Xiang L, Lu S, Mittwede PN, Clemmer JS, Hester RL (2014) Inhibition of NADPH oxidase prevents acute lung injury in obese rats following severe trauma. Am J Physiol Heart Circ Physiol 306(5):H684–H689. https://doi.org/10.1152/ajpheart.00868.2013 280. Yu WY, Li L, Wu F, Zhang HH, Fang J, Zhong YS, Yu CH (2020) Moslea Herba flavonoids alleviated influenza A virus-induced pulmonary endothelial barrier disruption via suppressing NOX4/ NF-kappaB/MLCK pathway. J Ethnopharmacol 253:112641. https://doi.org/10.1016/j.jep.2020.112641 281. Jin HZ, Yang XJ, Zhao KL, Mei FC, Zhou Y, You YD, Wang WX (2019) Apocynin alleviates lung injury by suppressing NLRP3 inflammasome activation and NF-kappaB signaling in acute pancreatitis. Int Immunopharmacol 75:105821. https://doi.org/10. 1016/j.intimp.2019.105821 282. Cui Y, Wang Y, Li G, Ma W, Zhou XS, Wang J, Liu B (2018) The Nox1/Nox4 inhibitor attenuates acute lung injury induced by ischemia-reperfusion in mice. PLoS One 13(12):e0209444. https://doi.org/10.1371/journal.pone.0209444 283. Lee I, Dodia C, Chatterjee S, Feinstein SI, Fisher AB (2014) Protection against LPS-induced acute lung injury by a mechanism-based inhibitor of NADPH oxidase (type 2). Am J Physiol Lung Cell Mol Physiol 306(7):L635–L644. https://doi. org/10.1152/ajplung.00374.2013 284. Yeligar SM, Harris FL, Hart CM, Brown LA (2012) Ethanol induces oxidative stress in alveolar macrophages via upregulation of NADPH oxidases. J Immunol 188(8):3648–3657. https://doi. org/10.4049/jimmunol.1101278 285. Hong Y, Woo S, Kim Y, Lee JJ, Hong JY (2020) Plasma concentrations of NOX4 are predictive of successful liberation from mechanical ventilation and 28-day mortality in intubated patients. Ann Transl Med 8(21):1376. https://doi.org/10.21037/ atm-20-4252 286. Collum SD, Amione-Guerra J, Cruz-Solbes AS, DiFrancesco A, Hernandez AM, Hanmandlu A, Youker K, Guha A, KarmoutyQuintana H (2017) Pulmonary hypertension associated with idiopathic pulmonary fibrosis: current and future perspectives. Can Respir J 2017:1430350. https://doi.org/10.1155/2017/1430350 287. Hood KY, Montezano AC, Harvey AP, Nilsen M, MacLean MR, Touyz RM (2016) Nicotinamide adenine dinucleotide phosphate oxidase-mediated redox signaling and vascular remodeling by 16alpha-hydroxyestrone in human pulmonary artery cells: implications in pulmonary arterial hypertension. Hypertension 68(3):796–808. https://doi.org/10.1161/HYPERTENSIONAHA. 116.07668 288. Pache JC, Carnesecchi S, Deffert C, Donati Y, Herrmann FR, Barazzone-Argiroffo C, Krause KH (2011) NOX-4 is expressed in thickened pulmonary arteries in idiopathic pulmonary fibrosis. Nat Med 17(1):31–32.; author reply 32-33. https://doi.org/10.1038/ nm0111-31 289. Lu X, Bijli KM, Ramirez A, Murphy TC, Kleinhenz J, Hart CM (2013) Hypoxia downregulates PPARgamma via an ERK1/2-NFkappaB-Nox4-dependent mechanism in human pulmonary artery smooth muscle cells. Free Radic Biol Med 63:151–160. https://doi. org/10.1016/j.freeradbiomed.2013.05.013 290. Li S, Tabar SS, Malec V, Eul BG, Klepetko W, Weissmann N, Grimminger F, Seeger W, Rose F, Hanze J (2008) NOX4 regulates ROS levels under normoxic and hypoxic conditions, triggers proliferation, and inhibits apoptosis in pulmonary artery adventitial fibroblasts. Antioxid Redox Signal 10(10):1687–1698. https://doi. org/10.1089/ars.2008.2035 291. Chen F, Barman S, Yu Y, Haigh S, Wang Y, Black SM, Rafikov R, Dou H, Bagi Z, Han W, Su Y, Fulton DJ (2014) Caveolin-1 is a negative regulator of NADPH oxidase-derived reactive oxygen species. Free Radic Biol Med 73:201–213. https://doi.org/10. 1016/j.freeradbiomed.2014.04.029

209 292. Mittal M, Gu XQ, Pak O, Pamenter ME, Haag D, Fuchs DB, Schermuly RT, Ghofrani HA, Brandes RP, Seeger W, Grimminger F, Haddad GG, Weissmann N (2012) Hypoxia induces Kv channel current inhibition by increased NADPH oxidase-derived reactive oxygen species. Free Radic Biol Med 52(6):1033–1042. https://doi.org/10.1016/j.freeradbiomed.2011. 12.004 293. Barman SA, Chen F, Su Y, Dimitropoulou C, Wang Y, Catravas JD, Han W, Orfi L, Szantai-Kis C, Keri G, Szabadkai I, Barabutis N, Rafikova O, Rafikov R, Black SM, Jonigk D, Giannis A, Asmis R, Stepp DW, Ramesh G, Fulton DJ (2014) NADPH oxidase 4 is expressed in pulmonary artery adventitia and contributes to hypertensive vascular remodeling. Arterioscler Thromb Vasc Biol 34(8):1704–1715. https://doi.org/10.1161/ ATVBAHA.114.303848 294. Goncharov DA, Kudryashova TV, Ziai H, Ihida-Stansbury K, DeLisser H, Krymskaya VP, Tuder RM, Kawut SM, Goncharova EA (2014) Mammalian target of rapamycin complex 2 (mTORC2) coordinates pulmonary artery smooth muscle cell metabolism, proliferation, and survival in pulmonary arterial hypertension. Circulation 129(8):864–874. https://doi.org/10.1161/ CIRCULATIONAHA.113.004581 295. Nisbet RE, Graves AS, Kleinhenz DJ, Rupnow HL, Reed AL, Fan TH, Mitchell PO, Sutliff RL, Hart CM (2009) The role of NADPH oxidase in chronic intermittent hypoxia-induced pulmonary hypertension in mice. Am J Respir Cell Mol Biol 40(5):601–609. https:// doi.org/10.1165/2008-0145OC 296. Nisbet RE, Bland JM, Kleinhenz DJ, Mitchell PO, Walp ER, Sutliff RL, Hart CM (2010) Rosiglitazone attenuates chronic hypoxia-induced pulmonary hypertension in a mouse model. Am J Respir Cell Mol Biol 42(4):482–490. https://doi.org/10.1165/ rcmb.2008-0132OC 297. Veith C, Kraut S, Wilhelm J, Sommer N, Quanz K, Seeger W, Brandes RP, Weissmann N, Schroder K (2016) NADPH oxidase 4 is not involved in hypoxia-induced pulmonary hypertension. Pulm Circ 6(3):397–400. https://doi.org/10.1086/687756 298. Kirkham PA, Barnes PJ (2013) Oxidative stress in COPD. Chest 144(1):266–273. https://doi.org/10.1378/chest.12-2664 299. Ichinose M, Sugiura H, Yamagata S, Koarai A, Shirato K (2000) Increase in reactive nitrogen species production in chronic obstructive pulmonary disease airways. Am J Respir Crit Care Med 162(2 Pt 1):701–706. https://doi.org/10.1164/ajrccm.162.2. 9908132 300. Montuschi P, Collins JV, Ciabattoni G, Lazzeri N, Corradi M, Kharitonov SA, Barnes PJ (2000) Exhaled 8-isoprostane as an in vivo biomarker of lung oxidative stress in patients with COPD and healthy smokers. Am J Respir Crit Care Med 162(3 Pt 1): 1175–1177. https://doi.org/10.1164/ajrccm.162.3.2001063 301. Rahman I, van Schadewijk AA, Crowther AJ, Hiemstra PS, Stolk J, MacNee W, De Boer WI (2002) 4-Hydroxy-2-nonenal, a specific lipid peroxidation product, is elevated in lungs of patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 166(4):490–495. https://doi.org/10.1164/rccm.2110101 302. Kluchova Z, Petrasova D, Joppa P, Dorkova Z, Tkacova R (2007) The association between oxidative stress and obstructive lung impairment in patients with COPD. Physiol Res 56(1):51–56. https://doi.org/10.33549/physiolres.930884 303. Hollins F, Sutcliffe A, Gomez E, Berair R, Russell R, Szyndralewiez C, Saunders R, Brightling C (2016) Airway smooth muscle NOX4 is upregulated and modulates ROS generation in COPD. Respir Res 17(1):84. https://doi.org/10.1186/s12931-0160403-y 304. Liu X, Hao B, Ma A, He J, Liu X, Chen J (2016) The expression of NOX4 in smooth muscles of small airway correlates with the disease severity of COPD. Biomed Res Int 2016:2891810. https://doi.org/10.1155/2016/2891810

210 305. Hao B, Sun R, Guo X, Zhang L, Cui J, Zhou Y, Hong W, Zhang Y, He J, Liu X, Li B, Ran P, Chen J (2021) NOX4-derived ROS promotes collagen I deposition in bronchial smooth muscle cells by activating noncanonical p38MAPK/Akt-mediated TGF-beta signaling. Oxid Med Cell Longev 2021:6668971. https://doi.org/10. 1155/2021/6668971 306. Guo X, Fan Y, Cui J, Hao B, Zhu L, Sun X, He J, Yang J, Dong J, Wang Y, Liu X, Chen J (2018) NOX4 expression and distal arteriolar remodeling correlate with pulmonary hypertension in COPD. BMC Pulm Med 18(1):111. https://doi.org/10.1186/ s12890-018-0680-y 307. Milara J, Peiro T, Serrano A, Guijarro R, Zaragoza C, Tenor H, Cortijo J (2014) Roflumilast N-oxide inhibits bronchial epithelial to mesenchymal transition induced by cigarette smoke in smokers with COPD. Pulm Pharmacol Ther 28(2):138–148. https://doi.org/ 10.1016/j.pupt.2014.02.001 308. Hernandez-Saavedra D, Sanders L, Perez MJ, Kosmider B, Smith LP, Mitchell JD, Yoshida T, Tuder RM (2017) RTP801 amplifies nicotinamide adenine dinucleotide phosphate oxidase-4-dependent oxidative stress induced by cigarette smoke. Am J Respir Cell Mol Biol 56(1):62–73. https://doi.org/10.1165/rcmb.2016-0144OC 309. Mehal WZ, Iredale J, Friedman SL (2011) Scraping fibrosis: expressway to the core of fibrosis. Nat Med 17(5):552–553. https://doi.org/10.1038/nm0511-552 310. Liepelt A, Tacke F (2015) Healing the scars of life-targeting redox imbalance in fibrotic disorders of the elderly. Ann Transl Med 3 (Suppl 1):S13. https://doi.org/10.3978/j.issn.2305-5839.2015. 03.34 311. Liu RM, Desai LP (2015) Reciprocal regulation of TGF-beta and reactive oxygen species: a perverse cycle for fibrosis. Redox Biol 6:565–577. https://doi.org/10.1016/j.redox.2015.09.009 312. Amara N, Goven D, Prost F, Muloway R, Crestani B, Boczkowski J (2010) NOX4/NADPH oxidase expression is increased in pulmonary fibroblasts from patients with idiopathic pulmonary fibrosis and mediates TGFbeta1-induced fibroblast differentiation into myofibroblasts. Thorax 65(8):733–738. https://doi.org/10.1136/ thx.2009.113456 313. Wermuth PJ, Mendoza FA, Jimenez SA (2019) Abrogation of transforming growth factor-beta-induced tissue fibrosis in mice with a global genetic deletion of Nox4. Lab Invest 99(4): 470–482. https://doi.org/10.1038/s41374-018-0161-1 314. Sato N, Takasaka N, Yoshida M, Tsubouchi K, Minagawa S, Araya J, Saito N, Fujita Y, Kurita Y, Kobayashi K, Ito S, Hara H, Kadota T, Yanagisawa H, Hashimoto M, Utsumi H, Wakui H, Kojima J, Numata T, Kaneko Y, Odaka M, Morikawa T, Nakayama K, Kohrogi H, Kuwano K (2016) Metformin attenuates lung fibrosis development via NOX4 suppression. Respir Res 17(1):107. https://doi.org/10.1186/s12931-016-0420-x 315. Rangarajan S, Bone NB, Zmijewska AA, Jiang S, Park DW, Bernard K, Locy ML, Ravi S, Deshane J, Mannon RB, Abraham E, Darley-Usmar V, Thannickal VJ, Zmijewski JW (2018) Metformin reverses established lung fibrosis in a bleomycin model. Nat Med 24(8):1121–1127. https://doi.org/10.1038/ s41591-018-0087-6 316. Jiang JX, Chen X, Serizawa N, Szyndralewiez C, Page P, Schroder K, Brandes RP, Devaraj S, Torok NJ (2012) Liver fibrosis and hepatocyte apoptosis are attenuated by GKT137831, a novel NOX4/NOX1 inhibitor in vivo. Free Radic Biol Med 53(2): 289–296. https://doi.org/10.1016/j.freeradbiomed.2012.05.007 317. Proell V, Carmona-Cuenca I, Murillo MM, Huber H, Fabregat I, Mikulits W (2007) TGF-beta dependent regulation of oxygen radicals during transdifferentiation of activated hepatic stellate cells to myofibroblastoid cells. Comp Hepatol 6:1. https://doi.org/ 10.1186/1476-5926-6-1 318. Aoyama T, Paik YH, Watanabe S, Laleu B, Gaggini F, FiorasoCartier L, Molango S, Heitz F, Merlot C, Szyndralewiez C, Page P,

L. Hecker et al. Brenner DA (2012) Nicotinamide adenine dinucleotide phosphate oxidase in experimental liver fibrosis: GKT137831 as a novel potential therapeutic agent. Hepatology 56(6):2316–2327. https:// doi.org/10.1002/hep.25938 319. Lan T, Kisseleva T, Brenner DA (2015) Deficiency of NOX1 or NOX4 prevents liver inflammation and fibrosis in mice through inhibition of hepatic stellate cell activation. PLoS One 10(7): e0129743. https://doi.org/10.1371/journal.pone.0129743 320. Bettaieb A, Jiang JX, Sasaki Y, Chao TI, Kiss Z, Chen X, Tian J, Katsuyama M, Yabe-Nishimura C, Xi Y, Szyndralewiez C, Schroder K, Shah A, Brandes RP, Haj FG, Torok NJ (2015) Hepatocyte nicotinamide adenine dinucleotide phosphate reduced oxidase 4 regulates stress signaling, fibrosis, and insulin sensitivity during development of steatohepatitis in mice. Gastroenterology 149(2):468–480.e410. https://doi.org/10.1053/j.gastro.2015. 04.009 321. Audzeyenka I, Rogacka D, Piwkowska A, Rychlowski M, Bierla JB, Czarnowska E, Angielski S, Jankowski M (2016) Reactive oxygen species are involved in insulin-dependent regulation of autophagy in primary rat podocytes. Int J Biochem Cell Biol 75: 23–33. https://doi.org/10.1016/j.biocel.2016.03.015 322. Nlandu-Khodo S, Dissard R, Hasler U, Schafer M, Pircher H, Jansen-Durr P, Krause KH, Martin PY, de Seigneux S (2016) NADPH oxidase 4 deficiency increases tubular cell death during acute ischemic reperfusion injury. Sci Rep 6:38598. https://doi.org/ 10.1038/srep38598 323. Santos CX, Hafstad AD, Beretta M, Zhang M, Molenaar C, Kopec J, Fotinou D, Murray TV, Cobb AM, Martin D, Zeh Silva M, Anilkumar N, Schroder K, Shanahan CM, Brewer AC, Brandes RP, Blanc E, Parsons M, Belousov V, Cammack R, Hider RC, Steiner RA, Shah AM (2016) Targeted redox inhibition of protein phosphatase 1 by Nox4 regulates eIF2alpha-mediated stress signaling. EMBO J 35(3):319–334. https://doi.org/10. 15252/embj.201592394 324. Nlandu Khodo S, Dizin E, Sossauer G, Szanto I, Martin PY, Feraille E, Krause KH, de Seigneux S (2012) NADPH-oxidase 4 protects against kidney fibrosis during chronic renal injury. J Am Soc Nephrol 23(12):1967–1976. https://doi.org/10.1681/ASN. 2012040373 325. Eid AA, Ford BM, Bhandary B, de Cassia CR, Block K, Barnes JL, Gorin Y, Choudhury GG, Abboud HE (2013) Mammalian target of rapamycin regulates Nox4-mediated podocyte depletion in diabetic renal injury. Diabetes 62(8):2935–2947. https://doi.org/10.2337/ db12-1504 326. Gorin Y, Block K, Hernandez J, Bhandari B, Wagner B, Barnes JL, Abboud HE (2005) Nox4 NAD(P)H oxidase mediates hypertrophy and fibronectin expression in the diabetic kidney. J Biol Chem 280(47):39616–39626 327. Eid AA, Lee DY, Roman LJ, Khazim K, Gorin Y (2013) Sestrin 2 and AMPK connect hyperglycemia to Nox4-dependent endothelial nitric oxide synthase uncoupling and matrix protein expression. Mol Cell Biol 33(17):3439–3460. https://doi.org/10.1128/MCB. 00217-13 328. Thallas-Bonke V, Jha JC, Gray SP, Barit D, Haller H, Schmidt HH, Coughlan MT, Cooper ME, Forbes JM, Jandeleit-Dahm KA (2014) Nox-4 deletion reduces oxidative stress and injury by PKC-alpha-associated mechanisms in diabetic nephropathy. Physiol Rep 2(11). https://doi.org/10.14814/phy2.12192 329. Jha JC, Thallas-Bonke V, Banal C, Gray SP, Chow BS, Ramm G, Quaggin SE, Cooper ME, Schmidt HH, Jandeleit-Dahm KA (2016) Podocyte-specific Nox4 deletion affords renoprotection in a mouse model of diabetic nephropathy. Diabetologia 59(2): 379–389. https://doi.org/10.1007/s00125-015-3796-0 330. Jha JC, Gray SP, Barit D, Okabe J, El-Osta A, Namikoshi T, Thallas-Bonke V, Wingler K, Szyndralewiez C, Heitz F, Touyz RM, Cooper ME, Schmidt HH, Jandeleit-Dahm KA (2014)

12

Nox4: From Discovery to Pathophysiology

Genetic targeting or pharmacologic inhibition of NADPH oxidase nox4 provides renoprotection in long-term diabetic nephropathy. J Am Soc Nephrol 25(6):1237–1254. https://doi.org/10.1681/ASN. 2013070810 331. Ben Mkaddem S, Pedruzzi E, Werts C, Coant N, Bens M, Cluzeaud F, Goujon JM, Ogier-Denis E, Vandewalle A (2010) Heat shock protein gp96 and NAD(P)H oxidase 4 play key roles in Toll-like receptor 4-activated apoptosis during renal ischemia/ reperfusion injury. Cell Death Differ 17(9):1474–1485. https://doi. org/10.1038/cdd.2010.26 332. Meng XM, Ren GL, Gao L, Yang Q, Li HD, Wu WF, Huang C, Zhang L, Lv XW, Li J (2018) NADPH oxidase 4 promotes cisplatin-induced acute kidney injury via ROS-mediated programmed cell death and inflammation. Lab Invest 98(1): 63–78. https://doi.org/10.1038/labinvest.2017.120 333. Gao L, Wu WF, Dong L, Ren GL, Li HD, Yang Q, Li XF, Xu T, Li Z, Wu BM, Ma TT, Huang C, Huang Y, Zhang L, Lv X, Li J, Meng XM (2016) Protocatechuic aldehyde attenuates cisplatininduced acute kidney injury by suppressing Nox-mediated oxidative stress and renal inflammation. Front Pharmacol 7:479. https:// doi.org/10.3389/fphar.2016.00479 334. Jung KJ, Min KJ, Park JW, Park KM, Kwon TK (2016) Carnosic acid attenuates unilateral ureteral obstruction-induced kidney fibrosis via inhibition of Akt-mediated Nox4 expression. Free Radic Biol Med 97:50–57. https://doi.org/10.1016/j.freeradbiomed.2016. 05.020 335. Lee DY, Wauquier F, Eid AA, Roman LJ, Ghosh-Choudhury G, Khazim K, Block K, Gorin Y (2013) Nox4 NADPH oxidase mediates peroxynitrite-dependent uncoupling of endothelial nitric-oxide synthase and fibronectin expression in response to angiotensin II: role of mitochondrial reactive oxygen species. J Biol Chem 288(40):28668–28686. https://doi.org/10.1074/jbc. M113.470971 336. Gregg JL, Turner RM 2nd, Chang G, Joshi D, Zhan Y, Chen L, Maranchie JK (2014) NADPH oxidase NOX4 supports renal tumorigenesis by promoting the expression and nuclear accumulation of HIF2alpha. Cancer Res 74(13):3501–3511. https://doi.org/ 10.1158/0008-5472.CAN-13-2979 337. Fitzgerald JP, Nayak B, Shanmugasundaram K, Friedrichs W, Sudarshan S, Eid AA, DeNapoli T, Parekh DJ, Gorin Y, Block K (2012) Nox4 mediates renal cell carcinoma cell invasion through hypoxia-induced interleukin 6- and 8-production. PLoS One 7(1): e30712. https://doi.org/10.1371/journal.pone.0030712 338. Chang G, Chen L, Lin HM, Lin Y, Maranchie JK (2012) Nox4 inhibition enhances the cytotoxicity of cisplatin in human renal cancer cells. J Exp Ther Oncol 10(1):9–18 339. Etoh T, Inoguchi T, Kakimoto M, Sonoda N, Kobayashi K, Kuroda J, Sumimoto H, Nawata H (2003) Increased expression of NAD(P)H oxidase subunits, NOX4 and p22phox, in the kidney of streptozotocin-induced diabetic rats and its reversibity by interventive insulin treatment. Diabetologia 46(10):1428–1437 340. Benter IF, Yousif MH, Dhaunsi GS, Kaur J, Chappell MC, Diz DI (2008) Angiotensin-(1-7) prevents activation of NADPH oxidase and renal vascular dysfunction in diabetic hypertensive rats. Am J Nephrol 28(1):25–33. https://doi.org/10.1159/000108758 341. Ilatovskaya DV, Blass G, Palygin O, Levchenko V, Pavlov TS, Grzybowski MN, Winsor K, Shuyskiy LS, Geurts AM, Cowley AW Jr, Birnbaumer L, Staruschenko A (2018) A NOX4/TRPC6 pathway in podocyte calcium regulation and renal damage in diabetic kidney disease. J Am Soc Nephrol 29(7):1917–1927. https://doi.org/10.1681/ASN.2018030280 342. Gorin Y, Cavaglieri RC, Khazim K, Lee DY, Bruno F, Thakur S, Fanti P, Szyndralewiez C, Barnes JL, Block K, Abboud HE (2015) Targeting NADPH oxidase with a novel dual Nox1/Nox4 inhibitor attenuates renal pathology in type 1 diabetes. Am J Physiol Renal

211 Physiol 308(11):F1276–F1287. https://doi.org/10.1152/ajprenal. 00396.2014 343. You YH, Quach T, Saito R, Pham J, Sharma K (2016) Metabolomics reveals a key role for fumarate in mediating the effects of NADPH oxidase 4 in diabetic kidney disease. J Am Soc Nephrol 27(2):466–481. https://doi.org/10.1681/ASN. 2015030302 344. Cha JJ, Min HS, Kim KT, Kim JE, Ghee JY, Kim HW, Lee JE, Han JY, Lee G, Ha HJ, Bae YS, Lee SR, Moon SH, Lee SC, Kim G, Kang YS, Cha DR (2017) APX-115, a first-in-class pan-NADPH oxidase (Nox) inhibitor, protects db/db mice from renal injury. Lab Invest 97(4):419–431. https://doi.org/10.1038/ labinvest.2017.2 345. Kwon G, Uddin MJ, Lee G, Jiang S, Cho A, Lee JH, Lee SR, Bae YS, Moon SH, Lee SJ, Cha DR, Ha H (2017) A novel pan-Nox inhibitor, APX-115, protects kidney injury in streptozotocininduced diabetic mice: possible role of peroxisomal and mitochondrial biogenesis. Oncotarget 8(43):74217–74232. https://doi.org/ 10.18632/oncotarget.18540 346. Rajaram RD, Dissard R, Jaquet V, de Seigneux S (2019) Potential benefits and harms of NADPH oxidase type 4 in the kidneys and cardiovascular system. Nephrol Dial Transplant 34(4):567–576. https://doi.org/10.1093/ndt/gfy161 347. Gorin Y, Block K (2013) Nox4 and diabetic nephropathy: with a friend like this, who needs enemies? Free Radic Biol Med 61:130– 142. https://doi.org/10.1016/j.freeradbiomed.2013.03.014 348. Maalouf RM, Eid AA, Gorin YC, Block K, Escobar GP, Bailey S, Abboud HE (2012) Nox4-derived reactive oxygen species mediate cardiomyocyte injury in early type 1 diabetes. Am J Physiol Cell Physiol 302(3):C597–C604. https://doi.org/10.1152/ajpcell.00331. 2011 349. Den Hartigh LJ, Omer M, Goodspeed L, Wang S, Wietecha T, O’Brien KD, Han CY (2017) Adipocyte-specific deficiency of NADPH oxidase 4 delays the onset of insulin resistance and attenuates adipose tissue inflammation in obesity. Arterioscler Thromb Vasc Biol 37(3):466–475. https://doi.org/10.1161/ ATVBAHA.116.308749 350. San Martin A, Du P, Dikalova A, Lassegue B, Aleman M, Gongora MC, Brown K, Joseph G, Harrison DG, Taylor WR, Jo H, Griendling KK (2007) Reactive oxygen species-selective regulation of aortic inflammatory gene expression in Type 2 diabetes. Am J Physiol Heart Circ Physiol 292(5):H2073–H2082. https://doi.org/ 10.1152/ajpheart.00943.2006 351. Wu X, Williams KJ (2012) NOX4 pathway as a source of selective insulin resistance and responsiveness. Arterioscler Thromb Vasc Biol 32(5):1236–1245. https://doi.org/10.1161/ATVBAHA.111. 244525 352. Sedeek M, Gutsol A, Montezano AC, Burger D, Nguyen Dinh Cat A, Kennedy CR, Burns KD, Cooper ME, Jandeleit-Dahm K, Page P, Szyndralewiez C, Heitz F, Hebert RL, Touyz RM (2013) Renoprotective effects of a novel Nox1/4 inhibitor in a mouse model of Type 2 diabetes. Clin Sci 124(3):191–202. https://doi. org/10.1042/CS20120330 353. Li J, Wang JJ, Yu Q, Chen K, Mahadev K, Zhang SX (2010) Inhibition of reactive oxygen species by Lovastatin downregulates vascular endothelial growth factor expression and ameliorates blood-retinal barrier breakdown in db/db mice: role of NADPH oxidase 4. Diabetes 59(6):1528–1538. https://doi.org/10.2337/ db09-1057 354. Meng W, Shah KP, Pollack S, Toppila I, Hebert HL, MI MC, Groop L, Ahlqvist E, Lyssenko V, Agardh E, Daniell M, Kaidonis G, Craig JE, Mitchell P, Liew G, Kifley A, Wang JJ, Christiansen MW, Jensen RA, Penman A, Hancock HA, Chen CJ, Correa A, Kuo JZ, Li X, Chen YI, Rotter JI, Klein R, Klein B, Wong TY, Morris AD, ASF D, Colhoun HM, Price AL, Burdon KP, Groop PH, Sandholm N, Grassi MA, Sobrin L, CNA P,

212 Wellcome Trust Case Control Consortium 2 (WTCCC2), Surrogate markers for Micro- and Macro-vascular hard endpoints for Innovative diabetes Tools (SUMMIT) Study Group (2018) A genome-wide association study suggests new evidence for an association of the NADPH Oxidase 4 (NOX4) gene with severe diabetic retinopathy in type 2 diabetes. Acta Ophthalmol 96(7):e811– e819. https://doi.org/10.1111/aos.13769 355. Handayaningsih AE, Iguchi G, Fukuoka H, Nishizawa H, Takahashi M, Yamamoto M, Herningtyas EH, Okimura Y, Kaji H, Chihara K, Seino S, Takahashi Y (2011) Reactive oxygen species play an essential role in IGF-I signaling and IGF-I-induced myocyte hypertrophy in C2C12 myocytes. Endocrinology 152(3): 912–921. https://doi.org/10.1210/en.2010-0981 356. Steinhorn B, Sartoretto JL, Sorrentino A, Romero N, Kalwa H, Abel ED, Michel T (2017) Insulin-dependent metabolic and inotropic responses in the heart are modulated by hydrogen peroxide from NADPH-oxidase isoforms NOX2 and NOX4. Free Radic Biol Med 113:16–25. https://doi.org/10.1016/j.freeradbiomed. 2017.09.006 357. Jiao W, Ji J, Li F, Guo J, Zheng Y, Li S, Xu W (2019) Activation of the NotchNox4reactive oxygen species signaling pathway induces cell death in high glucosetreated human retinal endothelial cells. Mol Med Rep 19(1):667–677. https://doi.org/10.3892/mmr.2018. 9637 358. Meng D, Mei A, Liu J, Kang X, Shi X, Qian R, Chen S (2012) NADPH oxidase 4 mediates insulin-stimulated HIF-1alpha and VEGF expression, and angiogenesis in vitro. PLoS One 7(10): e48393. https://doi.org/10.1371/journal.pone.0048393 359. Williams CR, Lu X, Sutliff RL, Hart CM (2012) Rosiglitazone attenuates NF-kappaB-mediated Nox4 upregulation in hyperglycemia-activated endothelial cells. Am J Physiol Cell Physiol 303(2):C213–C223. https://doi.org/10.1152/ajpcell. 00227.2011 360. Plecita-Hlavata L, Jaburek M, Holendova B, Tauber J, Pavluch V, Berkova Z, Cahova M, Schroder K, Brandes RP, Siemen D, Jezek P (2020) Glucose-stimulated insulin secretion fundamentally requires H2O2 signaling by NADPH oxidase 4. Diabetes 69(7): 1341–1354. https://doi.org/10.2337/db19-1130 361. Mandal CC, Ganapathy S, Gorin Y, Mahadev K, Block K, Abboud HE, Harris SE, Ghosh-Choudhury G, Ghosh-Choudhury N (2011) Reactive oxygen species derived from Nox4 mediate BMP2 gene transcription and osteoblast differentiation. Biochem J 433(2): 393–402. https://doi.org/10.1042/BJ20100357 362. Goettsch C, Babelova A, Trummer O, Erben RG, Rauner M, Rammelt S, Weissmann N, Weinberger V, Benkhoff S, Kampschulte M, Obermayer-Pietsch B, Hofbauer LC, Brandes RP, Schroder K (2013) NADPH oxidase 4 limits bone mass by promoting osteoclastogenesis. J Clin Invest 123(11):4731–4738. https://doi.org/10.1172/JCI67603 363. Sun J, Chen W, Li S, Yang S, Zhang Y, Hu X, Qiu H, Wu J, Xu S, Chu T (2021) Nox4 promotes RANKL-induced autophagy and osteoclastogenesis via activating ROS/PERK/eIF-2alpha/ATF4 pathway. Front Pharmacol 12:751845. https://doi.org/10.3389/ fphar.2021.751845 364. Chen L, Xiao J, Kuroda J, Ago T, Sadoshima J, Cohen RA, Tong X (2014) Both hydrogen peroxide and transforming growth factor beta 1 contribute to endothelial Nox4 mediated angiogenesis in endothelial Nox4 transgenic mouse lines. Biochim Biophys Acta 1842(12 Pt A):2489–2499. https://doi.org/10.1016/j.bbadis.2014. 10.007 365. Peshavariya HM, Chan EC, Liu GS, Jiang F, Dusting GJ (2014) Transforming growth factor-beta1 requires NADPH oxidase 4 for angiogenesis in vitro and in vivo. J Cell Mol Med 18(6): 1172–1183. https://doi.org/10.1111/jcmm.12263 366. Hakami NY, Dusting GJ, Peshavariya HM (2016) Trichostatin A, a histone deacetylase inhibitor suppresses NADPH Oxidase

L. Hecker et al. 4-derived redox signalling and angiogenesis. J Cell Mol Med 20(10):1932–1944. https://doi.org/10.1111/jcmm.12885 367. Peshavariya HM, Liu GS, Chang CW, Jiang F, Chan EC, Dusting GJ (2014) Prostacyclin signaling boosts NADPH oxidase 4 in the endothelium promoting cytoprotection and angiogenesis. Antioxid Redox Signal 20(17):2710–2725. https://doi.org/10.1089/ars. 2013.5374 368. Vogel J, Kruse C, Zhang M, Schroder K (2015) Nox4 supports proper capillary growth in exercise and retina neo-vascularization. J Physiol 593(9):2145–2154. https://doi.org/10.1113/jphysiol. 2014.284901 369. Yu MO, Park KJ, Park DH, Chung YG, Chi SG, Kang SH (2015) Reactive oxygen species production has a critical role in hypoxiainduced Stat3 activation and angiogenesis in human glioblastoma. J Neurooncol 125(1):55–63. https://doi.org/10.1007/s11060-0151889-8 370. Wang J, Hong Z, Zeng C, Yu Q, Wang H (2014) NADPH oxidase 4 promotes cardiac microvascular angiogenesis after hypoxia/reoxygenation in vitro. Free Radic Biol Med 69:278–288. https://doi. org/10.1016/j.freeradbiomed.2014.01.027 371. Zhuang J, Jiang T, Lu D, Luo Y, Zheng C, Feng J, Yang D, Chen C, Yan X (2010) NADPH oxidase 4 mediates reactive oxygen species induction of CD146 dimerization in VEGF signal transduction. Free Radic Biol Med 49(2):227–236. https://doi.org/ 10.1016/j.freeradbiomed.2010.04.007 372. Datla SR, Peshavariya H, Dusting GJ, Mahadev K, Goldstein BJ, Jiang F (2007) Important role of Nox4 type NADPH oxidase in angiogenic responses in human microvascular endothelial cells in vitro. Arterioscler Thromb Vasc Biol 27(11):2319–2324. https://doi.org/10.1161/ATVBAHA.107.149450 373. Helfinger V, Henke N, Harenkamp S, Walter M, Epah J, Penski C, Mittelbronn M, Schroder K (2016) The NADPH Oxidase Nox4 mediates tumour angiogenesis. Acta Physiol (Oxf) 216(4): 435–446. https://doi.org/10.1111/apha.12625 374. Li Y, Han N, Yin T, Huang L, Liu S, Liu D, Xie C, Zhang M (2014) Lentivirus-mediated Nox4 shRNA invasion and angiogenesis and enhances radiosensitivity in human glioblastoma. Oxid Med Cell Longev 2014:581732. https://doi.org/10.1155/2014/ 581732 375. Meitzler JL, Konate MM, Doroshow JH (2019) Hydrogen peroxide-producing NADPH oxidases and the promotion of migratory phenotypes in cancer. Arch Biochem Biophys 675:108076. https://doi.org/10.1016/j.abb.2019.108076 376. Meitzler JL, Makhlouf HR, Antony S, Wu Y, Butcher D, Jiang G, Juhasz A, Lu J, Dahan I, Jansen-Durr P, Pircher H, Shah AM, Roy K, Doroshow JH (2017) Decoding NADPH oxidase 4 expression in human tumors. Redox Biol 13:182–195. https://doi.org/10. 1016/j.redox.2017.05.016 377. Lee JK, Edderkaoui M, Truong P, Ohno I, Jang KT, Berti A, Pandol SJ, Gukovskaya AS (2007) NADPH oxidase promotes pancreatic cancer cell survival via inhibiting JAK2 dephosphorylation by tyrosine phosphatases. Gastroenterology 133(5): 1637–1648. https://doi.org/10.1053/j.gastro.2007.08.022 378. Luengo E, Trigo-Alonso P, Fernandez-Mendivil C, Nunez A, Campo MD, Porrero C, Garcia-Magro N, Negredo P, Senar S, Sanchez-Ramos C, Bernal JA, Rabano A, Hoozemans J, Casas AI, Schmidt H, Lopez MG (2021) Implication of type 4 NADPH oxidase (NOX4) in tauopathy. Redox Biol 49:102210. https://doi. org/10.1016/j.redox.2021.102210 379. Park MW, Cha HW, Kim J, Kim JH, Yang H, Yoon S, Boonpraman N, Yi SS, Yoo ID, Moon JS (2021) NOX4 promotes ferroptosis of astrocytes by oxidative stress-induced lipid peroxidation via the impairment of mitochondrial metabolism in Alzheimer’s diseases. Redox Biol 41:101947. https://doi.org/10. 1016/j.redox.2021.101947

12

Nox4: From Discovery to Pathophysiology

380. Jiang S, Zhao Y, Zhang T, Lan J, Yang J, Yuan L, Zhang Q, Pan K, Zhang K (2018) Galantamine inhibits beta-amyloid-induced cytostatic autophagy in PC12 cells through decreasing ROS production. Cell Prolif 51(3):e12427. https://doi.org/10.1111/cpr.12427 381. Nishimura A, Ago T, Kuroda J, Arimura K, Tachibana M, Nakamura K, Wakisaka Y, Sadoshima J, Iihara K, Kitazono T (2016) Detrimental role of pericyte Nox4 in the acute phase of brain ischemia. J Cereb Blood Flow Metab 36(6):1143–1154. https://doi.org/10.1177/0271678X15606456 382. Chen YH, Lu HI, Lo CM, Hsiao CC, Li SH (2020) NOX4 overexpression is a poor prognostic factor in patients undergoing curative esophagectomy for esophageal squamous cell carcinoma. Surgery 167(3):620–627. https://doi.org/10.1016/j.surg.2019. 11.017 383. Cross AR, Segal AW (2004) The NADPH oxidase of professional phagocytes—prototype of the NOX electron transport chain systems. Biochim Biophys Acta 1657(1):1–22. https://doi.org/10. 1016/j.bbabio.2004.03.008 384. Joo JH, Huh JE, Lee JH, Park DR, Lee Y, Lee SG, Choi S, Lee HJ, Song SW, Jeong Y, Goo JI, Choi Y, Baek HK, Yi SS, Park SJ, Lee JE, Ku SK, Lee WJ, Lee KI, Lee SY, Bae YS (2016) A novel pyrazole derivative protects from ovariectomy-induced osteoporosis through the inhibition of NADPH oxidase. Sci Rep 6:22389. https://doi.org/10.1038/srep22389 385. Chen Z, Tao S, Li X, Zeng X, Zhang M, Yao Q (2019) Anagliptin protects neuronal cells against endogenous amyloid beta (Aβ)induced cytotoxicity and apoptosis. Artif Cells Nanomed Biotechnol 47(1):2213–2220. https://doi.org/10.1080/21691401. 2019.1609979 386. Anvari E, Wikstrom P, Walum E, Welsh N (2015) The novel NADPH oxidase 4 inhibitor GLX351322 counteracts glucose intolerance in high-fat diet-treated C57BL/6 mice. Free Radic Res 49(11):1308–1318. https://doi.org/10.3109/10715762.2015. 1067697 387. Degasper C, Brunner A, Sampson N, Tsibulak I, Wieser V, Welponer H, Marth C, Fiegl H, Zeimet AG (2019) NADPH oxidase 4 expression in the normal endometrium and in endometrial cancer. Tumour Biol 41(2):1010428319830002. https://doi.org/10. 1177/1010428319830002 388. Boudreau HE, Casterline BW, Rada B, Korzeniowska A, Leto TL (2012) Nox4 involvement in TGF-beta and SMAD3-driven induction of the epithelial-to-mesenchymal transition and migration of breast epithelial cells. Free Radic Biol Med 53(7):1489–1499. https://doi.org/10.1016/j.freeradbiomed.2012.06.016 389. Hiraga R, Kato M, Miyagawa S, Kamata T (2013) Nox4-derived ROS signaling contributes to TGF-beta-induced epithelialmesenchymal transition in pancreatic cancer cells. Anticancer Res 33(10):4431–4438 390. Seo JM, Park S, Kim JH (2012) Leukotriene B4 receptor-2promotes invasiveness and metastasis of ovarian cancer cells through signal transducer and activator of transcription 3 (STAT3)-dependent up-regulation of matrix metalloproteinase 2. J Biol Chem 287(17):13840–13849. https://doi.org/10.1074/jbc. M111.317131 391. Diaz B, Shani G, Pass I, Anderson D, Quintavalle M, Courtneidge SA (2009) Tks5-dependent, nox-mediated generation of reactive oxygen species is necessary for invadopodia formation. Sci Signal 2(88):ra53. https://doi.org/10.1126/scisignal.2000368 392. Wang X, Liu Z, Sun J, Song X, Bian M, Wang F, Yan F, Yu Z (2021) Inhibition of NADPH oxidase 4 attenuates lymphangiogenesis and tumor metastasis in breast cancer. FASEB J 35(4):e21531. https://doi.org/10.1096/fj.202002533R 393. Ma WF, Boudreau HE, Leto TL (2021) Pan-cancer analysis shows TP53 mutations modulate the association of NOX4 with genetic programs of cancer progression and clinical outcome. Antioxidants (Basel) 10(2). https://doi.org/10.3390/antiox10020235

213 394. Lin XL, Yang L, Fu SW, Lin WF, Gao YJ, Chen HY, Ge ZZ (2017) Overexpression of NOX4 predicts poor prognosis and promotes tumor progression in human colorectal cancer. Oncotarget 8(20):33586–33600. https://doi.org/10.18632/ oncotarget.16829 395. Eun HS, Cho SY, Joo JS, Kang SH, Moon HS, Lee ES, Kim SH, Lee BS (2017) Gene expression of NOX family members and their clinical significance in hepatocellular carcinoma. Sci Rep 7(1): 11060. https://doi.org/10.1038/s41598-017-11280-3 396. Eun HS, Chun K, Song IS, Oh CH, Seong IO, Yeo MK, Kim KH (2019) High nuclear NADPH oxidase 4 expression levels are correlated with cancer development and poor prognosis in hepatocellular carcinoma. Pathology 51(6):579–585. https://doi.org/10. 1016/j.pathol.2019.05.004 397. Kaushik D, Ashcraft KA, Wang H, Shanmugasundaram K, Shah PK, Gonzalez G, Nazarullah A, Tye CB, Liss MA, Pruthi DK, Mansour AM, Chowdhury W, Bacich D, Zhang H, Watson AL, Block K, O'Keefe D, Rodriguez R (2020) Nuclear NADPH oxidase-4 associated with disease progression in renal cell carcinoma. Transl Res 223:1–14. https://doi.org/10.1016/j.trsl.2020. 05.009 398. Tang CT, Lin XL, Wu S, Liang Q, Yang L, Gao YJ, Ge ZZ (2018) NOX4-driven ROS formation regulates proliferation and apoptosis of gastric cancer cells through the GLI1 pathway. Cell Signal 46: 52–63. https://doi.org/10.1016/j.cellsig.2018.02.007 399. Chen YH, Chien CY, Fang FM, Huang TL, Su YY, Luo SD, Huang CC, Lin WC, Li SH (2018) Nox4 overexpression as a poor prognostic factor in patients with oral tongue squamous cell carcinoma receiving surgical resection. J Clin Med 7(12). https:// doi.org/10.3390/jcm7120497 400. Hensley K, Hall N, Subramaniam R, Cole P, Harris M, Aksenov M, Aksenova M, Gabbita SP, Wu JF, Carney JM et al (1995) Brain regional correspondence between Alzheimer’s disease histopathology and biomarkers of protein oxidation. J Neurochem 65(5):2146–2156. https://doi.org/10.1046/j. 1471-4159.1995.65052146.x 401. Majumder P, Chanda K, Das D, Singh BK, Chakrabarti P, Jana NR, Mukhopadhyay D (2021) A nexus of miR-1271, PAX4 and ALK/RYK influences the cytoskeletal architectures in Alzheimer’s disease and type 2 diabetes. Biochem J 478(17):3297–3317. https://doi.org/10.1042/BCJ20210175 402. Austin SA, Santhanam AV, d’Uscio LV, Katusic ZS (2015) Regional heterogeneity of cerebral microvessels and brain susceptibility to oxidative stress. PLoS One 10(12):e0144062. https://doi. org/10.1371/journal.pone.0144062 403. Thummayot S, Tocharus C, Jumnongprakhon P, Suksamrarn A, Tocharus J (2018) Cyanidin attenuates Abeta25-35-induced neuroinflammation by suppressing NF-kappaB activity downstream of TLR4/NOX4 in human neuroblastoma cells. Acta Pharmacol Sin 39(9):1439–1452. https://doi.org/10.1038/aps. 2017.203 404. Oguchi T, Ono R, Tsuji M, Shozawa H, Somei M, Inagaki M, Mori Y, Yasumoto T, Ono K, Kiuchi Y (2017) Cilostazol suppresses aβ-induced neurotoxicity in SH-SY5Y cells through inhibition of oxidative stress and MAPK signaling pathway. Front Aging Neurosci 9:337. https://doi.org/10.3389/fnagi.2017. 00337 405. Majumder P, Roy K, Singh BK, Jana NR, Mukhopadhyay D (2017) Cellular levels of Grb2 and cytoskeleton stability are correlated in a neurodegenerative scenario. Dis Model Mech 10(5):655–669. https://doi.org/10.1242/dmm.027748 406. Sagy-Bross C, Hadad N, Levy R (2013) Cytosolic phospholipase A2alpha upregulation mediates apoptotic neuronal death induced by aggregated amyloid-beta peptide1-42. Neurochem Int 63(6): 541–550. https://doi.org/10.1016/j.neuint.2013.09.007

214 407. Ha JS, Sung HY, Lim HM, Kwon KS, Park SS (2012) PI3KERK1/2 activation contributes to extracellular H2O2 generation in amyloid beta toxicity. Neurosci Lett 526(2):112–117. https://doi. org/10.1016/j.neulet.2012.08.023 408. Bruce-Keller AJ, Gupta S, Knight AG, Beckett TL, McMullen JM, Davis PR, Murphy MP, Van Eldik LJ, St Clair D, Keller JN (2011) Cognitive impairment in humanized APPxPS1 mice is linked to Abeta(1-42) and NOX activation. Neurobiol Dis 44(3):317–326. https://doi.org/10.1016/j.nbd.2011.07.012 409. Li H, Wang Y, Feng D, Liu Y, Xu M, Gao A, Tian F, Zhang L, Cui Y, Wang Z, Chen G (2014) Alterations in the time course of expression of the Nox family in the brain in a rat experimental cerebral ischemia and reperfusion model: effects of melatonin. J Pineal Res 57(1):110–119. https://doi.org/10.1111/jpi.12148 410. Kuroda J, Ago T, Nishimura A, Nakamura K, Matsuo R, Wakisaka Y, Kamouchi M, Kitazono T (2014) Nox4 is a major source of superoxide production in human brain pericytes. Journal of vascular research 51(6):429–438. https://doi.org/10.1159/ 000369930 411. Casas AI, Geuss E, Kleikers PWM, Mencl S, Herrmann AM, Buendia I, Egea J, Meuth SG, Lopez MG, Kleinschnitz C, Schmidt H (2017) NOX4-dependent neuronal autotoxicity and BBB breakdown explain the superior sensitivity of the brain to ischemic damage. Proc Natl Acad Sci U S A 114(46):12315–12320. https://doi.org/10.1073/pnas.1705034114 412. He W, Wang Q, Gu L, Zhong L, Liu D (2018) NOX4 rs11018628 polymorphism associates with a decreased risk and better shortterm recovery of ischemic stroke. Exp Ther Med 16(6):5258–5264. https://doi.org/10.3892/etm.2018.6874 413. Pan J, Lao L, Shen J, Huang S, Zhang T, Fan W, Yan M, Gu J, Liu W (2021) Utility of serum NOX4 as a potential prognostic biomarker for aneurysmal subarachnoid hemorrhage. Clin Chim Acta 517:9–14. https://doi.org/10.1016/j.cca.2021.02.007 414. Lee SR, An EJ, Kim J, Bae YS (2020) Function of NADPH oxidases in diabetic nephropathy and development of Nox inhibitors. Biomol Ther (Seoul) 28(1):25–33. https://doi.org/10. 4062/biomolther.2019.188 415. Wang X, Elksnis A, Wikstrom P, Walum E, Welsh N, Carlsson PO (2018) The novel NADPH oxidase 4 selective inhibitor GLX7013114 counteracts human islet cell death in vitro. PLoS One 13(9):e0204271. https://doi.org/10.1371/journal.pone. 0204271 416. Das S, Wikstrom P, Walum E, Lovicu FJ (2019) A novel NADPH oxidase inhibitor targeting Nox4 in TGFbeta-induced lens epithelial to mesenchymal transition. Exp Eye Res 185:107692. https:// doi.org/10.1016/j.exer.2019.107692 417. Dionysopoulou S, Wikstrom P, Walum E, Thermos K (2020) Effect of NADPH oxidase inhibitors in an experimental retinal model of excitotoxicity. Exp Eye Res 200:108232. https://doi. org/10.1016/j.exer.2020.108232 418. Szekeres FLM, Walum E, Wikstrom P, Arner A (2021) A small molecule inhibitor of Nox2 and Nox4 improves contractile function after ischemia-reperfusion in the mouse heart. Sci Rep 11(1): 11970. https://doi.org/10.1038/s41598-021-91575-8 419. Laleu B, Gaggini F, Orchard M, Fioraso-Cartier L, Cagnon L, Houngninou-Molango S, Gradia A, Duboux G, Merlot C, Heitz F, Szyndralewiez C, Page P (2010) First in class, potent, and orally bioavailable NADPH oxidase isoform 4 (Nox4) inhibitors for the treatment of idiopathic pulmonary fibrosis. J Med Chem 53(21):7715–7730. https://doi.org/10.1021/jm100773e 420. Sun Q, Zhang W, Zhong W, Sun X, Zhou Z (2017) Pharmacological inhibition of NOX4 ameliorates alcohol-induced liver injury in mice through improving oxidative stress and mitochondrial function. Biochim Biophys Acta Gen Subj 1861(1 Pt A):2912–2921. https://doi.org/10.1016/j.bbagen.2016.09.009

L. Hecker et al. 421. Nishio T, Hu R, Koyama Y, Liang S, Rosenthal SB, Yamamoto G, Karin D, Baglieri J, Ma HY, Xu J, Liu X, Dhar D, Iwaisako K, Taura K, Brenner DA, Kisseleva T (2019) Activated hepatic stellate cells and portal fibroblasts contribute to cholestatic liver fibrosis in MDR2 knockout mice. J Hepatol 71(3):573–585. https://doi. org/10.1016/j.jhep.2019.04.012 422. Zhao QD, Viswanadhapalli S, Williams P, Shi Q, Tan C, Yi X, Bhandari B, Abboud HE (2015) NADPH oxidase 4 induces cardiac fibrosis and hypertrophy through activating Akt/mTOR and NFkappaB signaling pathways. Circulation 131(7):643–655. https://doi.org/10.1161/CIRCULATIONAHA.114.011079 423. Zeng SY, Yang L, Yan QJ, Gao L, Lu HQ, Yan PK (2019) Nox1/4 dual inhibitor GKT137831 attenuates hypertensive cardiac remodelling associating with the inhibition of ADAM17dependent proinflammatory cytokines-induced signalling pathways in the rats with abdominal artery constriction. Biomed Pharmacother 109:1907–1914. https://doi.org/10.1016/j.biopha. 2018.11.077 424. Appukuttan B, Ma Y, Stempel A, Ashander LM, Deliyanti D, Wilkinson-Berka JL, Smith JR (2018) Effect of NADPH oxidase 1 and 4 blockade in activated human retinal endothelial cells. Clin Exp Ophthalmol 46(6):652–660. https://doi.org/10.1111/ceo. 13155 425. Wilkinson-Berka JL, Deliyanti D, Rana I, Miller AG, Agrotis A, Armani R, Szyndralewiez C, Wingler K, Touyz RM, Cooper ME, Jandeleit-Dahm KA, Schmidt HH (2014) NADPH oxidase, NOX1, mediates vascular injury in ischemic retinopathy. Antioxid Redox Signal 20(17):2726–2740. https://doi.org/10.1089/ars.2013.5357 426. Gray SP, Di Marco E, Okabe J, Szyndralewiez C, Heitz F, Montezano AC, de Haan JB, Koulis C, El-Osta A, Andrews KL, Chin-Dusting JP, Touyz RM, Wingler K, Cooper ME, Schmidt HH, Jandeleit-Dahm KA (2013) NADPH oxidase 1 plays a key role in diabetes mellitus-accelerated atherosclerosis. Circulation 127(18):1888–1902. https://doi.org/10.1161/ CIRCULATIONAHA.112.132159 427. Ford K, Hanley CJ, Mellone M, Szyndralewiez C, Heitz F, Wiesel P, Wood O, Machado M, Lopez MA, Ganesan AP, Wang C, Chakravarthy A, Fenton TR, King EV, Vijayanand P, Ottensmeier CH, Al-Shamkhani A, Savelyeva N, Thomas GJ (2020) NOX4 inhibition potentiates immunotherapy by overcoming cancer-associated fibroblast-mediated CD8 T-cell exclusion from tumors. Cancer Res 80(9):1846–1860. https://doi. org/10.1158/0008-5472.CAN-19-3158 428. Reutens AT, Jandeleit-Dahm K, Thomas M, Bach LA, Colman PG, Davis TME, D'Emden M, Ekinci EI, Fulcher G, Hamblin PS, Kotowicz MA, MacIsaac RJ, Morbey C, Simmons D, Soldatos G, Wittert G, Wu T, Cooper ME, Shaw JE (2019) A physician-initiated double-blind, randomised, placebo-controlled, phase 2 study evaluating the efficacy and safety of inhibition of NADPH oxidase with the first-in-class Nox-1/4 inhibitor, GKT137831, in adults with type 1 diabetes and persistently elevated urinary albumin excretion: protocol and statistical considerations. Contemp Clin Trials 90:105892. https://doi.org/ 10.1016/j.cct.2019.105892 429. De Livera AM, Reutens A, Cooper M, Thomas M, JandeleitDahm K, Shaw JE, Salim A (2020) Evaluating the efficacy and safety of GKT137831 in adults with type 1 diabetes and persistently elevated urinary albumin excretion: a statistical analysis plan. Trials 21(1):459. https://doi.org/10.1186/s13063-02004404-0 430. Augsburger F, Filippova A, Rasti D, Seredenina T, Lam M, Maghzal G, Mahiout Z, Jansen-Durr P, Knaus UG, Doroshow J, Stocker R, Krause KH, Jaquet V (2019) Pharmacological characterization of the seven human NOX isoforms and their inhibitors. Redox Biol 26:101272. https://doi.org/10.1016/j.redox.2019. 101272

Nox5: Molecular Regulation and Pathophysiology

13

Livia L. Camargo, Francisco Rios, Augusto Montezano, and Rhian M. Touyz

Abstract

Nox5 is considered to be the precursor NADPH oxidase in the evolution of the Noxs and is the most ancient member of the Nox family. It is widely conserved across species and five splice variants have been identified. Nox5 shares structural homology with other isoforms, specifically the catalytic core comprising six transmembrane helices chelating two hemes and a dehydrogenase domain that binds FAD and NADPH. Nox5 is distinct from Nox1–4 in that it has a calcium-binding region with EF-hand domains and is the only isoform that does not require NADPH oxidase subunits for its activation. Nox5 is activated by an increase in intracellular free calcium concentration and undergoes conformational change with consequent superoxide (O2•-) production. It is also regulated by postranslational modifications and interaction with regulatory partners. While there have been advances in the molecular biology of Nox5, there is a paucity of information on the pathophysiological role of this isoform. Nox5 may be important in sperm function and smooth muscle contraction and it has been implicated in various pathologies including cardiovascular disease, kidney disease and cancer. This chapter provides a comprehensive review of the discovery, regulation and function of Nox5 and the putative role in human health and disease. Keywords

NADPH oxidase 5 · Reactive oxygen species · EF-hands · Calmodulin · Calcium · Cardiovascular disease · Cancer

L. L. Camargo · F. Rios · A. Montezano · R. M. Touyz (✉) Research Institute of the McGill University Health Centre, McGill University, Montreal, Canada e-mail: [email protected]; [email protected]; augusto. [email protected]; [email protected]

1

Introduction

Eukaryotic Nox5 was originally identified in pachytene spermatocytes of testes and in B- and T-lymphocyte-rich regions in spleen and lymph nodes, but has now been found to be expressed in numerous tissues and cell types [1, 2]. Nox5 shares structural homology with other Nox isoforms in that it shares a catalytic core comprising six transmembrane helices with an intracellular C-and N-terminal [1, 2]. However Nox5 is distinct from Nox1–4, in that it has a calcium-binding region with multiple EF-hand domains in its N-terminal [3]. It is further distinguished from other Noxs in that it is the only isoform that does not require NADPH oxidase subunits for its activation [4]. Nox5 is rapidly activated by an increase in intracellular free calcium concentration ([Ca2+]i) [3]. Calcium binding induces a conformational change with consequent O2•- generation [1, 3]. Besides changes in [Ca2+]i, many other factors regulate Nox5 including postranslational modifications, interaction with molecular chaperones and regulatory partners and modulation by co-factors and signaling molecules [5–8]. Molecular chaperones including heat shock protein 90 (Hsp90), Hsp70 and heme also modulate Nox5 activity. Signaling proteins and co-factors that influence Nox5 include calmodulin, phosphatidic acid, arachidonate, protein kinases (protein kinase C (PKC), extracellular-regulated kinases (ERK)), and tyrosine kinases (c-Src, c-Abl) [4, 9–11]. Five splice variants have been identified, although the functional significance of these remain unclear [12]. Nox5 is widely expressed and has been identified in testes, spleen, lymph nodes, pancreas, placenta, bone marrow, uterus, kidney, vessels and stomach. In humans, Nox5 is abundant in cancer cells and seems to be the major source of reactive oxygen species (ROS) in spermatozoa, endothelial cells, vascular smooth muscle cells and fibroblasts. In physiological conditions Nox5 may be important in sperm motility and smooth muscle contraction, while in pathological

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_13

215

216

L. L. Camargo et al.

conditions Nox5-associated oxidative stress has been implicated in cardiovascular disease, kidney disease, and cancer. Nox5 was only identified in 2001. There is a relative paucity of information about the pathophysiology of Nox5, with less than 350 papers published to date with ‘Nox5’ as the key word (Pubmed). While this chapter provides a comprehensive and up to date appraisal of the molecular and cell biology of Nox5 and its putative role in health and disease, it should be emphasized that Nox5 research is still in its infancy.

2

Evolution and Discovery of Eukaryotic Nox5

Nox5 was the first isoform to evolve [1, 2]. Despite its early position in the evolutionary tree, it was amongst the last of the eukaryotic isoforms to be discovered and characterized. Accordingly, it is considered to be the precursor NADPH oxidase in the evolution of the Nox enzymes and is the most ancient member of the Nox family [13–15]. Nox5 is widely conserved across species, with Nox5 expressed in plants, arthropods, reptiles and mammals, yet curiously it is absent in most rodents [1, 2]. Six splice variants have been identified including Nox5α, Nox5β, Nox5γ, Nox5δ, Nox5ζ (termed long forms of Nox (Nox5L)) and Nox5ε or Nox5S (short form). Long forms of Nox5 are differentiated by differences in the sequences of calcium-binding domains of the N-terminal, while the truncated Nox5S variant is devoid of the calcium-binding domain [12]. The Nox5 gene seems to have originated from microorganisms resembling bacteria and α-proteobacteria and Nox5, or Nox5-like protein, is found in cyanobacteria, protozoa, plants, fungi, arthropods and the animal kingdom. Sequence alignment studies show a long evolutionary process of segregation in sub-groups, further supporting the theory that Nox5 evolved early [13, 14]. During the evolution of the ferric reductase enzymes, including Noxs, the addition of specialized control mechanisms to regulate enzymatic function were required. The particular addition of these regulatory elements, including EF-hands, allowed the Nox5-like (or Nox-EF) enzymes to be regulated by calcium linked to conformational changes and binding to regulatory proteins, such as calmodulin [15]. EF-hands with calcium- binding properties were observed in early versions of Nox5 from plants and dictyostelium as one single pair, while in mammalian species two pairs of EF-hands were acquired [16, 17]. The Nox5 structure represents the transition between the other members of the Nox family, Nox1–4 and Duoxs. Nox5 possesses a

longer N-terminus region with the calcium-binding motifs but lacks the peroxidise domain observed in Duoxs. The Nox5 gene has undergone a series of duplications and deletions, many related to the coding of EF-hands. This may have happened more recently in evolution leading to the non-monophyletic origin of Nox5 and to the diversion into Nox1–4-like and Duox-like proteins. Examples of deletion and duplication are found in numerous species, where the gene may be present in some plants and fungi but absent in others [16, 17]. Nox5 orthologs are observed in invertebrates such as sea urchins (echinoderms) and arthropods. In vertebrates, Nox5 is present across most species ranging from fish to mammals, but is absent in many rodents, including mice and rats. The evolutionary process whereby rodents lost Nox5 is unclear but Nox5 functions have likely been replaced by other isoforms, possibly Duoxs, which are also calcium-sensitive.

3

Structure of Nox5

Nox5 shares the core structural organization of other Noxs. As shown in Fig. 13.1 it possesses the characteristic six transmembrane α-helices, connected by 5 loops, containing four histidine residues that coordinate the binding of two heme groups and an intracellular C-terminal dehydrogenase domain that binds FAD and NADPH [1, 2]. In its N-terminal, Nox5 has four EF-hand motifs that bind four calcium ions in its N-terminal, essential for Nox5 activation. The crystal structure of the transmembrane region of Nox5 demonstrated that molecular O2 reduction occurs via electron flow from the intracellular side through NADPH, FAD and the heme groups resulting in production of O2•-, to the extracellular or luminal space [18]. The 3D model of Nox5 transmembrane domains revealed the presence of interactions between the cytosolic loops and the C-terminal dehydrogenase domain that function as a toggle switch influencing NADPH binding. Regulation of this interaction controls access of NADPH to the enzyme activating or deactivating the first step of the redox chain [18]. It undergoes oligomerization forming enzymatically active homotetramers. Oligomerization is dependent on the C-terminal dehydrogenase domain and plays a major role in Nox5 activation and stability [19].

4

Regulation of Nox5

Nox5 is a key source of reactive oxygen species in many cell types. In humans, while Nox2 is the primary O2•--generating Nox isoform in phagocytic cells including neutrophils, Nox5 seems to be the main Nox isoform in non-phagocytic cells.

13

Nox5: Molecular Regulation and Pathophysiology

217

Fig. 13.1 Schematic of Nox5 structure. Nox5 contains six transmembrane domains enclosing two heme groups (Fe) bound by four histidine residues (H) in the transmembrane domain III and V. The intracellular N-terminal domain contains four Ca2+ binding EF hands and a

dehydrogenase domain with binding sites for NADPH and FAD and is located in the C-terminal domain. Phosphorylation sites at serine and threonine residues are found in Nox5 C-terminal domain

Unlike the other Noxs (Nox1–4), Nox5 does not require binding to membrane (p22phox) or cytosolic subunits (p47phox, p67phox, p40phox) and small G proteins (Rac1/ 2 and Rap1A GTPases) for its activation. However, several other regulatory processes are involved in the control of Nox5 activity, including [Ca2+]i, modulation by regulatory signaling molecules and kinases, interaction with molecular chaperones, and post-translational modifications (Fig. 13.2).

activation. The interaction site for the regulatory Nox5-EF in the catalytic dehydrogenase site has been identified as the 637–660 segment, termed the regulatory EF-hand-binding domain (REFBD), and the 489–505 segment, which was previously identified as the phosphorylation region of Nox5 [20]. Initial studies demonstrating the dependence of Nox5 on calcium were performed in cell-free systems where it was clearly demonstrated that Nox5 oxidase activity is tightly linked to the cellular calcium concentration [3]. Calmodulin is critically involved in this process [21]. EF-hands seem to be calcium-specific, since magnesium does not compete with calcium, at least at magnesium concentrations up to 1 mM [22, 23]. However, at higher concentrations, magnesium seems to bind Nox5-EF3, although effects on Nox5 conformational structure and oxidase activity seem to be negligible [22, 23]. Full activation of Nox5 requires high [Ca2+]i unlikely to be observed in physiological conditions. Accordingly other mechanisms

4.1

Calcium-Dependent Activation of Nox5

A distinctive characteristic of Nox5 is the presence of the four calcium-binding EF-hand motifs within its N-terminal region, rendering reactive oxygen species generation by Nox5 calcium-dependent [1]. The regulatory N-terminal EF-hand domain interacts directly and in a calciumdependent manner with the catalytic C-terminal catalytic dehydrogenase domain of the enzyme, leading to its

218

L. L. Camargo et al.

Fig. 13.2 Distinctive characteristics of Nox5. Unlike Nox1–4, Nox5 activation is calcium-dependent and does not require binding to NADPH oxidase membrane or cytosolic subunits. However, several

other regulatory processes are involved in the regulation of Nox5 activity, such as interaction with molecular chaperones and regulatory proteins and posttranslational modifications

involving regulatory proteins have evolved to activate Nox5 at lower [Ca2+]i by increasing sensitivity to calcium [12].

calcium-deficient states [21]. When calcium is present, the C-terminal lobe of the EF-domain assumes an organized and more compact structure that allows its binding to the enzyme dehydrogenase domain of Nox5. This involves numerous conserved aspartate residues in the dehydrogenase domain. It is now evident that calcium induces an unfolded-to-folded transition of the EF-domain that promotes direct interaction with a conserved regulatory region, resulting in activation of Nox5 [21]. Calcium-bound calmodulin binds to Nox5 resulting in increase of its sensitivity to Ca2+, therefore promoting enzyme activation at low levels of [Ca2+]i [24, 25]. Nox5 is also known to associate with caveolin-1, the main structural protein of plasma membrane microdomain caveola. Binding of caveolin-1 to Nox5 leads to enzyme inactivation [26].

4.2

Regulatory Proteins

Nox5 activity is influenced by several regulatory proteins, some of which interact directly with Nox5 to influence Nox5 activity, calcium sensitivity and molecular stability and others influence Nox activity indirectly. Amongst the first proteins to be defined as a Nox5-regulatory protein was calmodulin [21]. The C-terminal of Nox5 contains a calmodulin-binding site, close to the EF-hands. Calmodulin plays an important role in calcium-dependent activation of Nox5. Studies investigating the structural properties of calcium binding to the EF-hands of Nox5 showed that calmodulin-like regulatory domains of Nox5 are partially unfolded and detached from the rest of the protein in

13

4.3

Nox5: Molecular Regulation and Pathophysiology

Molecular Chaperones

Molecular chaperones, such as heat shock proteins, facilitate conformational folding or unfolding of large proteins or multi-complex proteins and are essential in maintaining a healthy proteome [27]. Hsp90 and Hsp70 associate with Nox5 and modulate its activity. Hsp90 via its M domain binds directly to the C-terminal region and activates Nox5, while Hsp70 has inhibitory effects [12]. Hsp90 binding seems to have a dual role in that it stabilizes the dehydrogenase domain and it antagonizes formation of active Nox5 oligomers [28]. The conformational changes that occur with Nox5 activation cause displacement of Hsp90, thereby removing it from its inhibitory function [28]. Molecular mechanisms involved in these processes are complex and may involve heme, since Hsp90 is directly involved in heme insertion in heme proteins, such as nitric oxide synthase, hemoglobin, soluble guanylate cyclase [28] and Nox5. Nox5 is activated when the intracellular heme concentration is increased [28]. This is a Hsp90-dependent process because inhibition of Hsp90 blocks heme-induced activation of Nox5. Together these findings suggest that cell heme levels together with Hsp90 binding dynamically regulate Nox5 oligomerization [28]. This interaction between Hsp90 and heme may fine-tune Nox5 activity, especially in conditions associated with abnormal heme metabolism, such as anemia, hemorrhage, heart failure and immune disease [29]. Nox5 dissociation from Hsp90 and binding to Hsp70 leads to Nox5 ubiquitination and degradation by the proteasome [30].

4.4

Regulation of Nox5 by Nitric Oxide

Nitric oxide has been shown to both inhibit and activate Nox5. In endothelial and vascular smooth muscle cells, endogenously generated nitric oxide inhibits basal and agonist-stimulated activation of Nox5 and O2•- production [7]. A similar response has been observed with nitric oxide donors. Nitric oxide influences Nox5 activity by direct S-nitrosylation, which inhibits activation of Nox5. The ability of nitric oxide to suppress Nox5 activity may contribute to maintaining the redox balance between nitric oxide and O2•and prevention of oxidative stress [7]. Paradoxically, Nox5 itself seems to regulate nitric oxide synthase (NOS). In vascular cells increased Nox5-derived O2•- enhances endothelial NOS (eNOS) activity [31]. In pathological conditions, Nox5 induces eNOS uncoupling. This may accelerate production of peroxynitrite, which has injurious effects in the vasculature [31–33]. Nitric oxide has also been shown to activate Nox5 through a heme- and Hsp90-dependent process [28]. The (patho)physiological significance of this remains unclear.

219

4.5

Post-Translational Modifications

Protein post-translational modifications consist of chemical changes in proteins during or after synthesis. These modifications influence protein structure and activity and include protein cleavage, covalent addition of chemical groups such as phosphate (phosphorylation), acetyl (acetylation), palmitoyl (palmitoylation), among others. Noxs are regulated by various post-translational modifications, and many isoforms, except Nox5, are heavily glycosylated. Nox5 is regulated by phosphorylation, oxidation and nitrosylation, and recent studies indicate that it is also SUMOylated and palmitoylated [4].

4.5.1 Nox5 Phosphorylation Serine and threonine residues (Ser490, Thr494, Ser498, Ser475, Ser502, and Ser675) in the C-terminal domain of Nox5 are phosphorylated by numerous kinases including PKCα [34], ERK1/2) [35], Ca2+/calmodulin-dependent protein kinase II (CAM kinase II) [36] and tyrosine kinase c-Abl [11]. In addition, it has been demonstrated that Nox5 is regulated by the toll-like receptor pathway in podocytes, and that interleukin-1/4 receptor-associated kinase (IRAK1/ 4) influences Nox5 activity, potentially through Nox5 phosphorylation [37]. Nox5 phosphorylation may be another mechanism of Ca2+ sensitization, leading to increase in enzyme activity [6]. 4.5.2 Nox5 Oxidation Proteins can undergo oxidation by direct interaction with reactive oxygen species or with byproducts of oxidative stress. Several cysteine and methionine residues, highly susceptible to oxidative posttranslational modifications, are present in Nox5 [38]. Protein oxidation leads to structural changes in protein and consequently modifications in protein function. Nox5 oxidation results in enzyme inactivation. In the presence of high levels of reactive oxygen species, cysteine and methionine residues in the Ca2+-binding EF domain of Nox5 are oxidized, reducing Ca2+ binding. In addition, oxidation alters Nox5 secondary and tertiary structure, which decrease enzyme activation [39]. Nox5 oxidation leading to enzyme inactivation may be a potential protective mechanism against oxidative stress. 4.5.3 Nox5 S-Nitrosylation Protein S-nitrosylation is another oxidative posttranslational modification caused by reaction of cysteine residues in proteins with nitric oxide. S-nitrosylation sites (C107, C246, C519 and C694) are found in Nox5 by biotin switch assay and mass spectrometry [7]. Nitric oxide promotes reversible Nox5 S-nitrosylation leading to reduced enzyme activity. Reactive oxygen species generation by Nox5 is

220

L. L. Camargo et al.

reduced by exogenous and endogenous nitric oxide in a dosedependent manner, effects that are blocked by NOS inhibitors [7]. Therefore, S-nitrosylation may have a protective effect against Nox5-mediated oxidative stress.

4.5.4 SUMOylation of Nox5 Small ubiquitin-like modifier (SUMO) family of proteins can be conjugated to lysine residues in proteins, a posttranslational modification termed SUMOylation. Protein SUMOylation is a reversible process that unlike ubiquitin does not target protein for degradation, but rather affects protein stability, solubility, localization, and function [40]. SUMOylation modulates Nox5-dependent O2•- generation as increased expression of SUMO1 inhibits Nox5 activity while SUMO1 silencing increases Nox5-induced O2•production [41]. However, direct evidence of Nox5 SUMOylation has not yet been demonstrated. Mechanisms underlying the effect of SUMOylation on Nox5 activity are still unclear but may involve effects on other molecules including Nox5 regulatory proteins [12].

4.6

Genetic Regulation

The human Nox5 gene is found on chromosome 15, that undergoes differential splicing and alternative promoter use to generate the six different Nox5 isoforms [42]. The differences between the Nox5 isoforms are found in the N-terminus region, affecting its catalytic activity. Nox5α, Nox5β and Nox5γ are functionally active and generate reactive oxygen species, while Noxδ, Nox5ε and Noxζ seem to be inactive and do not generate detectable levels of reactive oxygen species.The main isoforms expressed in human cells are Nox5α and Nox5β and appear to be negatively regulated by Nox5ε, which inhibits Nox5-induced reactive oxygen species production [43]. Epigenetic changes, such as overexpression of histone deacetylase 2, increases the Nox5 gene promotor activity in vascular smooth muscle cells [44]. Several polymorphisms have been described within the coding sequence of human Nox5 gene that may influence Nox5 activity. Studies on single nucleotide polymorphisms (SNPs) within the coding region of Nox5 in Cos cells expressing various Nox5αβ mutants identified seven SNPs that are associated with reduced enzymatic function [45, 46].

5

Nox5 as a Proton Pump and Effect on Intracellular pH

During Nox activation, the electron donor NADPH provides two electrons that are translocated across the membrane to reduce O2 to O2•- and hence the primary function of Nox5 is the generation of O2•-. During this process two protons are

generated, which influence intracellular pH. Because of this, it has been suggested that Nox5 may have a dual function, firstly the transport of electrons to generate O2•-, and secondly the conductance of protons across membranes [4]. In phagocytes, Nox activation induces production of large quantities of reactive oxygen species, which leads to depolarization and decreased intracellular pH. These effects are directly linked to oxidase activity and seem to be especially important in intracellular pH regulation in plants and fungi [46]. With the discovery of Nox2, it was suggested that Nox2 is a phagocyte proton pump [4, 47]. However this notion has not been confirmed and more recently, specific voltage gated proton channels (Hv1) that associate with Noxs, were found to be responsible for conductance of protons [47]. It seems that Nox moves electrons while Hv1 transports protons in a coordinated manner. These transporters are closely interrelated and seem to have a symbiotic association where Hv1 is required for optimal generation of reactive oxygen species by Nox2 and Hv1 is influenced by Nox-induced electrogenic H+ efflux [47, 48]. While most studies showing this association focused on Nox2, other Noxs, and especially Nox5, may also be regulated by Hv1 and pH [4]. In addition to activated Nox5 having an effect on intracellular pH, changes in cellular acidic status may influence Nox5 itself, especially in the context of cancer. In esophageal cancer cells, exposure to acidic conditions increased Nox5 expression through ERK1/2, Rho kinase and cAMP response element binding protein [49]. This was attenuated by proton pump inhibitor treatment [49]. Based on these findings it has been suggested that in esophageal cancer, acid-induced Nox5 activation induces oxidative stress and acidification of cells, leading to cell injury and inflammation, processes that may be managed clinically with proton pump inhibitors (PPIs) [49]. This however awaits confirmation and the exact role of Nox5 as a proton pump still needs to be confirmed.

6

Cellular and Tissue Distribution of Nox5

The Nox5 gene is expressed in different tissues such as testes, spleen, lymph nodes, pancreas, placenta, bone marrow, uterus, vessels, kidney and stomach. In humans, Nox5 is abundant in cancer cells and seems to be the major source of reactive oxygen species in spermatozoa, endothelial cells, vascular smooth muscle cells and fibroblasts.

7

Subcellular Localization of Nox5

Within cells, Nox5 is found mainly in the perinuclear area, in a pattern consistent with the endoplasmic reticulum (ER) [50]. The ER is a site of protein synthesis, posttranslational modification and an important store of

13

Nox5: Molecular Regulation and Pathophysiology

intracellular calcium. Considering the calcium-sensitivity of Nox5, it is not surprising that it may localize in areas where calcium is abundant. In addition, reactive oxygen species are produced in the ER as byproducts of protein folding and ER function is redox-sensitive. Nox5 function in the ER is not yet clear, however it may be an area of crosstalk between calcium- and redox-sensitive signaling [51]. Nox5 seems to traffic between subcellular compartments to localize in specific sites where it associates with regulatory proteins. In particular Nox5 traffics from the ER to the cell membrane, possibly through processes involving actinbinding proteins. Molecular mechanisms governing Nox5 trafficking may involve the small GTPase Sar1 [52]. Additionally, Nox5 binds to phosphatidylinositol 4,5-bisphosphate, a regulatory lipid in the plasma membrane, through polybasic domains in its N-terminus resulting in translocation from internal membranes to the plasma membrane [9]. In the plasma membrane, Nox5 localizes in cholesterol-rich microdomains (lipid rafts/caveolae) [37], leading to association with regulatory proteins such as PKC, c-Src and Abl that influence its activation [42].

8

Nox5 and Signaling

Nox5, by increasing generation of reactive oxygen species, mediates signalling in several cell types through modification of kinases, transcription factors, ion channels and other signalling molecules [38] (Fig. 13.3). This occurs primarily via changes in the oxidative state of proteins. Among the many redox sensitive targets of Nox5 is the tyrosine kinase c-Src. While c-Src influences Nox5 activation, it is also regulated by Nox5-induced production of reactive oxygen species. Accordingly, c-Src is both upstream and downstream of Nox5. This circuitous system is especially important in human cancer cells and in vascular smooth muscle cells [53, 54]. Nox5-dependent activation of c-Src in vascular smooth muscle cells plays a key role in cytoskeletal organization and pro-contractile signalling [54].

9

Physiological Role of Nox5

The physiological role of Nox5 has been extensively studied in arthropods. In drosophila melanogaster the Nox5 analogue, dNox, is involved in hormone-stimulated smooth muscle contraction and laying of eggs [55]. In the arthropod rhodnius prolixus, a blood sucking insect, Nox5 (dNox) plays an important role in gastrointestinal smooth muscle contraction and peristalsis, hemoglobin digestion, egg production and hemolymph urate levels [56]. However, in humans the exact (patho)physiological role of Nox5 still remains unclear although it is likely involved in calcium-

221

activated, redox-dependent processes. Nox5 seems to be important in spermatozoa motility and sperm-oocyte fusion, while pathologically Nox5 has been implicated in many disease including cancer, cardiovascular disease and kidney disease [3].

9.1

Nox5 and Sperm Function

Nox5 is abundantly expressed in human and equine testes and spermatozoa [1, 57]. These findings, together with the fact that physiological processes in spermatozoa involve production of reactive oxygen species, suggested an important role for Nox5 in spermatozoa function. Of the Nox isoforms, only Nox5 is expressed in human testes. Nox5-induced generation of O2•- controls spermatozoa motility and spermoocyte fusion [10]. In spermatozoas, Nox5 is localized in the flagella region and acrosome and its function is regulated by the tyrosine kinase c-Abl and the Hv1 proton channel, which influence motility and sperm maturation [1, 10]. Nox5 is an important determinant of the redox state of human spermatozoa and ovum-fertilizing capacity. Accordingly, it has been suggested that targeting Nox5 may be an interesting strategy as a male contraceptive.

9.2

Nox5 and Contraction

The close association between Nox5 and calcium suggests that Nox5 would be an important element in the regulation of cellular processes regulated by calcium, including cell contraction, motility, and cytoskeletal organization. In support of this, Nox5 has been shown to be associated with activation of the cellular contractile machinery and pro-contractile signaling pathways including activation of calcium channels, phosphorylation of myosin light chain 20 (MLC20), activation of Rho kinase and inhibition of the myosin light-chain phosphatase regulatory subunit, myosin phosphatase target subunit 1 (MYPT1) [51]. In addition, in mice expressing human Nox5 in a vascular smooth muscle-specific manner, vascular contraction was amplified, pro-contractile vascular signaling was upregulated and cells underwent cytoskeletal reorganization [51]. These processes involve tyrosine kinases, such as c-Src, which influence Nox5-induced actin polymerization, and phosphorylation of MLC20 and focal adhesion kinase (FAK) in human vascular smooth muscle cells and contraction in Nox5 transgenic mice [54]. A role for Nox5 in contraction is further confirmed in Drosophila, since deletion of dNox leads to diminished ovarian muscle contraction and retention of mature eggs in ovaries and sterility [55]. Moreover, siRNA downregulation of dNox in the arthropod rhodnius prolixus attenuated gastrointestinal contraction [51].

222

L. L. Camargo et al.

Fig. 13.3 Role of Nox5 in cell signalling. Nox5 is activated by Ca2+ and produces reactive oxygen species (ROS) by the reduction of O2 to superoxide (O2•-), which in turn is dismutated to generate hydrogen peroxide (H2O2) spontaneously or catalysed by superoxide dismutase (SOD). Nox5 activity is also regulated by posttranslational modifications, such as phosphorylation by kinases PKC, c-Abl and ERK1/2, and interaction with protein partners calmodulin and Hsp90. ROS levels are normally controlled by many antioxidant systems, which

10

Nox5 and Disease

Although the physiological relevance of Nox5 in the cardiovascular system is still under investigation, many pre-clinical and clinical studies have associated Nox5 with cardiovascular and cerebrovascular diseases, heart and kidney disease and cancer (Fig. 13.4).

10.1 Hypertension Two genetic studies demonstrated that Noxs are associated with blood pressure control [58, 59], where putative roles for

maintain redox balance. Increased generation of ROS causes oxidative posttranslational modifications (OxPTM) of numerous proteins that influence redox signalling and cell function. Abbreviations: CAT, catalase; Prdx, peroxiredoxins; Trx, thioredoxins; GPx, gluthathione peroxidases; Nrf2, Nuclear factor erythroid 2-related factor 2; MMP, matrix metalloproteinases; MAPK, mitogen-activated protein kinases; GPCR, G protein-coupled receptor; GFR, growth factor receptor

Nox4 and Nox5 in hypertension have been suggested [58]. Nox5 expression is observed in human arteries and veins and is increased in arteries and vascular cells from hypertensive subjects [54, 60]. The predominant isoform, that is catalytically active, in the endothelium and vascular media layer of human vessels are Nox5α and Nox5β [60]. Aged mice expressing human Nox5 in endothelial cells develop hypertension, and have associated endothelial dysfunction due eNOS uncoupling [32]. Impaired endothelium-induced relaxation, a hallmark of vascular dysfunction in hypertension, is also observed in Nox5-vascular smooth muscle cell transgenic mice [51], suggesting that Nox5-induced hypercontractility is also associated with changes in endothelial cell function. Moreover, increased

13

Nox5: Molecular Regulation and Pathophysiology

223

Fig. 13.4 Pathophysiological role of Nox5. Nox5 is key regulator of redox- dependent signalling in many human cells. In pathological conditions increased Nox5 expression and activity lead to oxidative stress and aberrant redox dependent signalling. Increased Nox5 activation influences cellular processes such as contraction, proliferation, remodelling, inflammation, and fibrosis. These effects contribute to development of cardiovascular diseases, renal pathologies and cancer

expression of endothelial Nox5 causes an increase in retinal vascular permeability and expression of markers of angiogenesis and inflammation, important in retinopathies associated with diabetes and hypertension [61]. Nox5 activation is enhanced by many pro-hypertensive peptides. Angiotensin II (Ang II) and endothelin-1 (ET-1) induce redox signaling via Nox5-derived O2•- generation in human microvascular endothelial cells [62] and vascular smooth muscle cells [54], leading to proliferation, migration and inflammation, processes important in vascular remodelling in cardiovascular disease. In preeclampsia, a hypertensive disorder of pregnancy where an incomplete remodelling of spiral arteries occurs leading to reduced uterine blood flow and blood supply to the fetus, Nox5 DNA methylation was increased in regions coding for calcium binding [63]. Although the functional significance of these findings in relation to Nox5 biology awaits further understanding, hypermethylation of Nox5 usually leads to decreased gene transcription [64], which in turn may contribute to the abnormal uterine vascular remodelling observed in preeclampsia. We identified hypospadias as a novel risk factor for hypertension. The underlying molecular mechanisms involve vascular dysfunction and Nox5 upregulation. In boys with hypospadias, vascular contractility of dermal resistance

arteries was increased, an effect restored by Nox5 inhibition [50]. In addition, Nox5 expression and DNA methyltransferase activity were enhanced in vascular smooth muscle cells extracted from vascular tissue of hypospadias individuals [65]. Epigenetic modification is an important mechanism of regulation of gene expression, such as regulation of histone acetylation via histone deacetylases (HDACs). Previous studies show that Nox gene expression can be enhanced via HDACs [66] and HDAC inhibition decreases Nox5 expression in human aortic cells exposed to high glucose [67] and human lung fibroblasts and macrophages [68].

10.2 Atherosclerosis Human studies demonstrate that Nox5 is involved in vascular inflammatory diseases such as atherosclerosis and aortic aneurysm [69, 70]. In human coronary artery disease, calcium-dependent O2•- generation and Nox5 expression in endothelial and vascular smooth muscle cells were increased compared to control arteries [70]. Mechanisms associated with Nox5 and the pathogenesis of atherosclerosis are still unclear. In porcine vascular smooth muscle cells, Nox5 knockdown reduced fibroblast growth factor-induced activator protein 1 (AP-1) regulated transcription of calcium-

224

activated K+ channels and vascular smooth muscle cell migration [71]. Inflammation plays a major role in atherosclerosis and Nox5 is expressed in immune cells, where it has been shown to play a role in monocyte differentiation into dendritic cells [72]. Exposure of monocytes and monocyte-derived macrophages to pro-inflammatory and pro-atherogenic molecules, interferon-γ and oxidized low density lipoprotein (LDL), induce Nox5 expression and reactive oxygen species production increase [44]. Additionally, in human atherosclerotic plaques, Nox5 expression co-localized with histone acetyltransferase (HAT) enzymes in macrophages; and inhibition of HATs decreased Nox5 expression in human macrophages exposed to lipopolysaccharides, suggesting that epigenetic regulation of Nox5 may also play a role in the pathogenesis of atherosclerosis [73]. Of interest, studies in Nox5-deficient rabbits fed a cholesterol rich diet showed that plaque formation was increased in thoracic aortas, despite no differences in blood pressure or vascular contractility [74], suggesting that Nox5 may have different roles during the disease process and is vascular protective. The functional significance of Nox5 in atherosclerosis may also be species specific as not all experimental models of atherosclerosis suggested a role for Nox5, as in primates Nox2 instead of Nox5 was involved in atherosclerotic vascular injury [75].

10.3 Aortic Aneurysm and Vascular Calcification Abdominal aortic aneurysm is a major cause of cardiovascular morbidity and mortality, especially during aging. This involves changes in vascular phenotype to a pro-inflammatory and pro-fibrotic state. In addition, vascular calcification due to vascular smooth muscle cell switching to a synthetic phenotype may be important. Noxs have been implicated in these processes [76]. In particular in human aortic aneurysms, expression of Nox5 is increased and this is associated with oxidative stress, inflammation and calcification [77]. Nox5 may be especially important in vascular calcification considering its dependency on calcium for its activation. Nox5 has been identified as a key regulator of vascular smooth muscle cell function and differentiation and plays a role in calcium uptake via extracellular vesicles [78]. Calcium induces Nox5-mediated production of reactive oxygen species, which in turn influence extracellular vesicles, that are involved in calcification [78]. These processes seem to be amplified by nicotine and smoking, which are associated with redox-sensitive vascular damage and aneurysm formation [79]. In the context of diabetes endothelial Nox5 seems to play an important role in aortic aneurysm formation, but not in atherosclerotic disease. In pro-atherosclerotic, diabetic mice

L. L. Camargo et al.

in which human Nox5 was expressed in an endothelialspecific manner, there was no aggravation of atherosclerosis, but the rate of aneurysm formation was doubled [80]. This was attributed to direct effects of Nox5-induced generation of reactive oxygen species on collagen deposition and extracellular matrix remodelling.

10.4 Small Vessel Disease of the Brain In small vessel disease of the brain, Nox5 expression has been associated with stroke [81] and cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) [82]. In a mouse model where Nox5 expression was induced in endothelial and hematopoietic cells followed by a transient occlusion of the middle cerebral artery and 24 h reperfusion of the brain, the tMCAO stroke model, blood brain barrier (BBB) disruption, infarct size and neuromotor function were worsened compared to wild-type mice [81]. Moreover, in another study utilizing mice expressing Nox5 specifically in endothelial cells, the expression of markers of BBB integrity, zonula occludens-1 and occluding, were diminished followed by memory impairment in aged Nox5 mice [83]. CADASIL, the major cause of monogenetic stroke due to mutations of neurogenic locus notch homolog protein 3 (Notch3), is a disease of small vessels, typically of the brain, but also peripherally [84]. Peripheral resistance arteries obtained from patients with CADASIL exhibit vascular dysfunction, hypertrophic remodelling, aberrant Rho kinase activity and ER stress with associated increase in Nox5 expression and activity [82]. Nox5 inhibition attenuated vascular dysfunction and impaired signaling in CADASIL patients suggesting an important role for Nox5 in this condition [82]. Exact mechanisms whereby Notch3 mutation in CADASIL influence Nox5 await clarification.

10.5 Heart Disease In the heart, Nox5 expression is increased in cardiomyocytes and endothelial cells in cardiac tissue from patients with heart failure undergoing heart transplantation [85]. In experimental models, neonatal rat cardiomyocytes expressing Nox5 had an exaggerated response to Ang II-induced hypertrophy, characterized by increased surface area and expression of atrial natriuretic peptide, brain natriuretic peptide and β-myosin heavy chain (β-MHC) [85]. In a mouse model of Nox5 expression targeted to cardiomyocytes, Nox5 aggravated pressure-overload and Ang II-induced heart hypertrophy and fibrosis with associated decrease in heart contractility [85].

13

Nox5: Molecular Regulation and Pathophysiology

10.6 Kidney Disease In kidneys, upregulation of Nox5 is observed in diabetic nephropathy influencing renal function and blood pressure [86]. In mice expressing Nox5 specifically in podocytes, diabetes induction via streptozotocin, led to podocyte dysfunction, albuminuria and hypertension through mechanisms involving reactive oxygen species and toll-like receptors activation [87, 88]. Similar findings are observed when Nox5 expression is targeted to endothelial cells, vascular smooth muscle cells or mesangial cells [89], where Nox5 intensified the effects of diabetes on glomerulosclerosis, inflammation and fibrosis. Increased Nox5 expression and activity is also observed in renal proximal tubule cells from hypertensive patients [90] and in sepsis-induced acute kidney injury [91].

10.7 Hirschsprung Disease and Nox5 Hirschsprung disease is a congenital disorder where ganglionic cells in the distal gastrointestinal tract are missing. Clinically this is associated with tonic contraction and functional intestinal obstruction [92]. Genome-wide studies identified Nox5 as a candidate risk gene [92]. Association studies further demonstrated that Nox5 polymorphisms are linked to increased susceptibility to Hirschsprung disease [93]. Studies in zebrafish failed to demonstrate a pathogenetic role for Nox5 in Hirschsprung disease, although aganglionic segments of the gastrointestinal tract in Hirschsprung disease exhibited a significant decrease in Nox5 expression [94]. The functional significance of Nox5 variants and Hirschsprung disease is still unclear.

10.8 Nox5 and Cancer Overexpression and uncontrolled activity of Nox5 has been described in a variety of cancers, such as gastric cancer, malignant melanoma, breast cancer, prostate cancer, lung, brain, ovary and oesophageal cancer [95–99]. Exposure of Barrett’s esophagial cells to an acidic environment, causes an increase in expression of the enzyme DNA methyltransferase-1 in a Nox5- nuclear factor kappa-lightchain-enhancer of activated B cells (NFkB)-dependent manner [100]. Barrett’s esophagitis is a precancerous state and DNA methyltransferase-1 plays an important role in cancer development. In esophageal squamous cell carcinoma, Nox5 induced fibroblast and adipose-derived mesenchymal stem cells to differentiate into cancer-associated fibroblasts and release pro-inflammatory cytokines, facilitating malignancy

225

[101]. This effect of Nox5 was associated with c-Src/NFkB activation leading to secretion of interleukin 1β (IL1β), tumor necrosis factor α (TNFα) and lactate [101]. In terms of malignancy, Nox5 has also been associated with lymphoma aggressiveness [102], and effects may be associated with Nox5-induced resistance to apoptosis observed in lymphoma cell lines [103]. In prostate cancer, Nox5 is associated with elevated reactive oxygen species production and proliferation, via activation of PKCζ and c-Jun N-terminal kinase (JNK) [104]. Of interest, Nox5 has been associated with cancer cell sensitivity to chemotherapies, such as cisplatin due to the fact that Nox5 expression is increased by cisplatin in skin, breast and lung cancer cells which is associated with oxidative stress-induced cell death [105, 106]. In human ovarian adenocarcinoma cells, cisplatin resistance was related to a decrease in Nox5 expression followed by an increase in the gene expression of antioxidant enzymes [107].

11

Conclusions

Calcium-dependent Nox5 emerged early during evolution and is the most ancient member of the Nox family [108]. Noxs are distinctive in that they transport electrons across membranes to produce O2•- and/or H2O2, which are vital signaling molecules that influence cell function. Regulation, mechanisms of activation and tissue distribution of the Nox isoforms are distinct. This is especially relevant for Nox5, which is unique in that it generates O2•- from a single gene product, it does not require any NADPH oxidase subunits for its activation, it has a unique N-terminal extension that contains calcium-binding domains and it is not glycosylated. Moreover, it is not expressed in most rodents. Physiologically Nox5-induced O2•- generation regulates sperm function and smooth muscle contraction, while pathologically it has been implicated in cardiovascular disease, stroke, kidney and heart disease and cancer. The field of Nox5 pathophysiology is still young, but with advancements in Nox5 biochemistry and molecular and cell biology, development of new experimental models, further characterization of the Nox5 crystal structure and elucidation of the functional significance of Nox5 mutations and polymorphisms, the significance of Nox5 in human health and disease will become evident. Acknowledgements The authors have received funds from the British Heart Foundation (BHF) (RG/13/7/30099, RE/13/5/30177). RMT is supported through the Dr. Phil Gold Chair, McGill University, Montreal.

Conflicts There are no conflicts to declare.

226

References 1. Banfi B, Molnar G, Maturana A et al (2001) A Ca(2+)-activated NADPH oxidase in testis, spleen, and lymph nodes. J Biol Chem 276:37594–37601 2. Cheng G, Cao Z, Xu X et al (2001) Homologs of gp91phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269:131–140 3. Bánfi B, Tirone F, Durussel I et al (2004) Mechanism of Ca2+ activation of the NADPH oxidase 5 (NOX5). J Biol Chem 279: 18583–18591 4. Touyz RM, Anagnostopoulou A, Rios F et al (2019) NOX5: molecular biology and pathophysiology. Exp Physiol 104:605–616 5. Chen F, Haigh S, Yu Y et al (2015) Nox5 stability and superoxide production is regulated by C-terminal binding of Hsp90 and CO-chaperones. Free Radic Biol Med 89:793–805 6. Jagnandan D, Church JE, Banfi B et al (2007) Novel mechanism of activation of NADPH oxidase 5. calcium sensitization via phosphorylation. J Biol Chem 282:6494–6507 7. Qian J, Chen F, Kovalenkov Y et al (2012) Nitric oxide reduces NADPH oxidase 5 (Nox5) activity by reversible S-nitrosylation. Free Radic Biol Med 52:1806–1819 8. Panday A, Sahoo MK, Osorio D et al (2015) NADPH oxidases: an overview from structure to innate immunity-associated pathologies. Cell Mol Immunol 12:5–23 9. Kawahara T, Lambeth JD (2008) Phosphatidylinositol (4, 5)bisphosphate modulates Nox5 localization via an N-terminal polybasic region. Mol Biol Cell 19:4020–4031 10. Musset B, Clark RA, DeCoursey TE et al (2012) NOX5 in human spermatozoa: expression, function, and regulation. J Biol Chem 287:9376–9388 11. El Jamali A, Valente AJ, Lechleiter JD et al (2008) Novel redoxdependent regulation of NOX5 by the tyrosine kinase c-Abl. Free Radic Biol Med 44:868–881 12. Fulton DJR (2019) The molecular regulation and functional roles of NOX5. Methods Mol Biol 1982:353–375 13. Rizvi F, Heimann T, O'Brien WJ (2012) Expression of NADPH oxidase (NOX) 5 in rabbit corneal stromal cells. PLoS One 7: e34440 14. Miles JA, Egan JL, Fowler JA et al (2021) The evolutionary origins of peroxynitrite signalling. Biochem Biophys Res Commun 580: 107–112 15. Zhang X, Krause KH, Xenarios I et al (2013) Evolution of the ferric reductase domain (FRD) superfamily: modularity, functional diversification, and signature motifs. PLoS One 8:e58126 16. Bedard K, Jaquet V, Krause KH (2012) NOX5: from basic biology to signaling and disease. Free Radic Biol Med 52:725–734 17. Kawahara T, Quinn MT, Lambeth JD (2007) Molecular evolution of the reactive oxygen-generating NADPH oxidase (Nox/Duox) family of enzymes. BMC Evol Biol 7:109 18. Magnani F, Nenci S, Millana Fananas E et al (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci U S A 114:6764–6769 19. Kawahara T, Jackson HM, Smith SM et al (2011) Nox5 forms a functional oligomer mediated by self-association of its dehydrogenase domain. Biochemistry 50:2013–2025 20. Tirone F, Radu L, Craescu CT et al (2010) Identification of the binding site for the regulatory calcium-binding domain in the catalytic domain of NOX5. Biochemistry 49:761–771 21. Millana Fananas E, Todesca S, Sicorello A et al (2020) On the mechanism of calcium-dependent activation of NADPH oxidase 5 (NOX5). FEBS J 287:2486–2503 22. Wei CC, Fabry E, Hay E et al (2020) Metal binding and conformational studies of the calcium binding domain of NADPH oxidase

L. L. Camargo et al. 5 reveal its similarity and difference to calmodulin. J Biomol Struct Dyn 38:2352–2368 23. Aravind P, Chandra K, Reddy PP et al (2008) Regulatory and structural EF-hand motifs of neuronal calcium sensor-1: Mg 2+ modulates Ca 2+ binding, Ca 2+ -induced conformational changes, and equilibrium unfolding transitions. J Mol Biol 376: 1100–1115 24. Tirone F, Cox JA (2007) NADPH oxidase 5 (NOX5) interacts with and is regulated by calmodulin. FEBS Lett 581:1202–1208 25. Smith D, Lloyd L, Wei E et al (2022) Calmodulin binding to the dehydrogenase domain of NADPH oxidase 5 alters its oligomeric state. Biochem Biophys Rep 29:101198 26. Chen F, Barman S, Yu Y et al (2014) Caveolin-1 is a negative regulator of NADPH oxidase-derived reactive oxygen species. Free Radic Biol Med 73:201–213 27. Wang RY, Noddings CM, Kirschke E et al (2022) Structure of Hsp90-Hsp70-Hop-GR reveals the Hsp90 client-loading mechanism. Nature 601:460–464 28. Sweeny EA, Schlanger S, Stuehr DJ (2020) Dynamic regulation of NADPH oxidase 5 by intracellular heme levels and cellular chaperones. Redox Biol 36:101656 29. Khechaduri A, Bayeva M, Chang HC et al (2013) Heme levels are increased in human failing hearts. J Am Coll Cardiol 61:1884– 1893 30. Chen F, Yu Y, Qian J et al (2012) Opposing actions of heat shock protein 90 and 70 regulate nicotinamide adenine dinucleotide phosphate oxidase stability and reactive oxygen species production. Arterioscler Thromb Vasc Biol 32:2989–2999 31. Zhang Q, Malik P, Pandey D et al (2008) Paradoxical activation of endothelial nitric oxide synthase by NADPH oxidase. Arterioscler Thromb Vasc Biol 28:1627–1633 32. Elbatreek MH, Sadegh S, Anastasi E et al (2020) NOX5-induced uncoupling of endothelial NO synthase is a causal mechanism and theragnostic target of an age-related hypertension endotype. PLoS Biol 18:e3000885 33. Montezano AC, Touyz RM (2012) Reactive oxygen species and endothelial function--role of nitric oxide synthase uncoupling and Nox family nicotinamide adenine dinucleotide phosphate oxidases. Basic Clin Pharmacol Toxicol 110:87–94 34. Chen F, Yu Y, Haigh S et al (2014) Regulation of NADPH oxidase 5 by protein kinase C isoforms. PLoS One 9:e88405 35. Pandey D, Fulton DJ (2011) Molecular regulation of NADPH oxidase 5 via the MAPK pathway. Am J Physiol Heart Circ Physiol 300:H1336–H1344 36. Pandey D, Gratton JP, Rafikov R et al (2011) Calcium/calmodulindependent kinase II mediates the phosphorylation and activation of NADPH oxidase 5. Mol Pharmacol 80:407–415 37. Anagnostopoulou A, Camargo LL, Rodrigues D et al (2020) Importance of cholesterol-rich microdomains in the regulation of Nox isoforms and redox signaling in human vascular smooth muscle cells. Sci Rep 10:17818 38. Pendyala S, Natarajan V (2010) Redox regulation of Nox proteins. Respir Physiol Neurobiol 174:265–271 39. Petrushanko IY, Lobachev VM, Kononikhin AS et al (2016) Oxidation of capital ES, Cyrillicsmall a, Cyrillic2+-binding domain of NADPH oxidase 5 (NOX5): toward understanding the mechanism of inactivation of NOX5 by ROS. PLoS One 11:e0158726 40. Yang Y, He Y, Wang X et al (2017) Protein SUMOylation modification and its associations with disease. Open Biol 7:170167 41. Pandey D, Chen F, Patel A et al (2011) SUMO1 negatively regulates reactive oxygen species production from NADPH oxidases. Arterioscler Thromb Vasc Biol 31:1634–1642 42. Serrander L, Jaquet V, Bedard K et al (2007) NOX5 is expressed at the plasma membrane and generates superoxide in response to protein kinase C activation. Biochimie 89:1159–1167

13

Nox5: Molecular Regulation and Pathophysiology

43. Fulton DJ (2009) Nox5 and the regulation of cellular function. Antioxid Redox Signal 11:2443–2452 44. Manea SA, Todirita A, Raicu M et al (2014) C/EBP transcription factors regulate NADPH oxidase in human aortic smooth muscle cells. J Cell Mol Med 18:1467–1477 45. Wang Y, Chen F, Le B et al (2014) Impact of Nox5 polymorphisms on basal and stimulus-dependent ROS generation. PLoS One 9: e100102 46. Segal AW (2016) NADPH oxidases as electrochemical generators to produce ion fluxes and turgor in fungi, plants and humans. Open Biol 6:50 47. Henderson LM, Chappell JB, Jones OT (1987) The superoxidegenerating NADPH oxidase of human neutrophils is electrogenic and associated with an H+ channel. Biochem J 246:325–329 48. DeCoursey TE (2016) The intimate and controversial relationship between voltage-gated proton channels and the phagocyte NADPH oxidase. Immunol Rev 273:194–218 49. Li D, Deconda D, Li A et al (2019) Effect of proton pump inhibitor therapy on NOX5, mPGES1 and iNOS expression in Barrett's Esophagus. Sci Rep 9:16242 50. Ahmarani L, Avedanian L, Al-Khoury J et al (2013) Whole-cell and nuclear NADPH oxidases levels and distribution in human endocardial endothelial, vascular smooth muscle, and vascular endothelial cells. Can J Physiol Pharmacol 91:71–79 51. Montezano AC, De Lucca CL, Persson P et al (2018) NADPH oxidase 5 is a pro-contractile Nox isoform and a point of cross-talk for calcium and redox signaling-implications in vascular function. J Am Heart Assoc 7:e009388 52. Kiyohara T, Miyano K, Kamakura S et al (2018) Differential cell surface recruitment of the superoxide-producing NADPH oxidases Nox1, Nox2 and Nox5: the role of the small GTPase Sar1. Genes Cells 23:480–493 53. Chen J, Wang Y, Zhang W et al (2020) Membranous NOX5derived ROS oxidizes and activates local Src to promote malignancy of tumor cells. Signal Transduct Target Ther 5:139 54. Camargo LL, Montezano AC, Hussain M et al (2022) Central role of c-Src in NOX5- mediated redox signalling in vascular smooth muscle cells in human hypertension. Cardiovasc Res 118:1359– 1373 55. Ritsick DR, Edens WA, Finnerty V et al (2007) Nox regulation of smooth muscle contraction. Free Radic Biol Med 43:31–38 56. Gandara ACP, Dias FA, de Lemos PC et al (2021) Urate and NOX5 control blood digestion in the hematophagous insect. Front Physiol 12:633093 57. Sabeur K, Ball BA (2007) Characterization of NADPH oxidase 5 in equine testis and spermatozoa. Reproduction 134:263–270 58. Kraja AT, Cook JP, Warren HR et al (2017) New blood pressureassociated loci identified in meta-analyses of 475 000 individuals. Circ Cardiovasc Genet 10:e001778 59. Han X, Hu Z, Chen J et al (2017) Associations between genetic variants of NADPH oxidase-related genes and blood pressure responses to dietary sodium intervention: the GenSalt study. Am J Hypertens 30:427–434 60. Pandey D, Patel A, Patel V et al (2012) Expression and functional significance of NADPH oxidase 5 (Nox5) and its splice variants in human blood vessels. Am J Physiol Heart Circ Physiol 302: H1919–H1928 61. Deliyanti D, Alrashdi SF, Touyz RM et al (2020) Nox (NADPH oxidase) 1, Nox4, and Nox5 promote vascular permeability and neovascularization in retinopathy. Hypertension 75:1091–1101 62. Montezano AC, Burger D, Paravicini TM et al (2010) Nicotinamide adenine dinucleotide phosphate reduced oxidase 5 (Nox5) regulation by angiotensin II and endothelin-1 is mediated via calcium/calmodulin-dependent, rac-1-independent pathways in human endothelial cells. Circ Res 106:1363–1373

227 63. Yeung KR, Chiu CL, Pidsley R et al (2016) DNA methylation profiles in preeclampsia and healthy control placentas. Am J Physiol Heart Circ Physiol 310:H1295–H1303 64. Zhu C, Yu ZB, Chen XH et al (2011) DNA hypermethylation of the NOX5 gene in fetal ventricular septal defect. Exp Ther Med 2: 1011–1015 65. Lucas-Herald AK, Montezano AC, Alves-Lopes R et al (2022) Vascular dysfunction and increased cardiovascular risk in hypospadias. Eur Heart J 43:1832–1845 66. Siuda D, Zechner U, El Hajj N et al (2012) Transcriptional regulation of Nox4 by histone deacetylases in human endothelial cells. Basic Res Cardiol 107:283 67. Manea SA, Antonescu ML, Fenyo IM et al (2018) Epigenetic regulation of vascular NADPH oxidase expression and reactive oxygen species production by histone deacetylase-dependent mechanisms in experimental diabetes. Redox Biol 16:332–343 68. Chen F, Li X, Aquadro E et al (2016) Inhibition of histone deacetylase reduces transcription of NADPH oxidases and ROS production and ameliorates pulmonary arterial hypertension. Free Radic Biol Med 99:167–178 69. Kigawa Y, Miyazaki T, Lei XF et al (2017) Functional heterogeneity of Nadph oxidases in atherosclerotic and aneurysmal diseases. J Atheroscler Thromb 24:1–13 70. Guzik TJ, Chen W, Gongora MC et al (2008) Calcium-dependent NOX5 nicotinamide adenine dinucleotide phosphate oxidase contributes to vascular oxidative stress in human coronary artery disease. J Am Coll Cardiol 52:1803–1809 71. Gole HK, Tharp DL, Bowles DK (2014) Upregulation of intermediate-conductance Ca2+-activated K+ channels (KCNN4) in porcine coronary smooth muscle requires NADPH oxidase 5 (NOX5). PLoS One 9:e105337 72. Manea A, Manea SA, Gan AM et al (2015) Human monocytes and macrophages express NADPH oxidase 5; a potential source of reactive oxygen species in atherosclerosis. Biochem Biophys Res Commun 461:172–179 73. Vlad ML, Manea SA, Lazar AG et al (2019) Histone acetyltransferase-dependent pathways mediate upregulation of NADPH oxidase 5 in human macrophages under inflammatory conditions: a potential mechanism of reactive oxygen species overproduction in atherosclerosis. Oxidative Med Cell Longev 2019: 3201062 74. Petheo GL, Kerekes A, Mihalffy M et al (2021) Disruption of the NOX5 gene aggravates atherosclerosis in rabbits. Circ Res 128: 1320–1322 75. Stanic B, Pandey D, Fulton DJ et al (2012) Increased epidermal growth factor-like ligands are associated with elevated vascular nicotinamide adenine dinucleotide phosphate oxidase in a primate model of atherosclerosis. Arterioscler Thromb Vasc Biol 32:2452– 2460 76. Siu KL, Li Q, Zhang Y et al (2017) NOX isoforms in the development of abdominal aortic aneurysm. Redox Biol 11:118–125 77. Guzik B, Sagan A, Ludew D et al (2013) Mechanisms of oxidative stress in human aortic aneurysms--association with clinical risk factors for atherosclerosis and disease severity. Int J Cardiol 168: 2389–2396 78. Furmanik M, Chatrou M, van Gorp R et al (2020) Reactive oxygen-forming Nox5 links vascular smooth muscle cell phenotypic switching and extracellular vesicle-mediated vascular calcification. Circ Res 127:911–927 79. Petsophonsakul P, Burgmaier M, Willems B et al (2022) Nicotine promotes vascular calcification via intracellular Ca2+-mediated, Nox5-induced oxidative stress, and extracellular vesicle release in vascular smooth muscle cells. Cardiovasc Res 118:2196–2210 80. Ho F, Watson AMD, Elbatreek MH et al (2022) Endothelial reactive oxygen-forming NADPH oxidase 5 is a possible player

228 in diabetic aortic aneurysm but not atherosclerosis. Sci Rep 12: 11570 81. Casas AI, Kleikers PW, Geuss E et al (2019) Calcium-dependent blood-brain barrier breakdown by NOX5 limits postreperfusion benefit in stroke. J Clin Invest 129:1772–1778 82. Neves KB, Harvey AP, Moreton F et al (2019) ER stress and Rho kinase activation underlie the vasculopathy of CADASIL. JCI Insight 4:e131344 83. Cortes A, Solas M, Pejenaute A et al (2021) Expression of endothelial NOX5 alters the integrity of the blood-brain barrier and causes loss of memory in aging mice. Antioxidants (Basel) 10: 1311 84. Fouillade C, Monet-Leprêtre M, Baron-Menguy C et al (2012) Notch signalling in smooth muscle cells during development and disease. Cardiovasc Res 95:138–146 85. Zhao GJ, Zhao CL, Ouyang S et al (2020) Ca(2+)-dependent NOX5 (NADPH oxidase 5) exaggerates cardiac hypertrophy through reactive oxygen species production. Hypertension 76: 827–838 86. Holterman CE, Thibodeau JF, Towaij C et al (2014) Nephropathy and elevated BP in mice with podocyte-specific NADPH oxidase 5 expression. J Am Soc Nephrol 25:784–797 87. Jha JC, Banal C, Okabe J et al (2017) NADPH oxidase Nox5 accelerates renal injury in diabetic nephropathy. Diabetes 66: 2691–2703 88. Holterman CE, Boisvert NC, Thibodeau JF et al (2019) Podocyte NADPH oxidase 5 promotes renal inflammation regulated by the toll-like receptor pathway. Antioxid Redox Signal 30:1817–1830 89. Jha JC, Dai A, Holterman CE et al (2019) Endothelial or vascular smooth muscle cell-specific expression of human NOX5 exacerbates renal inflammation, fibrosis and albuminuria in the Akita mouse. Diabetologia 62:1712–1726 90. Yu P, Han W, Villar VA et al (2014) Unique role of NADPH oxidase 5 in oxidative stress in human renal proximal tubule cells. Redox Biol 2:570–579 91. Ge QM, Huang CM, Zhu XY et al (2017) Differentially expressed miRNAs in sepsis-induced acute kidney injury target oxidative stress and mitochondrial dysfunction pathways. PLoS One 12: e0173292 92. Kim JH, Cheong HS, Sul JH et al (2014) A genome-wide association study identifies potential susceptibility loci for Hirschsprung disease. PLoS One 9:e110292 93. Shin JG, Seo JY, Seo JM et al (2019) Association analysis of NOX5 polymorphisms with Hirschsprung disease. J Pediatr Surg 54:1815–1819 94. Wang J, Xiao J, Meng X et al (2021) NOX5 is expressed aberrantly but not a critical pathogenetic gene in Hirschsprung disease. BMC Pediatr 21:153

L. L. Camargo et al. 95. Antony S, Jiang G, Wu Y et al (2017) NADPH oxidase 5 (NOX5)induced reactive oxygen signaling modulates normoxic HIF-1alpha and p27 (Kip1) expression in malignant melanoma and other human tumors. Mol Carcinog 56:2643–2662 96. Dho SH, Kim JY, Lee KP et al (2017) STAT5A-mediated NOX5-L expression promotes the proliferation and metastasis of breast cancer cells. Exp Cell Res 351:51–58 97. Kalatskaya I (2016) Overview of major molecular alterations during progression from Barrett's esophagus to esophageal adenocarcinoma. Ann N Y Acad Sci 1381:74–91 98. Roy K, Wu Y, Meitzler JL et al (2015) NADPH oxidases and cancer. Clin Sci (Lond) 128:863–875 99. Antony S, Wu Y, Hewitt SM et al (2013) Characterization of NADPH oxidase 5 expression in human tumors and tumor cell lines with a novel mouse monoclonal antibody. Free Radic Biol Med 65:497–508 100. Hong J, Li D, Wands J et al (2013) Role of NADPH oxidase NOX5-S, NF-kappaB, and DNMT1 in acid-induced p16 hypermethylation in Barrett's cells. Am J Phys Cell Phys 305: C1069–C1079 101. Chen J, Wang Y, Zhang W et al (2021) NOX5 mediates the crosstalk between tumor cells and cancer-associated fibroblasts via regulating cytokine network. Clin Transl Med 11:e472 102. Goncalves JDS, Carvalho FL, Coutinho I et al (2020) NADPH oxidase 5 upregulation is associated with lymphoma aggressiveness. Rev Assoc Med Bras (1992) 66:210–215 103. Carnesecchi S, Rougemont AL, Doroshow JH et al (2015) The NADPH oxidase NOX5 protects against apoptosis in ALK-positive anaplastic large-cell lymphoma cell lines. Free Radic Biol Med 84:22–29 104. Holl M, Koziel R, Schafer G et al (2016) ROS signaling by NADPH oxidase 5 modulates the proliferation and survival of prostate carcinoma cells. Mol Carcinog 55:27–39 105. Dho SH, Kim JY, Kwon ES et al (2015) NOX5-L can stimulate proliferation and apoptosis depending on its levels and cellular context, determining cancer cell susceptibility to cisplatin. Oncotarget 6:39235–39246 106. Park S, Oh SS, Lee KW et al (2018) NDRG2 contributes to cisplatin sensitivity through modulation of BAK-to-Mcl-1 ratio. Cell Death Dis 9:30 107. Kalinina EV, Andreev YA, Petrova AS et al (2018) Redoxdependent expression of genes encoding NADPH oxidase 5 and the key antioxidant enzymes during formation of drug resistance of tumor cells to cisplatin. Bull Exp Biol Med 165:678–681 108. Massari M, Nicoll CR, Marchese S, Mattevi A, Mascotti ML (2022) Evolutionary and structural analyses of the NADPH oxidase family in eukaryotes reveal an initial calcium dependency. Redox Biol 12(56):102436

DUOX1 and DUOX2, DUOXA1 and DUOXA2

14

Françoise Miot and Xavier De Deken

Abstract

Keywords

DUOX1 and DUOX2 constitute a subgroup of long (~1500 amino acids) seven transmembrane domain NADPH oxidases. In addition to the catalytic core common to NOX1–5, comprising NADPH- and FAD-binding sites, heme arrangement for electron transfer from NADPH across the membrane to O2, they possess a N-terminal extracellular peroxidase homologous domain followed by an intracellular loop with two Ca++ EF-hand binding sites. They produce H2O2 extracellularly when correctly processed with their DUOXA at the plasma membrane of the cell. DUOX1 and 2 were initially isolated from the thyroid. DUOX/DUOXA complexes produce H2O2 required as co-substrate for the thyroperoxidase involved in thyroid hormone synthesis. DUOX2 and to a lesser extent DUOXA2 genes are frequently mutated and non-functional variants are frequently associated with congenital hypothyroidism, but with variable penetrance and hypothyroid phenotypes ranging from transient to permanent hypothyroidism and partial to total iodide organification defect. DUOX1 and 2 are also expressed on epithelial surfaces of the airways, salivary gland ducts and DUOX2 along the gastrointestinal digestive tract. Associated with lactoperoxidase, they constitute an efficient host defense mechanism against bacterial and viral infections. In the gut, DUOX2 is robustly induced to neutralize microbial proliferation and to maintain immune homeostasis. Deleterious variants of DUOX2 associated with congenital hypothyroidism could therefore increase the susceptibility to develop inflammatory bowel disease.

DUOX · DUOXA · Thyroid hormone synthesis · Dityrosine crosslinking · Host defense

F. Miot (✉) · X. De Deken Institut de Recherche Interdisciplinaire en Biologie Humaine et Moléculaire (IRIBHM) DuoxLab, Faculty of Medicine, Université libre de Bruxelles, Brussels, Belgium e-mail: [email protected]; [email protected]

1

The Story of DUOX and the Thyroid: H2O2 Is Needed for Thyroid Hormone Synthesis

Thyroid is the endocrine gland responsible for thyroid hormone (TH), tetraiodothyronine, L-thyroxine (T4) and L-triiodothyronine (T3), synthesis. TH synthesis and secretion in the blood stream are under the control of the thyroid stimulating hormone (TSH), secreted by the pituitary, via the protein G –coupled TSH receptor. T4 and T3 are the only iodine-containing compounds with established physiologic significance as essential regulators of metabolic processes. TH synthesis is an oxidative process that takes place in the organized thyroid functional unit called the follicle. A single layer of polarized epithelial cells forms the envelope of this spherical structure with a central lumen containing colloidal thyroglobulin (TG), the storage site of iodinated molecules. Iodide is taken up into the thyrocyte through the basolateral located sodium/iodide symporter (NIS) where it is 20–50 fold concentrated. The central steps of TH synthesis take place at the apical membrane-lumen boundary of the follicular thyrocyte (Fig. 14.1) [1]. First, iodide is oxidized and covalently binds to TG tyrosyl residues. Secondly, iodinated tyrosyl residues (mono- and di-iodotyrosines, MIT and DIT, respectively) are coupled to form T4 (2 DIT) and T3 (1 MIT and 1 DIT) (Fig. 14.2). Iodide oxidation and coupling of iodotyrosines, called iodide organification, are catalyzed by the thyroperoxidase (TPO) anchored via its C-terminal transmembrane domain at the apical plasma membrane of the thyrocyte. TPO is a heme containing peroxidase that requires hydrogen peroxide as final electron acceptor. When TH are needed, iodinated TG (TGI) is internalized by micropinocytosis at the apical pole of thyrocytes, conveyed

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_14

229

230

F. Miot and X. De Deken

Fig. 14.1 Thyroid hormone synthesis in the thyroid follicle. At the basal pole of the thyrocyte, iodide is taken up from the blood through the Na+/I- symporter (NIS) and transported to the apical pole facing the lumen via iodide channel anoctamin-1(ANO1) and anion exchanger pendrin (PDS). At the apical membrane-lumen boundary thyroperoxidase (TPO) catalyzes, in the presence of H2O2 generated by DUOX/DUOXA complex, iodide oxidation, covalent binding of oxidized iodide to tyrosyl residues in thyroglobulin (TG-TGI) and

coupling of mono- and di-iodothyrosines (MIT, DIT) to form T3 and T4. After TGI endocytosis, T3 and T4 are released from TGI by proteolytic cleavage and secreted at the basal membrane in the blood through the monocarboxylate transporter 8 (MCT8). MIT and DIT are deiodinated by the iodotyrosine dehalogenase (DEHAL1) and the released iodide is recycled for hormone synthesis. Thyroid hormone synthesis and secretion are under the control of thyrotropin via the protein G-coupled TSH receptor (TSHr)

to endosomes and lysosomes and digested by proteases among which cathepsins B, D, H and L. After TG digestion, T3 and T4 are released in the blood stream at the basolateral membrane via thyroid hormone transporters, including the monocarboxylate transporter 8 (MCT8). Free MIT and DIT are deiodinated by the iodotyrosine dehalogenase 1 (IYD/DEHAL1) releasing iodide that is recycled for further TH synthesis.

H2O2 necessary for thyroid oxidative processes was suggested already in 1971 to be the product of a reaction occurring at the apical plasma membrane of the thyrocyte. This reaction would be catalyzed by an enzyme using NADPH as a substrate produced in the cytosol through the stimulation of the pentose phosphate cycle and activated by calcium [2]. Further in vitro biochemical studies in various thyroid models had shown that this enzyme is a membrane-

Fig. 14.2 Structural formulae of thyroid hormones, T4 and T3, and precursor compounds, MIT and DIT

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

bound NADPH-dependent flavoprotein [3–10] that produces extracellular H2O2 used as co-factor by TPO. H2O2 is the limiting step of protein iodination and therefore of TH synthesis when iodide supply is sufficient [11, 12]. In human thyroid, H2O2 production and iodination process are stimulated by the calcium-phosphatidylinositol pathway [13]. The quantity of H2O2 produced is huge especially in stimulated thyrocytes; it is comparable to the amount of the reactive oxygen species (ROS) produced by an activated leukocyte. While an activated leukocyte lives a few hours, thyrocytes have a lifespan of 7 years in adults [14]. Therefore, it was soon suggested that the activity of the H2O2 generator should be highly regulated and that the thyrocyte should develop effective defense systems to protect itself from oxidative damage.

2

DUOX1 and DUOX2: Cloning from the Thyroid

More than twenty years have passed between the initial biochemical characterization and the cloning of DUOX sequences encoding the catalytic enzymatic cores of the thyroid H2O2 generating system. Two independent studies uncovered almost simultaneously their molecular nature. Starting from a purified fraction of pig thyroid membranebound NADPH flavoprotein and using micro-sequences to design gene-specific primers for rapid amplification of the 3’ cDNA ends, Dupuy et al. isolated p138Tox in human (1210 amino acids) and pig (1207 amino acids). These proteins showed similarities with a predicted 1506-amino acid protein encoded by a Caenorhabditis elegans (C.elegans) gene. P138Tox turned out to be DUOX2 lacking the first 338 residues [15]. Simultaneously, De Deken et al. cloned two cDNAs encoding NADPH oxidases using a strategy based on the functional similarities between ROS generation in the leukocytes and the thyroid according to the hypothesis that one of the components of the thyroid system would belong to the known gp91phox/NOX2 enzyme displaying sequence similarities. Screenings of cDNA libraries constructed from primo-cultured human thyrocytes at low stringency with a NOX2 probe yielded two distinct human cDNA clones corresponding to sequences harboring open reading frames of 1551 (initially called ThOX1) and 1548 amino acids (ThOX2), respectively [16]. The encoded polypeptides displayed 83% sequence similarity and were clearly related to NOX2 (53 and 47% similarity over the conserved NOX catalytic core). The 500 amino acids of the N-terminus ecto-domain showed a similarity of 43% with TPO. The presence of this peroxidase homology domain (PHD) led to change the nomenclature of these two new NADPH oxidases hence called DUOX for DUal OXidase. However, mammal DUOX does not exhibit a peroxidase

231

activity presumably because their PHD lacks critical amino acid residues for the peroxidase function [17, 18]. PHD displays, on the other hand, cysteine residues essential for the proper maturation of DUOX [19–21]. This PHD is followed by a first transmembrane segment preceding a large cytosolic domain containing two Ca++ binding EF-hand motifs. Finally, the protein ends with the C-terminal NOX catalytic core portion. This conserved region encompasses the six transmembrane segments, including the four histidines and one arginine critical for heme binding, and the COOH intracellular extremity enclosing the conserved FAD- and NADPH-binding sites common to all NOX enzymes (Fig. 14.3a). DUOX proteins are glycosylated. Both human DUOX1 and DUOX2 have two N-glycosylation states: the high mannose glycosylated form found in the endoplasmic reticulum, that runs by gel electrophoresis at 180 kDa, and a fully glycosylated form processed through the Golgi and located at the plasma membrane that runs at 190 kDa [20, 22]. DUOX protein is detected by immunohistochemistry predominantly at the plasma membrane as shown in human thyroid in close vicinity of thyroperoxidase (Fig. 14.4). DUOX and TPO make part of a protein complex called the thyroxisome wherein the producer and the consumer of H2O2 are closely associated to promote hormonogenesis and limit damaging H2O2 leakage inside the cell [23–26]. Human DUOX1 and DUOX2 genes are co-localized on chromosome 15q15.3, span 75 kb in an opposite transcriptional orientations and are separated by a ~ 16 kb region (Fig. 14.5). DUOX1 is more telomeric, spans 36 kb and is composed of 35 exons; the two first being non-coding. DUOX2 spans 21.5 kb and is composed of 34 exons; the first being non-coding. The promoters of both genes do not resemble each other and differ from promoters of the thyroid-specific TG, TPO genes. The two DUOX genes are highly expressed in the thyroid and no significant TSH-dependent up-regulation of DUOX transcription is observed in humans in contrast with the TSH stimulation of the transcriptional activity of NIS, TG, TPO promoters [27]. The onset of DUOX expression in mouse thyroid embryonic development points to DUOX as a thyroid differentiation marker. DUOX proteins are expressed just after the thyroid precursor cells have completed their migration from the primitive pharynx and reached their final location around the trachea [28]. This final morphological maturation in mouse begins with the expression of TG at embryonic day 14 (E14) followed one day later (E15) with the expression of TPO, NIS, TSH receptor and DUOX concomitant with the appearance of iodinated TG, the precursor of thyroid hormones [29, 30].

232

F. Miot and X. De Deken

Fig. 14.3 Schematic structures of DUOX2 (A) and DUOXA2 (B) proteins with genetic alterations (red dots) identified in congenital hypothyroidism. Only the most frequent missense, nonsense, frameshift mutations and in-frame deletions are localized. Black, deleterious mutations (< 60% residual activity); green, functional missense variant (>60% residual activity); blue, mutation not functionally tested

3

DUOX-DUOXA Is the Active Enzyme Complex

If the extracellular accumulation of DUOX-dependent H2O2 after activation of TSH stimulated pathway in various thyroid cellular systems (thyroid slices, open follicles, thyroid cell lines. . .) was easily quantified by classical methods, like oxidation of Amplex Red or homovanillic acid assay [31, 32], H2O2 generation by recombinant DUOX proteins could not be detected. Transfected in several cell lines (CHO, Hek293, NIH3T3, Hela and PLB-XCGD), DUOX proteins were retained in the endoplasmic reticulum in their immature form and did not produce extracellular H2O2 [22]. Attempts to reconstitute DUOX activity by co-transfection with known activators or stabilizers of NOX enzymes (p67phox, p47phox, p22phox,. . . ..) or EFP-1, a thioredoxin-related protein interacting with DUOX [33], all failed. The reconstitution of a DUOX-based functional H2O2 generating system was

made finally possible thanks to the uncovering of the DUOX maturation factors: DUOXA [34]. Two human DUOXA paralogs were initially identified as enriched genes in the thyroid. These two genes are located on chromosome 15 in a tail to tail orientation in the 16 kb-DUOX1/DUOX2 intergenic region. DUOX and DUOXA genes are organized in tandem in an operon-like unit and share a common bidirectional promoter; DUOXA1 facing DUOX1 and DUOXA2 facing DUOX2 (Fig. 14.5). Remarkably, the bidirectional DUOX/DUOXA organization is conserved throughout the vertebrate lineage whereas Teleosts have a single DUOX/ DUOXA arrangement and C. elegans and Drosophila melanogaster (D.melanogaster) lack a DUOXA homolog in the vicinity of their respective duox locus [34]. Human DUOXA2 open reading frame spans 6 exons and encodes a 320 amino acid protein composed of five transmembrane segments, a first large external loop presenting N-glycosylation sites between the second and third transmembrane helices and a C-terminal cytoplamic region

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

233

Fig. 14.3 (continued)

(Fig. 14.3b). Four alternatively spliced DUOXA1 variants have been identified; the most expressed transcript, DUOXA1α, encodes a 343 amino acid protein, which is the closest homolog of DUOXA2 (58% identity of sequence) and adopts the same predicted structure. Transfection of either DUOX or DUOXA alone does not result in H2O2 generation. Remarkably, co-transfection of DUOX with DUOXA rescues DUOX activity, the release of H2O2 from the cells being blocked by diphenyliodonium (DPI). Like DUOX, DUOXA protein alone is mainly retained in the endoplasmic reticulum. When co-expressed with DUOX, the two proteins are co-transported to the plasma membrane where they form the active complexes [35]. Only the DUOX1/DUOXA1 and DUOX2/DUOXA2 pairs produce the highest levels of H2O2 as they undergo glycosylation during transition from the endoplasmic reticulum to the Golgi complex. DUOX2/DUOXA1 complex produces less H2O2 but rather superoxide anions and DUOX1/DUOXA2 is less potent to produce H2O2 [36, 37]. The type of released DUOX-dependent ROS is dictated by defined sequences in DUOXA. The C-terminal end and the second intracellular loop of DUOXA1 are required for full activity of DUOX1, while active DUOX2 depends on the integrity of the N-terminal extremity of DUOXA2 [37]. The reconstitution of a DUOX functional H2O2 producing system with homolog pairs makes it possible to measure and

compare the intrinsic enzymatic activities of DUOX1 and DUOX2 and to define how they are controlled. The basal NADPH oxidase activity of both isoenzymes is dependent on calcium via the two EF-hand Ca++ binding motifs. The H2O2 generation by DUOX/DUOXA is also activated through direct serine/threonine phosphorylation. DUOX1 activity is stimulated via protein kinase A-mediated phosphorylation on serine 955. DUOX2 is stimulated by nanomolar concentrations of phorbol 12-myristate 13-acetate (PMA) via protein kinase C dependent phosphorylation [38]. In primary cultured human thyrocytes, activation of cyclic AMP dependent pathway by forskolin and protein kinase C pathway by PMA lead to an increase of H2O2 production suggesting that both DUOX proteins are involved in thyroid hormone synthesis by providing H2O2 to TPO. The association of DUOXA as activator /maturation factor with DUOX in a functional enzyme complex is demonstrated in vivo in the double knockout Duoxa1-Duoxa2 mouse model (Duoxa-/-). Duoxa-/- mice present a maturation defect of Duox proteins retained in the endoplasmic reticulum and a loss of H2O2 release from their thyroid. They develop severe postnatal congenital hypothyroidism with a goitre and extremely low level of circulating T4 [39]. The association of DUOX/DUOXA in active complexes has been elucidated by cryo-electron microscopy (cryo-EM). Two studies simultaneously identify the co-assembling of

234 Fig. 14.4 Localization of DUOX and TPO on a section of human thyroid by immunohistochemistry. (a) immunostaining of DUOX; the protein is localized at the apical pole of all thyroid cells and at the apical plasma membrane as pointed by the arrows. (b) No labeling in the immunostaining with preimmune serum. (c) Immunostaining of thyroperoxidase is also localized at the apical pole of the cells

F. Miot and X. De Deken

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

235

Fig. 14.5 Schematic representation of the genomic organization of the human DUOX-DUOXA gene locus on chromosome 15q15.3. The exons are represented as vertical bars

DUOX and DUOXA in an active enzymatic complex localized at the plasma membrane of the cell. The cryo-EM structures obtained in low- and high-calcium states reveal the human DUOX1/DUOXA1 complex as a symmetric 2:2 heterotetramer stabilized by inter-subunit interactions. Ca++ binding to the EF-hands enhances the catalytic activity of DUOX1 (decreased Km and increased kcat of the NADPH oxidation reaction) by stabilizing the dehydrogenase domain in a conformation that allows electron transfer [40]. The cryo-EM structures of mouse Duox1/Duoxa1 obtained in the presence and absence of NADPH and analyzed by Ji Sun [41] reveal that the dimer of dimers is rather in an inactive state contrary to the active hetero-dimer, suggesting that oligomerization could be a new level of regulation of DUOX activity. The differences in the composition of the active complexes observed in these two studies are likely to be due to the different local map quality of certain domains of DUOX1 rather than a difference in species (human vs mouse) or the preparation procedure. In the two proposed models, defined interactions between DUOX and DUOXA at the membrane contribute to the stabilization of the complex and the activation of DUOX. See Chap. 30 by Jing-Xiang Wu, Ji Sun, and Lei Chen.

4

DUOX-DUOXA Defects in Congenital Hypothyroidism

Defects in DUOX/DUOXA were rapidly recognized as possible causes of hypothyroidism in two mouse models: the B6 (129)-Duox2thyd/J mouse strain, harboring a spontaneous homozygous Duox2 inactivating missense mutation (V674G) [42] and the Duoxa-/- knockout mouse with truncated Duoxa genes modelling complete deficiency of both DUOX isoenzymes [39]. The two mouse models present a severe form of hypothyroidism with low circulating T4, high serum levels of TSH, a goitre and a substantial growth retardation. Duox1 knockout mice present no obvious thyroid function defect but an altered urinary bladder function [43]. In humans, mutations in DUOX and/or DUOXA genes are pointed responsible for congenital hypothyroidism (CH) due to thyroid dyshormonogenesis in patients born with a hyperplastic thyroid or developing a postnatal goitre when T4 treatment is delayed after birth. CH affect 1 in 3000 newborns [44]. About 85% of CH result of thyroid dysgenesis where the thyroid is ectopic, missing, or severely hypoplastic. Germline

mutations in genes commonly associated with thyroid development such as PAX8, NKX2, FOXE1, HHEX were identified in a small portion of patients with CH with thyroid dysgenesis. Other genes involved in thyroid development remain to be identified. In the other 15% of CH cases, hypothyroidism is associated with disorders inherited in an autosomal recessive manner in genes involved in TH synthesis. Multiple mutations in the genes involved in differentiated thyroid function (TPO, TG, SLC5A5, SLC26A4) have been reported. Inactivating mutations in these genes lead to altered proteins involved in one of the steps of TH biosynthesis and secretion: NIS (encoded by SLC5A5) responsible for iodide trapping; Pendrin (encoded by SLC26A4) responsible for iodide discharge beyond the apical membrane; TG, when iodinated, is the main substrate of TH synthesis; DEHAL1 responsible for iodide recycling; TPO and DUOX-DUOXA responsible for the iodide organification. Homozygous mutations in TPO are associated with a total iodide organification defect (IOD) and a severe CH phenotype; while mutations in DUOX and DUOXA are associated to variable hypothyroid phenotypes ranging from transient to permanent hypothyroidism and partial to total IOD. In 2002, the characterization of the first inactivating DUOX2 mutations causing IOD undoubtedly demonstrated the essential role played by DUOX2 in TH synthesis [45]. The authors of this study reported that bi-allelic mutations of DUOX2 cause permanent CH, while monoallelic mutations lead to transient CH. However, a plethora of DUOX2 mutations found and described to date (around 110 mutations) refutes this simple pattern in about 40% of CH cases. The permanent and transient nature of CH is not directly related to the number of inactivated DUOX2 alleles [46]. This is clearly illustrated by a case we had the opportunity to analyse. CH was detected in a new born (high TSH and low total T4 levels, mildly enlarged thyroid) and was transient. Indeed, T4 treatment was given from the age of 1 month to 17 years, after which it was stopped and the patient remained euthyroid. We found complete inactivation of DUOX2 caused by a partial genomic deletion of one allele inherited from the mother associated with a paternally inherited missense mutation (c.4552G > A, p.Gly1518Ser). The deleted fragment encompassed the entire COOHterminal end of the catalytic core and Gly1518Ser DUOX2 mutated protein was weakly expressed at the cell surface of transfected cells but was non-functional [47]. Complete inactive DUOX2 was presumably rescued by DUOX1/DUOXA1

236

in this patient at adult age when TH requirements are lower than during early childhood. Reported DUOX2 variants in more than 200 unrelated CH patients include missense, nonsense, in frame deletions, splice-site and frame shift mutations. These mutations are distributed over the entire length of the protein: about one third of the mutations are found in the extracellular PHD, one third in the first intracellular loop and one third in the domain comprising the catalytic core common to all NOX proteins. We have represented a part of the reported mutations associated with CH and have indicated with different colours the consequence on DUOX activity measured by a functional test when available (Fig. 14.3a). The severity of hypothyroidism generated by these mutations is variable; from mild to severe associated with mono-allelic or bi-allelic DUOX2 inactivation. Most of the time, hypothyroidism is transient but in rare cases can persist into adulthood [48–51]. Intra-familial variabilities in phenotype have also been reported in siblings harbouring the same genetic defects. For example, four siblings of one family carrying the same compound heterozygous DUOX2 mutations, p.L479SfsX2 and p.K628RfsX10, present either permanent or transient CH [52]. The prevalence of DUOX2 mutations among CH patients is quite variable but generally high: 29 to 83% in China [53– 56], 43% in Japan [57], 56% in Thailand [58], 30 to 45% in Italy ([59, 60], and 35% in Korea [61]. In case of suspicion of inherited CH, the best criteria for a DUOX2 genetic screening are the presence of a goitre, a partial IOD, a low T4/TSH serum ratio, high serum TG levels and a transient phenotype [46, 62]. The prevalence of DUOXA2 variants in CH is much lower, around 1%. About 15 DUOXA2 mutations associated with CH have been reported to date, most of them are shown in Fig. 14.3b [49, 63–70]. The first homozygous missense p. Y246X mutation was reported in 2008 in a Chinese patient suffering from a mild permanent CH [63]. The corresponding recombinant mutant protein was inactive in vitro. Four additional cases with the same mutation presented various severities of hypothyroid phenotype [65, 67, 68, 70]. Intrafamilial variabilities were also reported for DUOXA2 affected patients. Two siblings of a Japanese family harbouring the identical homozygous loss-of-function mutation p.Y138X DUOXA2 had a totally different thyroid phenotype: the girl suffered from a permanent CH while her elder brother was not clinically affected [64]. The transient nature of neonatal CH frequently reported in cases with DUOX2 (see above) and DUOXA2 deficiency reflects a compensatory H2O2 production by DUOX1/ DUOXA1 that maintains a euthyroid status when the peak demand for TH biosynthesis declines later in childhood and adulthood [47, 59, 71]. A not compensated DUOX2 mutated-dependent iodide organification defect results in a severe permanent CH. This

F. Miot and X. De Deken

was the case in two CH siblings that carried DUOX2 homozygous nonsense non-functional variant (p.R434X) in addition to the first uncovered homozygous DUOX1 mutation (c.1823-1G.C) resulting in aberrant splicing and in an inactive truncated protein (p.V607DfsX43). The inability of DUOX1 to produce H2O2 and so to compensate DUOX2 deficiency probably explain the unusual severe CH phenotype observed in these affected patients [72]. The mild transient CH phenotype of a patient harbouring the missense DUOXA2 p.C183R and a deletion of 43kbp encompassing DUOX2-DUOXA2-DUOXA1 genomic region but conserving one DUOX2-, two DUOX1- and one DUOXA1- functional alleles supports also the existence of a rescue mechanism via DUOXA1 and indicates the high level of functional redundancy in the DUOX/DUOXA system [66]. The variety of observed phenotypes associated with DUOX2 and DUOXA2 deficiencies suggest that the degree of hypothyroidism could likely be influenced by genetic modifiers and environmental factors like nutritional iodide. High dietary iodide intake delays the appearance of hypothyroidism [73, 74] and leads to less severe CH phenotypes generally observed in the States and in Japan where the nutritional iodide (iodinated salt, seafood) intake is elevated compared to European counties [75].

5

DUOX-DUOXA and Thyroid Carcinogenesis

Several lines of evidence indicate that H2O2 produced in large quantity in the thyroid could promote carcinogenesis. Beside DUOX1 and DUOX2, thyroid expresses NOX4 localized at the endoplasmic reticulum and in vicinity of the nucleus. DNA damage and subsequent senescence observed in the early stages of thyroid tumorigenesis result from the activation of H-RasV12 that triggers NOX4 dependent H2O2 production [76]. Moreover large quantities of H2O2 produced during TH synthesis and not properly consumed by TPO or antioxidants like catalase, gluthatione peroxidase, peroxiredoxin have been proposed to cause oxidative DNA damage resulting in the high spontaneous mutation rate observed in the thyroid compared to other tissues [77, 78]. DUOX2-TPO association at the plasma membrane constitutes a physical barrier to limit diffusion of H2O2 protecting thyrocytes from oxidative damage [23, 25]. Patients with mutations in TPO suffer from severe hypothyroidism with a goitre presenting nodules that frequently develop into tumours [79]. Constitutive co-activation in a mouse model of the calcium-phosphatidylinositol-PKC pathway, that controls DUOX activity, and adenylate cyclase-cAMP-PKA pathway, that controls cell proliferation, promotes malignant

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

transformation of thyroid follicular cells [80] suggesting again that excess of DUOX-mediated H2O2 might account for thyroid tumorigenesis. It has been clearly shown that H2O2 induces DNA singleand double-strand breaks (DSB) and that thyrocytes have developed efficient antioxidant and repair transcriptional programs [81–84]. The genomic instability following DSBs contributes to chromosomal rearrangement like the oncogenic RET/PTC which is an early event in the oncogenesis of papillary thyroid cancers [85–87]. Accidental and therapeutic exposures to ionizing radiation is a high risk factor for thyroid cancer during childhood [88– 90]. The effects of radiation occur during and immediately after the exposure (early effects) but the genomic instabilities are maintained over the time (delayed effects) [91]. ROS seem to be associated to these early effects, as well as the late bystander effects on the non-irradiated cells [92]. Irradiation of human thyroid cell lines or thyroid primary cultures provokes an immediate first wave of ROS due to water radiolysis and a delayed DUOX1-dependent H2O2 production several days after irradiation. In these conditions, DUOX1 expression is increased via Interleukin-13 (IL-13) after p38 MAPK activation and induces persistent DNA damage and growth arrest suggesting a key role of DUOX1-mediated H2O2 production in long-term persistent radiation -induced DNA damage. Elevated levels of DUOX1 mRNA and protein expression have been reported in a small series of radio-induced thyroid tumours [93]. However, DUOX1 does not emerge as differentially expressed from studies comparing the expression profiles of sporadic (non-radio-induced) and radiation-induced papillary thyroid tumors from the Chernobyl Tissue Bank [94, 95]. This does not exclude the proposal that high levels of DUOX1 might be required for thyroid tumorigenesis. An animal model overexpressing inducible DUOX in the thyroid would provide a helpful tool to determine the potential implication of H2O2 at early and late stage of tumorigenesis.

6

DUOX in Association with Peroxidases Is Involved in Tyrosine Cross-Linking

DUOX, by supplying H2O2 to peroxidases, plays a critical role in tyrosine cross-linking in proteins of extracellular matrix in living species. Most of the tissues expressing DUOX co-express a dedicated peroxidase to catalyze the oxidation reactions resulting in cross-linked bonds in proteins. We have already seen that, in thyroid, oxidized iodide is covalently bound to TG tyrosines. The iodinated tyrosyl residues, mono- or di-iodinated tyrosines, are linked via a phenoxy-ether bond to form T4 or T3 (Fig. 14.2). These reactions are catalyzed by TPO, the thyroid specific

237

peroxidase, in presence of H2O2 provided by DUOX. TPO and DUOX are both localized at the apical membrane of the thyrocyte so that the consumer and the supplier of H2O2 are in close vicinity. Hereafter are some examples of cooperativity of DUOX with peroxidases to form tyrosine cross-links in proteins essential for extracellular matrix stabilization and cell viability of lower organisms. The nematode cuticle is a collagenous extracellular matrix composed of cross-linked proteins constituting a protective physical barrier toward the external environment. Impairment of biosynthesis of such cuticle in C.elegans gives rise to a blistered phenotype. C. elegans expresses in the hypodermis one functional DUOX called Ce-DUOX1/BLI-3 with a peroxidase active PHD. Two point mutations in the extracellular peroxidase domain that impair heme binding severely hinder the formation of the cuticle without affecting ROS production [96]. An additional peroxidase Mlt-7, co-expressed with Duox/Bli-3 in the hypodermis, is implicated in the formation of the cuticle. Disruption of mlt-7 gene cause dramatic changes in tyrosine cross-linking patterns of cuticle collagen proteins [97]. Functional cooperative relationship between Ce-DUOX/BLI-3 and MLT-7 peroxidase enables the stabilization by tyrosine cross-links in collagen of a proper extracellular matrix and ensures post-embryonic viability of the worm. See Chap. 27 by Danielle A. Garsin. Sea urchin eggs demonstrate a respiratory burst at fertilization. The high consumption of O2 corresponds to the activation of Udx1, the Duox sea urchin homolog. Udx1 produces large quantities of H2O2 immediately used by the secreted ovoperoxidase resulting in the formation of covalent di-tyrosine bonds between the proteins of the fertilization envelope. The hardening of the envelope constitutes a physical block to polyspermy and protects the embryo from further harmful effect of H2O2 [98]. In insects, eggshell hardening through chorion protein cross-linking occurs after the oxidation of tyrosine residues and subsequent formation of di-tyrosine bonds. In Rhodnius prolixus, the vector of Chagas disease, the ovary Duox is the source of the H2O2 that supports di-tyrosine-mediated chorion protein cross-linking. RNAi silencing of Duox decreases H2O2 production and provokes a desiccation of the embryo showing that eggshell hardening is essential for waterproofing and embryo development [99]. Hematophagous insects like Anopheles gambiae, vector of malaria, secrete a peritrophic matrix (PM) in response to a blood meal in the midgut where commensal bacteria proliferate extensively after blood feeding. The PM is a semipermeable layer of chitin polymers that surrounds the blood meal and prevents blood cells and gut bacteria from coming in direct contact with midgut epithelium. Mucins are secreted between the PM and the midgut epithelial cells and formed layers of cross-linked proteins. Tyrosine cross-linking results

238

F. Miot and X. De Deken

from oxidation reactions catalyzed by a peculiar secreted peroxidase IMPer, for Immunomodulary Peroxidase, in the presence of H2O2 provided by DUOX of the midgut epithelial cells. The network of di-tyrosines in the gut of Anopheles gambiae protects the microbiota by preventing activation of epithelial immunity. It provides also a suitable environment for the maintenance and proliferation of Plasmodium parasite as it prevents activation of antimalarial response mediated by nitric oxide synthase [100]. The phytophagous insect Riptortus pedestris is a notorious pest of soybean plants. This bug possesses in its midgut a symbiotic organ consisting of crypts in the lumen of which mutualistic aerobic symbiotic bacteria Burkholderia insecticola proliferate. DUOX of the gut epithelial cells in concert with an unidentified peroxidase enables nutritional gut symbiosis by mediating the formation and stabilization of a tracheal network enveloping the symbiotic organ via di-tyrosine bonds in tracheal matrix proteins. The formed respiratory network can ensure oxygen supply to the symbiotic organ and to the aerobic gut symbiont [101]. See Chap. 28 by Ana Caroline.P. Gandara and Pedro L. Oliveira.

7

DUOX and Host Defense

In addition to their high thyroid expression, DUOX are expressed on epithelial surfaces of the airways, salivary gland ducts and along the gastrointestinal digestive tract [102]. DUOX of salivary glands, rectum, trachea and bronchi in concert with the lactoperoxidase (LPO) present in glandular secretions constitutes an efficient host defense system wherein LPO oxidizes thiocyanate (SCN-) into the antimicrobial compound, hypothiocyanite (OSCN-), in the presence of H2O2 produced by DUOX [103–107].

7.1

DUOX Immune Function in the Respiratory Epithelium

The lung is particularly exposed to an aggressive environment and the airway epithelium continuously encounters a large array of microbes. Both DUOX1 and DUOX2 are expressed in the respiratory epithelium. DUOX properties have been mostly studied in primary culture of epithelial airway cells and cell lines and some confirmed in the knockout mouse models Duoxa-/and Duox1-/-. The expression of DUOX isoforms in the airways is differently regulated according the differentiation state of the cell. In tracheobronchial epithelial cells, all-trans retinoic acid induces DUOX2 expression [108] while differentiation of fetal epithelium in mature alveolar type II cells induces the expression of DUOX1 [109]. A study aimed at the

localization of DUOX1-DUOXA1 and DUOX2-DUOXA2 complexes in undifferentiated and differentiated polarized lung epithelial cells shows DUOX1 expression at the surface of non-ciliated cells while DUOX2 is expressed at the surface of ciliated cells [110]. In the airways, DUOX1 was first considered as the main isoform responsible for extracellular H2O2 production in response to bacterial infection. DUOX1-mediated H2O2 is the co-substrate of LPO within the airway secretions to form bactericidal OSCN- [104]. However, DUOX2- LPOSCN- constitutes as well an efficient antimicrobial mechanism in the upper and lower airways. This defense mechanism is defective in the airways of patients with cystic fibrosis partly due to a reduced SCN- transport by the deficient cystic fibrosis transmembrane conductance regulator (CFTR) [105, 111] and an inhibition of DUOX activity by pyocyanin, a virulence factor secreted by Pseudomonas aeruginosa, the major pathogen invading the airways of these patients [112]. DUOX1 and DUOX2 gene expression were shown to be differentially regulated in response to immunomodulatory Th1 and Th2 cytokines. In primary respiratory tract epithelial cultures treated with multiple cytokines, DUOX2 expression is increased by Interferon-ɣ (IFNɣ,), a Th1 cytokine, while DUOX1 expression is stimulated by IL-4 and IL-13, Th2 cytokines, highlighting for the first time a mechanism by which ROS production can be regulated in the respiratory tract as part of the host defense response [106, 113, 114]. Invading pathogens are detected by immune cells through specific pattern recognition receptors such as Toll-like receptor (TLR) and nucleotide-binding oligomerization domainlike receptors, NOD-like receptor [115]. These immune receptors recognize pathogen-associated molecular patterns (PAMPs) such as viral dsRNA (TLR3), lipopolysaccharide (TLR4), bacterial flagellin (TLR5), and bacterial muramyldipeptide (NOD2). A direct link exits between activation of TLR signaling and NOX-mediated ROS production [116–118]. In normal human nasal primary epithelial (NHNE) cells, bacterial flagellin promotes Ca++ dependent DUOX2-mediated H2O2 release and triggers innate immune response via TLR5 activation including IL-8 production and mucin, MUC5AC, expression [119]. DUOX1 has also been shown to be involved in TLR-mediated immune response. Human bronchial epithelial (HBE) cells treated with lipopolysaccharide release ATP that binds purinergic receptors causing the increase of intracellular calcium and the activation of DUOX1. H2O2 produced by DUOX1 can then promote the activation of the epidermal growth factor receptor (EGFR) and the subsequent signaling pathway leading to the production of IL-8 [120]. Likewise, DUOX1 critically contributes to sustained EGFR activation in allergic asthma through direct oxidation of cysteines in the EGFR and an indirect mechanism that leads to increased amphiregulin (ligand of EGFR)

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

production, IL-33 production, neutrophilic inflammation, mucous cell metaplasia and subepithelial fibrosis. All these features are attenuated in Duox1-/- mice or after intratracheal administration of DUOX1–targeted siRNA in mice wherein an allergic inflammation was induced by house dust mite instillation [121, 122]. Related findings were also reported in Duoxa-/- mice, showing again a lack of neutrophilia, type2 cytokine response and mucous cell metaplasia in an ovalbumin allergic asthma model without being able in this case to attribute the effect to one or the other DUOX protein [123]. DUOX2 plays also a role in the mucosal host defense during viral infection of the respiratory tract. The potential antiviral effect of DUOX2 was initially observed with the induction of DUOX2 transcripts in human tracheobronchial epithelial cells stimulated by poly (I:C), a mimic of viral dsRNA, or infected by Rhinovirus 1B [113]. Later, DUOX2-DUOXA2 expression was shown to be up-regulated in airway epithelial cells, at transcriptional and proteins levels, in response to a broad number of respiratory viruses including Sendai virus (SeV), respiratory syncitial virus (RSV), Influenza A virus (IAV), rhinovirus (RV) [124–127]. DUOX2 expression in response to virus infection depends on virus replication and on the autocrine/paracrine action of virus-induced cytokines. During virus replication, viral RNA moieties, the viral PAMPS, are recognized by TLRs and cytosolic retinoic acid inducible gene I (RIG-I)-like receptors (RLRs) and initiate the induction of antiviral and pro-inflammatory cytokines and chemokines gene expression [128]. Among the cytokines secreted after airway epithelium infection by SeV, IFNβ has been shown acting in synergy with the Tumor Necrosis Factor-α (TNFα) to induce the expression of DUOX2-DUOXA2 through the activation of Signal Transducer and Activator of Transcription-2/Interferon-Regulatory Factor-9 (STAT2/IRF-9) pathway [124, 129]. Beside the cytokine response, DUOX2 contributes also to the antiviral response by acting as support of LPO activity in forming OSCN- by the above described bactericidal system DUOX/LPO/SCN- [130, 131].

7.2

DUOX Immune Function in the Gastrointestinal Tract

The gut epithelial cell and mucus layers play a pivotal role as the first physical barrier against external factors and maintains a symbiotic relationship with commensal bacteria. It physically separates the gut associated lymphoid tissue from the highly antigen loaded luminal environment by minimizing the contact between the host and the commensal

239

microbiota and protects against potential pathogens by developing an adequate innate immune defense response. The latter consists of mucus secretion by goblet cells, production of antimicrobial peptides, immune effectors and generation of ROS mostly by NOX1 and DUOX2 [132, 133]. DUOX2 is expressed in the epithelium throughout the gastrointestinal tract of vertebrates, including human, with a maximal expression in the ileum [102, 134, 135]. Expression of DUOX2 homologs in the gut epithelium and their roles in innate immune defense are evolutionarily conserved. Silencing duox in D.melanogaster gut causes a severe defect of defense against infection by various pathogens and increases the mortality rate. Reintroduction of Drosophila or human DUOX rescues ROS generation, limits bacterial proliferation in the gut, and decreases the mortality of these flies [136, 137]. The reduction of expression by RNAi or mutations in Ce-Duox1/Bli-3 in C. elegans intestine increases the susceptibility to Enterococcus faecalis infection confirming that Ce-DUOX1/BLI-3-mediated H2O2 released from mucosal surfaces is an ancient conserved innate immune defense mechanism [138]. Likewise, knockdown of duox in zebrafish impairs the capacity of larvae to control the growth of enteric Salmonella highlighting that the bactericidal function of DUOX is conserved in lower vertebrates [139]. See Chap. 29 by S.M. Sabbir Alam and Daniel M. Suter. DUOX enzymes have also be shown to mediate H2O2 release from mouse gastric epithelium to prevent Helicobacter felis infection and inflammation. Indeed, infection with Helicobacter felis induces the expression of DUOX2/DUOXA2 in the epithelium of the stomach of wild type mice with a limited colonization of the mucus layer by the bacteria and signs of inflammation. The lack of functional DUOX in Duoxa-/- mice increases the mucosal colonization by Helicobacter and exacerbates acute and chronic gastritis with higher expression of pro-inflammatory markers than in control animals [140]. In the same way, DUOX2 is up-regulated in the mouse intestine infected by segmented filamentous bacteria, a group of host-adapted commensal bacteria. DUOX2-dependent H2O2 restricts the bacterial access to the lymphoid tissues. This repellent effect already clearly shown in vitro [141, 142] contributes to a reduction of the acute microbiota-induced immune response and therefore to gut immune homeostasis. The up-regulation of DUOX2 in response to mucosal dysbiosis is generally found in inflammatory bowel diseases (IBD), the most common of which are ulcerative colitis and Crohn’s disease (CD). IBD pathogenesis is characterized by an altered commensal microbial composition accompanied of epithelial dysfunction that results in increased bacterial translocation and finally an immune inflammatory stimulation. A transcriptome and microbiome study of intestinal biopsies from IBD patients shows that the expansion of

240

F. Miot and X. De Deken

Table 14.1 Tissue distribution and main functions of DUOX-DUOXA ROS species O2.-

Tissue expression Thyroid

Associated peroxidase TPO

Function DUOX/DUOXA Thyroid hormone synthesis

DUOX1 + DUOXA1 mature

H2O2

Airways, digestive tract

LPO

Host defense, signaling

DUOX2 immature DUOX2 + DUOXA2 mature

O2.-

Thyroid

TPO

Thyroid hormone synthesis

H2O2

Airways, digestive tract, salivary gland

LPO

Host defense-immune homeostasis, signaling

NOX DUOX1 immature

Involvement in human disease Congenital hypothyroidism Lung bacterial and viral infections inflammation, asthma Congenital hypothyroidism Lung Viral infection. Gastrointestinal inflammation, IBD

DUOX: DUal OXidase; DUOXA: Dual oxidase maturation factor Immature: refers to the proteins in absence of their maturation factor; mature: refers to the proteins associated with their maturation factor H2O2: hydrogen peroxide; O2.-: superoxide anion TPO: thyroperoxidase; LPO:lactoperoxidase IBD: Inflammatory Bowel Disease

mucosa-associated Proteobacteria is accompanied by an overexpression of DUOX2/DUOXA2 genes that rank in the top 5 most up-regulated genes [143]. The critical role of DUOX2 in IBD pathogenesis has been demonstrated by the identification of rare inactivating DUOX2 variants in young patients with Very Early Onset Inflammatory Bowel Disease (VEOIBD) [144, 145] and in adult patients with CD [146]. Loss-of- function mutations in DUOX2 are the most frequent genetic variants in CH presenting a low penetrance with inter-and intra-familial variability; see above. The IBD patients harboring the few identified DUOX2 variants have a normal thyroid function without signs of CH probably thanks to the compensation of H2O2 produced by DUOX1 in the thyroid and not present in intestinal epithelial cells. It was therefore difficult at this stage to argue that the risk of developing IBD would be related to CH in carriers of DUOX2 mutations. However, a retrospective analysis of a cohort of 42,922 patients with CH compared with a matched in size, age, gender non-CH group indicates that the CH patients have a 73% higher overall IBD prevalence with increased odds in patients with transient hypothyroidism for which partial defects in DUOX2 system is a common feature [147]. A multiomic phenome-wide association study (PheWAS) with phenotypic data including clinical-, microbiome- and plasma metabolite-data of 2872 individuals identifies a significant association between rare loss-of-function DUOX2 variants with increased plasma level of interleukin-17C (IL-17C), which is also present in mucosal biopsies of IBD patients. Duoxa-/- mice reproduce increased IL-17C induction in the ileum of mice presenting a dysbiosis. Thus, the association of deleterious DUOX2 variants with an elevation of plasma IL-17C, a preclinical hallmark of disturbed microbiota-immune homeostasis, may increase the susceptibility to develop IBD, suggesting that these DUOX2 variants could be a genetic risk factor in the pathogenesis of IBD [148].

8

Conclusion (Table 14.1)

DUOX has been studied for 20 years. The main aspect, but not all, have been discussed in this chapter. The story began with the cloning of two human isoforms DUOX1 and DUOX2 in the thyroid followed by the identification of the two maturation factors DUOXA1 and DUOXA2 which made it possible to measure the NADPH oxidase activity of DUOX. DUOX interacts with DUOXA at the plasma membrane of epithelial cells. Mature DUOX enzymes produce H2O2. The first function attributed to DUOX1/2 is the supply of H2O2 to thyroperoxidase in the process of thyroid hormone synthesis. Mutations in DUOX2 are often associated with congenital hypothyroidism. DUOX1 and 2 are also expressed in airway and digestive tract epithelial cells where their role in host defense, immune response and signaling has become indisputable. Acknowledgements The authors acknowledge the support for their own research over the years of the “Fonds de la Recherche Scientifique (FRS-FNRS)”, the “Fonds Yvonne Smits” and “Fonds Dr JP Naets” managed by the “Fondation Roi Baudouin”.

References 1. Rousset B, Dupuy C, Miot F, Dumont JE (2015) Thyroid hormone synthesis and secretion. In: Feingold K, Anawalt B, Boyce A (eds) Thyroid disease manager, updated 20. MDText.com, Inc., South Dartmouth, pp 1–66 2. Dumont JE (1971) The action of thyrotropin on thyroid metabolism. Vitam Horm 29:287–412. https://doi.org/10.1016/S00836729(08)60051-5 3. Massart C, Hoste C, Virion A et al (2011) Cell biology of H2O2 generation in the thyroid: investigation of the control of dual oxidases (DUOX) activity in intact ex vivo thyroid tissue and cell lines. Mol Cell Endocrinol 343:32–44. https://doi.org/10.1016/j. mce.2011.05.047

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

4. Björkman ER (1992) Hydrogen peroxide generation and its regulation in frtl-5 and porcine thyroid cells. Endocrinology 130:393– 399. https://doi.org/10.1210/endo.130.1.1309340 5. Björkman U, Ekholm R (1984) Generation of H2O2 in isolated porcine thyroid follicles. Endocrinology 115:392–398. https://doi. org/10.1210/endo-115-1-392 6. Virion A, Michot JL, Dème D et al (1984) NADPH-dependent H2O2 generation and peroxidase activity in thyroid particular fraction. Mol Cell Endocrinol 36:95–105. https://doi.org/10.1016/ 0303-7207(84)90088-1 7. Dème D, Virion A, Hammou NA, Pommier J (1985) NADPHdependent generation of H 2 O 2 in a thyroid particulate fraction requires Ca 2+. FEBS Lett 186:107–110. https://doi.org/10.1016/ 0014-5793(85)81349-1 8. Nakamura Y, Ogihara S, Ohtaki S (1987) Activation by ATP of calcium-dependent NADPH-oxidase generating hydrogen peroxide in thyroid plasma membranes. J Biochem 102:1121–1132. https://doi.org/10.1093/oxfordjournals.jbchem.a122150 9. Dupuy C, Kaniewski J, Dème D et al (1989) NADPH-dependent H2O2 generation catalyzed by thyroid plasma membranes. Studies with electron scavengers. Eur J Biochem 185:597–603. https://doi. org/10.1111/j.1432-1033.1989.tb15155.x 10. Carvalho DP, Dupuy C, Gorin Y et al (1996) The Ca2+- and reduced nicotinamide adenine dinucleotide phosphate-dependent hydrogen peroxide generating system is induced by thyrotropin in porcine thyroid cells. Endocrinology 137:1007–1012. https://doi. org/10.1210/endo.137.3.8603567 11. Corvilain B, Van Sande J, Dumont JE (1988) Inhibition by iodide of iodide binding to proteins: the “Wolff-Chaikoff” effect is caused by inhibition of H2O2 generation. Biochem Biophys Res Commun 154:1287–1292. https://doi.org/10.1016/0006-291X(88)90279-3 12. Corvilain B, Van Sande J, Laurent E, Dumont JE (1991) The H2O2-generating system modulates protein iodination and the activity of the pentose phosphate pathway in dog thyroid. Endocrinology 128:779–785. https://doi.org/10.1210/endo-128-2-779 13. Corvilain B, Laurent E, Lecomte M et al (1994) Role of the cyclic adenosine 3′,5′-monophosphate and the phosphatidylinositolCa2+ cascades in mediating the effects of thyrotropin and iodide on hormone synthesis and secretion in human thyroid slices. J Clin Endocrinol Metab 79:152–159 14. Coclet J, Foureau F, Ketelbant P et al (1989) Cell population kinetics in dog and human adult thyroid. Clin Endocrinol 31: 655–665. https://doi.org/10.1111/j.1365-2265.1989.tb01290.x 15. Dupuy C, Ohayon R, Valent A et al (1999) Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. J Biol Chem 274:37265–37269. https://doi.org/10.1074/jbc.274.52.37265 16. De Deken X, Wang D, Many M-C et al (2000) Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275:23227–23233. https://doi.org/ 10.1074/jbc.M000916200 17. Meitzler JL, Ortiz de Montellano PR (2011) Structural stability and heme binding potential of the truncated human dual oxidase 2 (DUOX2) peroxidase domain. Arch Biochem Biophys 512: 197–203. https://doi.org/10.1016/j.abb.2011.05.021 18. Meitzler JL, Ortiz de Montellano PR (2009) Caenorhabditis elegans and human dual oxidase 1 (DUOX1) “peroxidase” domains. J Biol Chem 284:18634–18643. https://doi.org/10.1074/ jbc.M109.013581 19. Meitzler JL, Hinde S, Banfi B et al (2013) Conserved cysteine residues provide a protein-protein interaction surface in dual oxidase (DUOX) proteins. J Biol Chem 288:7147–7157. https://doi. org/10.1074/jbc.M112.414797 20. Morand S, Agnandji D, Noel-Hudson M-S et al (2004) Targeting of the dual oxidase 2 N-terminal region to the plasma membrane. J Biol Chem 279:30244–30251. https://doi.org/10.1074/jbc. M405406200

241 21. Carré A, Louzada RAN, Fortunato RS et al (2015) When an intramolecular Disulfide bridge governs the interaction of DUOX2 with its partner DUOXA2. Antioxid Redox Signal 23: 724–733. https://doi.org/10.1089/ars.2015.6265 22. De Deken X, Wang D, Dumont JE, Miot F (2002) Characterization of ThOX proteins as components of the thyroid H2O2-generating system. Exp Cell Res 273:187–196. https://doi.org/10.1006/excr. 2001.5444 23. Fortunato RS, Lima de Souza EC, Hassani RA et al (2010) Functional consequences of dual oxidase-Thyroperoxidase interaction at the plasma membrane. J Clin Endocrinol Metabol 95:5403– 5411. https://doi.org/10.1210/jc.2010-1085 24. Song Y, Driessens N, Costa M et al (2007) Roles of hydrogen peroxide in thyroid physiology and disease. J Clin Endocrinol Metabol 92:3764–3773. https://doi.org/10.1210/jc.2007-0660 25. Song Y, Ruf J, Lothaire P et al (2010) Association of Duoxes with thyroid peroxidase and its regulation in thyrocytes. J Clin Endocrinol Metabol 95:375–382. https://doi.org/10.1210/jc. 2009-1727 26. De Deken X, Corvilain B, Dumont JE, Miot F (2014) Roles of DUOX-mediated hydrogen peroxide in metabolism, host defense, and signaling. Antioxid Redox Signal 20:2776–2793. https://doi. org/10.1089/ars.2013.5602 27. Pachucki J, Wang D, Christophe D, Miot F (2004) Structural and functional characterization of the two human ThOX/Duox genes and their 5′-flanking regions. Mol Cell Endocrinol 214:53–62. https://doi.org/10.1016/j.mce.2003.11.026 28. De Felice M, Di Lauro R (2004) Thyroid development and its disorders: genetics and molecular mechanisms. Endocr Rev 25: 722–746. https://doi.org/10.1210/er.2003-0028 29. Postiglione MP, Parlato R, Rodriguez-Mallon A et al (2002) Role of the thyroid-stimulating hormone receptor signaling in development and differentiation of the thyroid gland. Proc Natl Acad Sci U S A 99:15462–15467. https://doi.org/10.1073/PNAS.242328999 30. Milenkovic M, De Deken X, Jin L et al (2007) Duox expression and related H2O2 measurement in mouse thyroid: onset in embryonic development and regulation by TSH in adult. J Endocrinol 192:615–626. https://doi.org/10.1677/JOE-06-0003 31. Ameziane-El-Hassani R, Morand S, Boucher J-L et al (2005) Dual Oxidase-2 has an intrinsic Ca2+-dependent H2O2-generating activity. J Biol Chem 280:30046–30054. https://doi.org/10.1074/ jbc.M500516200 32. Bénard B, Brault J (1971) Production of peroxide in the thyroid. L’union medicale du Canada 100:701–705 33. Wang D, De Deken X, Milenkovic M et al (2005) Identification of a novel partner of Duox. J Biol Chem 280:3096–3103. https://doi. org/10.1074/jbc.M407709200 34. Grasberger H, Refetoff S (2006) Identification of the maturation factor for dual oxidase: evolution of an eukaryotic operon equivalent. J Biol Chem 281:18269–18272. https://doi.org/10.1074/jbc. C600095200 35. Grasberger H, De Deken X, Miot F et al (2007) Missense mutations of dual oxidase 2 (DUOX2) implicated in congenital hypothyroidism have impaired trafficking in cells reconstituted with DUOX2 maturation factor. Mol Endocrinol 21:1408–1421. https://doi.org/ 10.1210/me.2007-0018 36. Morand S, Ueyama T, Tsujibe S et al (2009) Duox maturation factors form cell surface complexes with Duox affecting the specificity of reactive oxygen species generation. FASEB J 23:1205– 1218. https://doi.org/10.1096/fj.08-120006 37. Hoste C, Dumont JE, Miot F, De Deken X (2012) The type of DUOX-dependent ROS production is dictated by defined sequences in DUOXA. Exp Cell Res 318:2353–2364. https://doi. org/10.1016/j.yexcr.2012.07.007

242 38. Rigutto S, Hoste C, Grasberger H et al (2009) Activation of dual oxidases Duox 1 and Duox2. J Biol Chem 284:6725–6734. https:// doi.org/10.1074/jbc.M806893200 39. Grasberger H, De Deken X, Mayo OB et al (2012) Mice deficient in dual oxidase maturation factors are severely hypothyroid. Mol Endocrinol 26:481–492. https://doi.org/10.1210/me.2011-1320 40. Wu J-X, Liu R, Song K, Chen L (2021) Structures of human dual oxidase 1 complex in low-calcium and high-calcium states. Nat Commun 12:155–165. https://doi.org/10.1038/s41467-02020466-9 41. Sun J (2020) Structures of mouse DUOX1–DUOXA1 provide mechanistic insights into enzyme activation and regulation. Nat Struct Mol Biol 27:1086–1093. https://doi.org/10.1038/s41594020-0501-x 42. Johnson KR, Marden CC, Ward-Bailey P et al (2007) Congenital hypothyroidism, dwarfism, and hearing impairment caused by a missense mutation in the mouse dual oxidase 2 gene, Duox2. Mol Endocrinol 21:1593–1602. https://doi.org/10.1210/me.2007-0085 43. Donkó Á, Ruisanchez É, Orient A et al (2010) Urothelial cells produce hydrogen peroxide through the activation of Duox 1. Free Radic Biol Med 49:2040–2048. https://doi.org/10.1016/j. freeradbiomed.2010.09.027 44. Deladoëy J, von Oettingen J, van Vliet G (2005) Hypothyroidism in infants and children. In: Braverman L, Coope D, Kopp PA (eds) Werner & Ingbar’s the thyroid a fundamental and clinical text, 11th edn. Lippincott Williams & Wilkins, Philadelphia 45. Moreno JC, Bikker H, Kempers MJE et al (2002) Inactivating mutations in the gene for thyroid oxidase 2 (THOX2) and congenital hypothyroidism. N Engl J Med 347:95–102. https://doi.org/10. 1056/NEJMoa012752 46. Muzza M, Rabbiosi S, Vigone MC et al (2014) The clinical and molecular characterization of patients with dyshormonogenic congenital hypothyroidism reveals specific diagnostic clues for DUOX2 defects. J Clin Endocrinol Metabol 99:E544–E553. https://doi.org/10.1210/jc.2013-3618 47. Hoste C, Rigutto S, van Vliet G et al (2010) Compound heterozygosity for a novel hemizygous missense mutation and a partial deletion affecting the catalytic core of the H2O2-generating enzyme DUOX2 associated with transient congenital hypothyroidism. Hum Mutat 31:E1304–E1319. https://doi.org/10.1002/humu. 21227 48. Muzza M, Fugazzola L (2017) Disorders of H2O2 generation. Best Pract Res Clin Endocrinol Metab 31:225–240. https://doi.org/10. 11844/cjcb.2013.10.0124 49. Peters C, Nicholas AK, Schoenmakers E et al (2019) DUOX2/ DUOXA2 mutations frequently cause congenital hypothyroidism that evades detection on Newborn screening in the United Kingdom. Thyroid 29:790–801. https://doi.org/10.1089/thy.2018.0587 50. Zheng Z, Yang L, Sun C et al (2020) Genotype and phenotype correlation in a cohort of Chinese congenital hypothyroidism patients with DUOX2 mutations. Ann Transl Med 8:1649–1649. https://doi.org/10.21037/atm-20-7165 51. De Deken X, Miot F (2019) DUOX defects and their roles in congenital hypothyroidism. In: Knaus UG, Leto TL (eds) Methods in molecular biology, 2019th ed Humana, New York, pp. 667–693 52. Maruo Y, Takahashi H, Soeda I et al (2008) Transient congenital hypothyroidism caused by biallelic mutations of the dual oxidase 2 gene in Japanese patients detected by a neonatal screening program. J Clin Endocrinol Metabol 93:4261–4267. https://doi. org/10.1210/jc.2008-0856 53. Fu C, Zhang S, Su J et al (2015) Mutation screening of DUOX2 in Chinese patients with congenital hypothyroidism. J Endocrinol Investig 38:1219–1224. https://doi.org/10.1007/s40618-0150382-8 54. Fu C, Luo S, Zhang S et al (2016) Next-generation sequencing analysis of DUOX2 in 192 Chinese subclinical congenital

F. Miot and X. De Deken hypothyroidism (SCH) and CH patients. Clin Chim Acta 458:30– 34. https://doi.org/10.1016/j.cca.2016.04.019 55. Tan M, Huang Y, Jiang X et al (2016) The prevalence, clinical, and molecular characteristics of congenital hypothyroidism caused by DUOX2 mutations: a population-based cohort study in Guangzhou. Horm Metab Res 48:581–588. https://doi.org/10. 1055/s-0042-112224 56. Jiang H, Wu J, Ke S et al (2016) High prevalence of DUOX2 gene mutations among children with congenital hypothyroidism in central China. Eur J Med Genet 59:526–531. https://doi.org/10.1016/j. ejmg.2016.07.004 57. Narumi S, Muroya K, Asakura Y et al (2011) Molecular basis of thyroid Dyshormonogenesis: genetic screening in populationbased Japanese patients. J Clin Endocrinol Metabol 96:E1838– E1842. https://doi.org/10.1210/jc.2011-1573 58. Sorapipatcharoen K, Tim-Aroon T, Mahachoklertwattana P et al (2020) DUOX2 variants are a frequent cause of congenital primary hypothyroidism in Thai patients. Endocr Connect 9:1121–1134. https://doi.org/10.1530/EC-20-0411 59. Rabbiosi S, Vigone MC, Cortinovis F et al (2013) Congenital hypothyroidism with Eutopic thyroid gland: analysis of clinical and biochemical features at diagnosis and after re-evaluation. J Clin Endocrinol Metabol 98:1395–1402. https://doi.org/10.1210/ jc.2012-3174 60. De Marco G, Agretti P, Montanelli L et al (2011) Identification and functional analysis of novel dual oxidase 2 (DUOX2) mutations in children with congenital or subclinical hypothyroidism. J Clin Endocrinol Metabol 96:E1335–E1339. https://doi.org/10.1210/jc. 2010-2467 61. Park KJ, Park HK, Kim YJ et al (2016) DUOX2 mutations are frequently associated with congenital hypothyroidism in the Korean population. Ann Lab Med 36:145–153. https://doi.org/10. 3343/alm.2016.36.2.145 62. Dufort G, Larrivée-Vanier S, Eugène D et al (2019) Wide Spectrum of DUOX2 deficiency: from life-threatening compressive Goiter in infancy to lifelong Euthyroidism. Thyroid 29:1018– 1022. https://doi.org/10.1089/thy.2018.0461 63. Zamproni I, Grasberger H, Cortinovis F et al (2008) Biallelic inactivation of the dual oxidase maturation factor 2 (DUOXA2) gene as a novel cause of congenital hypothyroidism. J Clin Endocrinol Metabol 93:605–610. https://doi.org/10.1210/jc. 2007-2020 64. Sugisawa C, Higuchi S, Takagi M et al (2017) Homozygous DUOXA2 mutation (p.Tyr138*) in a girl with congenital hypothyroidism and her apparently unaffected brother: case report and review of the literature. Endocr J 64:1–6. https://doi.org/10.1507/ endocrj.EJ16-0564 65. Zheng X, Ma S, Guo M et al (2017) Compound heterozygous mutations in the DUOX2/DUOXA2 genes cause congenital hypothyroidism. Yonsei Med J 58:888–890. https://doi.org/10.3349/ ymj.2017.58.4.888 66. Hulur I, Hermanns P, Nestoris C et al (2011) A single copy of the recently identified dual oxidase maturation factor (DUOXA) 1 gene produces only mild transient hypothyroidism in a patient with a novel biallelic DUOXA2 mutation and monoallelic DUOXA1 deletion. J Clin Endocrinol Metabol 96:E841–E845. https://doi.org/10.1210/jc.2010-2321 67. Yi R, Zhu W, Yang L et al (2013) A novel dual oxidase maturation factor 2 gene mutation for congenital hypothyroidism. Int J Mol Med 31:467–470. https://doi.org/10.3892/ijmm.2012.1223 68. Liu S, Liu L, Niu X et al (2015) A novel missense mutation (I26M) in DUOXA2 causing congenital Goiter hypothyroidism impairs NADPH oxidase activity but not protein expression. J Clin Endocrinol Metabol 100:1225–1229. https://doi.org/10.1210/jc. 2014-3964

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

69. Zheng X, Ma S, Qiu Y et al (2016) A novel c.554+5C>T mutation in the DUOXA2 gene combined with p.R885Q mutation in the DUOX2 gene causing congenital hypothyroidism. J Clin Res Pediatr Endocrinol 8:224–227. https://doi.org/10.1515/JPEM. 2011.408 70. Yang L-X, Ma S-G, Qiu Y-L, Zheng X (2016) Heterozygous mutations of the DUOXA2 and DUOX2 genes in dizygotic twins with congenital hypothyroidism. Clin Lab 62:849–854. https://doi. org/10.7754/Clin.Lab.2015.150840 71. Maruo Y, Nagasaki K, Matsui K et al (2016) Natural course of congenital hypothyroidism by dual oxidase 2 mutations from the neonatal period through puberty. Eur J Endocrinol 174:453–463. https://doi.org/10.1530/EJE-15-0959 72. Aycan Z, Cangul H, Muzza M et al (2017) Digenic DUOX1 and DUOX2 mutations in cases with congenital hypothyroidism. J Clin Endocrinol Metabol 102:3085–3090. https://doi.org/10.1210/jc. 2017-00529 73. Vigone MC, Fugazzola L, Zamproni I et al (2005) Persistent mild hypothyroidism associated with novel sequence variants of the DUOX2 gene in two siblings. Hum Mutat 26:395–403. https:// doi.org/10.1002/humu.9372 74. Kasahara T, Narumi S, Okasora K et al (2013) Delayed onset congenital hypothyroidism in a patient with DUOX2 mutations and maternal iodine excess. Am J Med Genet A 161A:214–217. https://doi.org/10.1002/ajmg.a.35693 75. Zimmermann MB, Jooste PL, Pandav CS (2008) Iodine-deficiency disorders. Lancet (London, England) 372:1251–1262. https://doi. org/10.1016/S0140-6736(08)61005-3 76. Weyemi U, Lagente-Chevallier O, Boufraqech M et al (2012) ROS-generating NADPH oxidase NOX4 is a critical mediator in oncogenic H-Ras-induced DNA damage and subsequent senescence. Oncogene 31:1117–1129. https://doi.org/10.1038/onc. 2011.327 77. Krohn K, Maier J, Paschke R (2007) Mechanisms of disease: hydrogen peroxide, DNA damage and mutagenesis in the development of thyroid tumors. Nat Clin Pract Endocrinol Metab 3:713– 720. https://doi.org/10.1038/ncpendmet0621 78. Maier J, van Steeg H, van Oostrom C et al (2006) Deoxyribonucleic acid damage and spontaneous mutagenesis in the thyroid gland of rats and mice. Endocrinology 147:3391–3397. https://doi.org/10.1210/en.2005-1669 79. Knobel M, Medeiros-Neto G (2003) An outline of inherited disorders of the thyroid hormone generating system. Thyroid 13: 771–801. https://doi.org/10.1089/105072503768499671 80. Ledent C, Denef JF, Cottecchia S et al (1997) Costimulation of adenylyl cyclase and phospholipase C by a mutant alpha 1B-adrenergic receptor transgene promotes malignant transformation of thyroid follicular cells. Endocrinology 138:369–378. https://doi.org/10.1210/ENDO.138.1.4861 81. Driessens N, Versteyhe S, Ghaddhab C et al (2009) Hydrogen peroxide induces DNA single- and double-strand breaks in thyroid cells and is therefore a potential mutagen for this organ. Endocr Relat Cancer 16:845–856. https://doi.org/10.1677/ERC-09-0020 82. Ghaddhab C, Kyrilli A, Driessens N et al (2019) Factors contributing to the resistance of the thyrocyte to hydrogen peroxide. Mol Cell Endocrinol 481:62–70. https://doi.org/10.1016/j. mce.2018.11.010 83. Versteyhe S, Driessens N, Ghaddhab C et al (2013) Comparative analysis of the thyrocytes and T cells: responses to H2O2 and radiation reveals an H2O2-induced antioxidant transcriptional program in thyrocytes. J Clin Endocrinol Metabol 98:E1645–E1654. https://doi.org/10.1210/jc.2013-1266 84. Kyrilli A, Gacquer D, Detours V et al (2020) Dissecting the role of thyrotropin in the DNA damage response in human thyrocytes after 131I, B radiation and H2O2. J Clin Endocrinol Metabol 105:839– 853. https://doi.org/10.1210/clinem/dgz185

243 85. Viglietto G, Chiappetta G, Martinez-Tello FJ et al (1995) RET/PTC oncogene activation is an early event in thyroid carcinogenesis. Oncogene 11:1207–1210 86. Ameziane-El-Hassani R, Boufraqech M, Lagente-Chevallier O et al (2010) Role of H 2 O 2 in RET/PTC1 chromosomal rearrangement produced by ionizing radiation in human thyroid cells. Cancer Res 70:4123–4132. https://doi.org/10.1158/0008-5472.CAN09-4336 87. Ameziane-El-Hassani R, Buffet C, Leboulleux S, Dupuy C (2019) Oxidative stress in thyroid carcinomas: biological and clinical significance. Endocr Relat Cancer 26:R131–R143. https://doi.org/ 10.1530/ERC-18-0476 88. Ron E, Lubin JH, Shore RE et al (2012) Thyroid cancer after exposure to external radiation: a pooled analysis of seven studies. Radiat Res 178:AV43–AV60. https://doi.org/10.1667/RRAV05.1 89. Williams ED, Abrosimov A, Bogdanova T et al (2004) Thyroid carcinoma after Chernobyl latent period, morphology and aggressiveness. Br J Cancer 90:2219–2224. https://doi.org/10. 1038/sj.bjc.6601860 90. Nikiforov YE, Nikiforova MN (2011) Molecular genetics and diagnosis of thyroid cancer. Nat Rev Endocrinol 7:569–580. https://doi.org/10.1038/nrendo.2011.142 91. Suzuki K, Ojima M, Kodama S, Watanabe M (2003) Radiationinduced DNA damage and delayed induced genomic instability. Oncogene 22:6988–6993. https://doi.org/10.1038/SJ.ONC. 1206881 92. Lorimore SA, Coates PJ, Wright EG (2003) Radiation-induced genomic instability and bystander effects: inter-related nontargeted effects of exposure to ionizing radiation. Oncogene 22:7058–7069. https://doi.org/10.1038/SJ.ONC.1207044 93. Ameziane-El-Hassani R, Talbot M, de Souza Dos Santos MC et al (2015) NADPH oxidase DUOX1 promotes long-term persistence of oxidative stress after an exposure to irradiation. Proc Natl Acad Sci 112:5051–5056. https://doi.org/10.1073/pnas.1420707112 94. Detours V, Delys L, Libert F et al (2007) Genome-wide gene expression profiling suggests distinct radiation susceptibilities in sporadic and post-Chernobyl papillary thyroid cancers. Br J Cancer 97:818–825. https://doi.org/10.1038/sj.bjc.6603938 95. Dom G, Tarabichi M, Unger K et al (2012) A gene expression signature distinguishes normal tissues of sporadic and radiationinduced papillary thyroid carcinomas. Br J Cancer 107:994–1000. https://doi.org/10.1038/bjc.2012.302 96. Meitzler JL, Brandman R, Ortiz de Montellano PR (2010) Perturbed Heme binding is responsible for the blistering phenotype associated with mutations in the Caenorhabditis elegans dual oxidase 1 (DUOX1) peroxidase domain. J Biol Chem 285:40991– 41000. https://doi.org/10.1074/jbc.M110.170902 97. Thein MC, Winter AD, Stepek G et al (2009) Combined extracellular matrix cross-linking activity of the peroxidase MLT-7 and the dual oxidase BLI-3 is critical for post-embryonic viability in Caenorhabditis elegans. J Biol Chem 284:17549–17563. https:// doi.org/10.1074/jbc.M900831200 98. Wong JL, Créton R, Wessel GM (2004) The oxidative burst at fertilization is dependent upon activation of the dual oxidase Udx 1. Dev Cell 7:801–814. https://doi.org/10.1016/j.devcel.2004. 10.014 99. Dias FA, Gandara ACP, Queiroz-Barros FG et al (2013) Ovarian dual oxidase (Duox) activity is essential for insect eggshell hardening and waterproofing. J Biol Chem 288:35058–35067. https://doi. org/10.1074/jbc.M113.522201 100. Kumar S, Molina-Cruz A, Gupta L et al (2010) A peroxidase/dual oxidase system modulates midgut epithelial immunity in Anopheles gambiae. Science 327:1644–1648. https://doi.org/10.1126/ science.1184008 101. Jang S, Mergaert P, Ohbayashi T et al (2021) Dual oxidase enables insect gut symbiosis by mediating respiratory network formation.

244 Proc Natl Acad Sci U S A 118:1–12. https://doi.org/10.1073/pnas. 2020922118 102. Geiszt M, Witta J, Baff J et al (2003) Dual oxidases represent novel hydrogen peroxide sources supporting mucosal surface host defense. FASEB J 17:1–14. https://doi.org/10.1096/fj.02-1104fje 103. Conner GE, Salathe M, Forteza R (2002) Lactoperoxidase and hydrogen peroxide metabolism in the airway. Am J Respir Crit Care Med 166:S57–S61. https://doi.org/10.1164/rccm.2206018 104. Forteza R, Salathe M, Miot F et al (2005) Regulated hydrogen peroxide production by Duox in human airway epithelial cells. Am J Respir Cell Mol Biol 32:462–469. https://doi.org/10.1165/rcmb. 2004-0302OC 105. Moskwa P, Lorentzen D, Excoffon KJD et al (2007) A novel host defense system of airways is defective in cystic fibrosis. Am J Respir Crit Care Med 175:174–183. https://doi.org/10.1164/rccm. 200607-1029OC 106. Gattas MV, Forteza R, Fragoso M et al (2009) Oxidative epithelial host defense is regulated by infectious and inflammatory stimuli. Free Radic Biol Med 47:1450–1458. https://doi.org/10.1016/j. freeradbiomed.2009.08.017 107. Sarr D, Toth E, Gingerich A, Rada B (2018) Antimicrobial actions of dual oxidases and lactoperoxidase. J Microbiol 56:373–386. https://doi.org/10.1007/s12275-018-7545-1 108. Linderholm AL, Onitsuka J, Xu C et al (2010) All- trans retinoic acid mediates DUOX2 expression and function in respiratory tract epithelium. Am J Phys Lung Cell Mol Phys 299:L215–L221. https://doi.org/10.1152/ajplung.00015.2010 109. Fischer H, Gonzales LK, Kolla V et al (2007) Developmental regulation of DUOX1 expression and function in human fetal lung epithelial cells. Am J Physiol Lung Cell Mol Physiol 292: L1506–L1514. https://doi.org/10.1152/ajplung.00029.2007 110. Luxen S, Noack D, Frausto M et al (2009) Heterodimerization controls localization of Duox-DuoxA NADPH oxidases in airway cells. J Cell Sci 122:1238–1247. https://doi.org/10.1242/jcs. 044123 111. Conner GE, Wijkstrom-Frei C, Randell SH et al (2007) The lactoperoxidase system links anion transport to host defense in cystic fibrosis. FEBS Lett 581:271–278. https://doi.org/10.1016/j. febslet.2006.12.025 112. Rada B, Lekstrom K, Damian S et al (2008) The pseudomonas toxin pyocyanin inhibits the dual oxidase-based antimicrobial system as it imposes oxidative stress on airway epithelial cells. J Immunol 181:4883–4893. https://doi.org/10.4049/jimmunol.181. 7.4883 113. Harper RW, Xu C, Eiserich JP et al (2005) Differential regulation of dual NADPH oxidases/peroxidases, Duox1 and Duox2, by Th1 and Th2 cytokines in respiratory tract epithelium. FEBS Lett 579: 4911–4917. https://doi.org/10.1016/j.febslet.2005.08.002 114. Hill T, Xu C, Harper RW (2010) IFNγ mediates DUOX2 expression via a STAT-independent signaling pathway. Biochem Biophys Res Commun 395:270–274. https://doi.org/10.1016/j. bbrc.2010.04.004 115. Iwasaki A, Medzhitov R (2015) Control of adaptive immunity by the innate immune system. Nat Immunol 16:343–353. https://doi. org/10.1038/ni.3123 116. Grandvaux N, Soucy-Faulkner A, Fink K (2007) Innate host defense: Nox and Duox on phox’s tail. Biochimie 89:1113–1122. https://doi.org/10.1016/j.biochi.2007.04.008 117. Yang C-S, Shin D-M, Kim K-H et al (2009) NADPH oxidase 2 interaction with TLR2 is required for efficient innate immune responses to mycobacteria via cathelicidin expression. J Immunol 182:3696–3705. https://doi.org/10.4049/jimmunol.0802217 118. Lv J, He X, Wang H et al (2017) TLR4-NOX2 axis regulates the phagocytosis and killing of Mycobacterium tuberculosis by macrophages. BMC Pulm Med 17:194–202. https://doi.org/10. 1186/s12890-017-0517-0

F. Miot and X. De Deken 119. Joo J-H, Ryu J-H, Kim C-H et al (2012) Dual oxidase 2 is essential for the toll-like receptor 5-mediated inflammatory response in airway mucosa. Antioxid Redox Signal 16:57–70. https://doi.org/ 10.1089/ars.2011.3898 120. Boots AW, Hristova M, Kasahara DI et al (2009) ATP-mediated activation of the NADPH oxidase DUOX1 mediates airway epithelial responses to bacterial stimuli. J Biol Chem 284:17858– 17867. https://doi.org/10.1074/jbc.M809761200 121. Habibovic A, Hristova M, Heppner DE et al (2016) DUOX1 mediates persistent epithelial EGFR activation, mucous cell metaplasia, and airway remodeling during allergic asthma. JCI Insight 1:1–25. https://doi.org/10.1172/jci.insight.88811 122. Hristova M, Habibovic A, Veith C et al (2016) Airway epithelial dual oxidase 1 mediates allergen-induced IL-33 secretion and activation of type 2 immune responses. J Allergy Clin Immunol 137:1545–1556. https://doi.org/10.1016/j.jaci.2015.10.003 123. Chang S, Linderholm A, Franzi L et al (2013) Dual oxidase regulates neutrophil recruitment in allergic airways. Free Radic Biol Med 65:38–46. https://doi.org/10.1016/j.freeradbiomed. 2013.06.012 124. Fink K, Martin L, Mukawera E et al (2013) IFNβ/TNFα synergism induces a non-canonical STAT2/IRF9-dependent pathway triggering a novel DUOX2 NADPH oxidase-mediated airway antiviral response. Cell Res 23:673–690. https://doi.org/10.1038/cr.2013.47 125. Strengert M, Jennings R, Davanture S et al (2014) Mucosal reactive oxygen species are required for antiviral response: role of Duox in influenza a virus infection. Antioxid Redox Signal 20:2695–2709. https://doi.org/10.1089/ars.2013.5353 126. Grandvaux N, Mariani M, Fink K (2015) Lung epithelial NOX/DUOX and respiratory virus infections. Clin Sci 128:337– 347. https://doi.org/10.1042/CS20140321 127. Kim HJ, Seo YH, An S et al (2018) Chemiluminescence imaging of Duox2-derived hydrogen peroxide for longitudinal visualization of biological response to viral infection in nasal mucosa. Theranostics 8:1798–1807. https://doi.org/10.7150/thno.22481 128. Ioannidis I, McNally B, Willette M et al (2012) Plasticity and virus specificity of the airway epithelial cell immune response during respiratory virus infection. J Virol 86:5422–5436. https://doi.org/ 10.1128/JVI.06757-11 129. Mariani MK, Dasmeh P, Fortin A et al (2019) The combination of IFN β and TNF induces an antiviral and immunoregulatory program via non-canonical pathways involving STAT2 and IRF9. Cell 8:919. https://doi.org/10.3390/cells8080919 130. Cegolon L, Salata C, Piccoli E et al (2014) In vitro antiviral activity of hypothiocyanite against A/H1N1/2009 pandemic influenza virus. Int J Hyg Environ Health 217:17–22. https://doi.org/10. 1016/j.ijheh.2013.03.001 131. Mikola H, Waris M, Tenovuo J (1995) Inhibition of herpes simplex virus type 1, respiratory syncytial virus and echovirus type 11 by peroxidase-generated hypothiocyanite. Antivir Res 26:161–171. https://doi.org/10.1016/0166-3542(94)00073-h 132. Goto Y, Ivanov II (2013) Intestinal epithelial cells as mediators of the commensal–host immune crosstalk. Immunol Cell Biol 91: 204–214. https://doi.org/10.1038/icb.2012.80 133. Burgueño JF, Fritsch J, Santander AM et al (2019) Intestinal epithelial cells respond to chronic inflammation and dysbiosis by synthesizing H2O2. Front Physiol 10:1484. https://doi.org/10. 3389/fphys.2019.01484 134. Ameziane-El-Hassani R, Benfares N, Caillou B et al (2005) Dual oxidase2 is expressed all along the digestive tract. Am J Physiol Lung Cell Mol Physiol 288:G933–G942. https://doi.org/10.1152/ ajpgi.00198.2004 135. Grasberger H, Gao J, Nagao-Kitamoto H et al (2015) Increased expression of DUOX2 is an epithelial response to mucosal dysbiosis required for immune homeostasis in mouse intestine.

14

DUOX1 and DUOX2, DUOXA1 and DUOXA2

Gastroenterology 149:1849–1859. https://doi.org/10.1053/j.gastro. 2015.07.062 136. Ha E-M, Oh C-T, Bae YS, Lee W-J (2005) A direct role for dual oxidase in drosophila gut immunity. Science 310:847–850. https:// doi.org/10.1126/science.1117311 137. Ha E-M, Oh C-T, Ryu J-H et al (2005) An antioxidant system required for host protection against gut infection in drosophila. Dev Cell 8:125–132. https://doi.org/10.1016/j.devcel.2004.11.007 138. Chávez V, Mohri-Shiomi A, Garsin D (2009) Ce-Duox1/BLI-3 generates reactive oxygen species as a protective innate immune mechanism in Caenorhabditis elegans. Infect Immun 77:4983– 4989. https://doi.org/10.1128/IAI.00627-09 139. Flores MV, Crawford KC, Pullin LM et al (2010) Dual oxidase in the intestinal epithelium of zebrafish larvae has anti-bacterial properties. Biochem Biophys Res Commun 400:164–168. https:// doi.org/10.1016/j.bbrc.2010.08.037 140. Grasberger H, El-Zaatari M, Dang DT, Merchant JL (2013) Dual oxidases control release of hydrogen peroxide by the gastric epithelium to prevent helicobacter felis infection and inflammation in mice. Gastroenterology 145:1045–1054. https://doi.org/10.1053/j. gastro.2013.07.011 141. Allaoui A, Botteaux A, Dumont JE et al (2009) Dual oxidases and hydrogen peroxide in a complex dialogue between host mucosae and bacteria. Trends Mol Med 15:571–579. https://doi.org/10. 1016/j.molmed.2009.10.003 142. Botteaux A, Hoste C, Dumont JE et al (2009) Potential role of Noxes in the protection of mucosae: H2O2 as abacterial repellent.

245 Microbes Infect 11:537–544. https://doi.org/10.1016/j.micinf. 2009.02.009 143. Haberman Y, Tickle TL, Dexheimer PJ et al (2014) Pediatric Crohn disease patients exhibit specific ileal transcriptome and microbiome signature. J Clin Investig 124:3617–3633. https://doi. org/10.1172/JCI75436 144. Hayes P, Dhillon S, O’Neill K et al (2015) Defects in nicotinamideadenine dinucleotide phosphate oxidase genes NOX1 and DUOX2 in very early onset inflammatory bowel disease. Cell Mol Gastroenterol Hepatol 1:489–502. https://doi.org/10.1016/j. jcmgh.2015.06.005 145. Parlato M, Charbit-Henrion F, Hayes P et al (2017) First identification of biallelic inherited DUOX2 inactivating mutations as a cause of very early onset inflammatory bowel disease. Gastroenterology 153:609–611.e3. https://doi.org/10.1053/j.gastro.2016. 12.053 146. Levine AP, Pontikos N, Schiff ER et al (2016) Genetic complexity of Crohn’s disease in two large Ashkenazi Jewish families. Gastroenterology 151:698–709. https://doi.org/10.1053/j.gastro.2016. 06.040 147. Grasberger H, Noureldin M, Kao TD et al (2018) Increased risk for inflammatory bowel disease in congenital hypothyroidism supports the existence of a shared susceptibility factor. Sci Rep 8:10158– 10163. https://doi.org/10.1038/s41598-018-28586-5 148. Grasberger H, Magis AT, Sheng E et al (2021) DUOX2 variants associate with preclinical disturbances in microbiota-immune homeostasis and increased inflammatory bowel disease risk. J Clin Investig 131:e141676. https://doi.org/10.1172/JCI141676

Part III NADPH Oxidase Regulators

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1)

15

Pham My-Chan Dang and Jamel El-Benna

Abstract

Keywords

The enzyme responsible for superoxide anion production in phagocytes is called the phagocyte NADPH oxidase. It is a multicomponent enzyme system resulting from the assembly upon activation of four cytosolic proteins (p47phox, p67phox, p40phox and Rac1 or Rac2) with two transmembrane proteins (p22phox and gp91phox, which form the cytochrome b558). The gp91phox is the catalytic subunit of the phagocyte NADPH oxidase and was the first NADPH oxidase to be discovered, renamed today as NOX2. Since then, a family of NOX enzymes, comprising NOX1 to NOX5 and the two DUOX, DUOX1 and DUOX2 has been characterized. NOX1 was the first homologue of gp91phox to be identified, and now refers to a multicomponent enzyme complex composed of three cytosolic proteins (NOXO1, a p47phox homologue, NOXA1, a p67phox homologue and Rac1) with two transmembrane proteins (NOX1 and p22phox). NOX1and NOX2-derived ROS are essential for innate immunity and other physiological functions; however, excessive ROS production can induce tissue injury, contributing to inflammatory diseases. Thus, NOX1 and NOX2 activation must be tightly regulated in time and space in order to limit ROS production. p47phox and NOXO1 play a major role in the regulation and organization of the NOX2 and NOX1 complexes, respectively, through the interactions of specific protein domains and via phosphorylation. This chapter aims to provide new insights on the role of p47phox and NOXO1 in NOX2 and NOX1 regulation and activation.

NADPH oxidase · NOX2 · NOX1 · p47phox · NOXO1 · Protein phosphorylation · Neutrophil · Epithelial cells

Pham My-Chan Dang and Jamel El-Benna contributed equally to this chapter. P. M.-C. Dang · J. El-Benna (✉) Centre de Recherche sur l’Inflammation (CRI), INSERM-U1149, CNRS-ERL8252, Laboratoire d’Excellence Inflamex, Université de Paris, Faculté de Médecine Xavier Bichat, Paris, France e-mail: [email protected]; [email protected]

1

Overview

Phagocytic cells such as polymorphonuclear neutrophils, eosinophils, monocytes and macrophages are the first line of defense against microbes such as bacteria, parasites and fungi [1–3]. Upon phagocytosis of microbes, phagocytes produce high amounts of superoxide anion (O2-.), which is the source of other reactive oxygen species (ROS) such as hydrogen peroxide (H2O2) and hypochlorous acid (HOCl) [4–6]. From a historitical point of view, initial experiments in 1932 by Baldridge and Gerard using a Warburg manometer found that phagocytosis was accompagnied by a “burst of oxygen consumption” [7]. In 1961, Iyer et al. used [14C]formate, which is oxidized to 14CO2 in the presence of catalase, and reported the production of hydrogen peroxide by neutrophils [8]. The source of hydrogen peroxide remained unknown until the findings by Babior et al. in 1973 [9]. Using the cytochrome c reduction assay in combination with superoxide dismutase, they discovered that neutrophils produced superoxide anion that was then converted into hydrogen peroxide. The rapid increase in oxygen uptake and the abrupt ROS production that occured during neutrophil activation was then called the “respiratory burst”. The discovery by Babior et al. [9] opened up a new research field in phagocyte biology. The enzyme producing superoxide anion was called the phagocyte NADPH oxidase because it uses cytosolic NADPH as the electron donor to reduce molecular oxygen into superoxide anion [10]. The enzyme was characterized in neutrophils, monocytes and macrophages and its membrane and cytosolic components were identified in the eighties and nineties by several groups (reviewed in [11, 12]).

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_15

249

250

P. M.-C. Dang and J. El-Benna

Today, the structure of the phagocyte NADPH oxidase is well known. It is a multicomponent enzyme system, which is composed of several membrane and cytosolic proteins that are assembled upon activation (reviewed in [11, 12]). The phagocyte NADPH oxidase has an efficient electron transporter, which was previously referred as the cytochrome b558 for its spectral absorbance signature, and is a heterodimer comprised of two transmembrane proteins, e.g., gp91phox (phox: phagocyte oxidase) and p22phox. The activity of the cytochrome b558 can be switched “ON” and “OFF” and is regulated by the four cytosolic proteins, p47phox, p67phox, p40phox and Rac2 (in neutrophils) or Rac1 (in monocytes and macrophages) (Fig. 15.1). The gp91phox is the core enzyme as it can bind NADPH and FAD, has two hemes, and is able to transfer electrons from NADPH to oxygen to produce superoxide anion [13, 14]. p22phox is a stabilizing subunit for gp91phox and binds to p47phox during activation [15, 16]. p47phox is the regulatory subunit that organizes the assembly of the complex as it interacts with the cytosolic “p67phox-p40phox complex” to allow translocation of all three proteins from the cytosol to the membranes [17, 18]. p67phox is considered as the activator subunit as it binds and stimulates gp91phox enzymatic activity via an activation domain (AD) [19, 20]. The GTPases Rac1 or Rac2 are also required for gp91phox enzymatic activity in monocytes/macrophages and neutrophils, respectively [21, 22]. Finally, p40phox is an

enhancer of the phagocyte NADPH oxidase activation [23, 24]. The vital importance of the phagocyte NADPH oxidase is demonstrated by the genetic immunodeficiency disorder called chronic granulomatous disease (CGD) in which phagocytes do not produce ROS and affected patients have infections often more fatal than in healthy individuals [25, 26]. CGD is a rare disease found in 1 over 200,000 to 250,000 individuals. It is caused by a gene mutation in one of the NADPH oxidase components, the most frequent CGD form being caused by mutations in the gp91phox/CYBB gene (65% of CGD patients), followed by mutation in the p47phox/NCF1 gene (20% of CGD patients), the p22phox/ CYBA gene (less than 5% of CGD patients), the p67phox/ NCF2 gene (less than 5% of CGD patients), and the p40phox/ NCF4 and Rac2/RAC2 genes (both forms less than 5% of CGD patients) [25, 26]. ROS production by various non-phagocytic cells such as epithelial and endothelial cells was described for years, although the sources of ROS were not identified [27– 29]. Based on a sequence homology search with the gp91phox/CYBB gene, Lambeth’s group described the first human homologue of gp91phox, which was called NOX1 (for NADPH oxidase 1) [30]. Later, several other homologues of gp91phox were described in several tissues such as kidney, lung, colon, inner ear and thyroid [31– 37]. These homologues were then cloned and grouped

NOX1

The phagocyte NADPH oxidase (NOX2) 2O2㼻

2 O2

2 O2

2e-

NOX2/ gp91phox p p

Rac2

2e-

p22phox

NOX1

FAD NADPH

p40phox p67phox p

2O2㼻 -

p47phox

p p p

Rac1

p22phox

FAD NADPH

NOXA1

NOXO1

p p

Fig. 15.1 Structure of the phagocyte NADPH oxidase (NOX2) and of NOX1. The active phagocyte NADPH oxidase complex is composed of two membrane proteins (gp91phox/NOX2 and p22phox) and four proteins (p67phox, p47phox, p40phox and Rac1 or Rac2) that are present in the cytosol prior to activation. The active phagocyte NOX1

complex is composed of two membrane proteins (NOX1 and p22phox) and three cytosolic proteins (NOXO1, NOXA1 and Rac1). The activated NADPH oxidase uses cytosolic NADPH to reduce oxygen into superoxide anion. (P) denotes phosphorylation

15

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1)

under the acronym “NOX”, for NADPH oxidase (NOX1 to NOX5) or “DUOX” for dual oxidase (DUOX1 and DUOX2) [38, 39]. The major difference between the phagocytic gp91phox (now called NOX2) and the other NOXs is that NOX2 requires stimulation to form an active complex between membrane and cytosolic components, and ROS production by NOX2 occurs on the external face of the plasma membrane, releasing high amounts of ROS into the phagosome or the extracellular space, whereas the other NOXs produce lower amounts of ROS [30–39]. These data suggested that ROS play a role in several cellular functions such as local tissue-specific bactericidal activity and intracellular signaling, stimulating major scientific interest in understanding the mechanisms controlling their production. NOX1 shares 56% amino-acid identity with NOX2 [30], and like NOX2, it has six transmembrane domains, contains two hemes, and binds FAD and NADPH. NOX1 mRNA is highly expressed in epithelial cells from the colon, but is also detected in the stomach, uterus, prostate, and smooth muscle cells [38, 39]. As with NOX2, NOX1 is expressed with p22phox in membranes and is regulated by Rac1. However, NOX1 has specific cytosolic partners, i.e., NOX organizer 1 (NOXO1, the homologue of p47phox), and NOX activator 1 (NOXA1, the homologue of p67phox) [40–42] (Fig. 15.1). p47phox and NOXO1 are the organizer and regulatory subunits of NOX2 and NOX1, respectively, and their properties are discussed in this chapter.

2

p47phox, the Phagocyte NADPH Oxidase Organizer and Regulator

2.1

History of the Discovery of p47phox

4

p47phox

NH2

recessive CGD [50]. Several groups then attempted to purify it in order to produce specific antibodies to sequence the protein and clone the gene [47–49]. Cloning of the p47phox/ Neutrophil Cytosolic Factor 1 (NCF1) gene was achieved in 1989 by two different groups [51, 52].

2.2

Structure of p47phox

The human p47phox/NCF1 gene sequence encodes for a protein composed of 390 amino acids, with an estimated molecular weight of 44.7 kDa, which migrates at almost 47 kDa by SDS-polyacrylamide gel electrophoresis [51, 52]. It is abundant in the cytosol of human neutrophils as it has been estimated to be present at 100–150 ng/106 cells [53, 54]. Native non-phosphorylated p47phox is a very basic protein with an isoelectric point of 9.8 [55, 56]. Its N-terminal amino acid sequence has a phox homology domain (PX; amino acids 4–128), and the center of the protein contains two Src homology 3 (SH3) domains (amino acids 156–216 (SH3A) and amino acids 226–286 (SH3B)) [11, 51, 52]. The C-terminal sequence of p47phox is very basic, rich in serines as potential phosphorylation sites, and contains an autoinhibitory region (AIR) (amino acids 297–340) and a proline rich region (PRR) (amino acids 360–368) [11, 12, 57] (Fig. 15.2).

2.3

The discovery that the NADPH oxidase could be activated in a cell-free system [43–46], suggested that the enzyme required membrane and cytosolic factors to be catalically active [47–49]. p47phox was first identified in 1985 as a 44-kDa phosphorylated protein that was missing in stimulated neutrophils from patients with autosomal 128

PX

303

251

Phosphorylation of p47phox

Using 32P labeling of neutrophils, it was shown that several agonists that stimulate ROS production, such as phorbol 12-myristate 13-acetate (PMA), formyl-Met-Leu-Phe (fMLF) and opsonized zymosan induced the phosphorylation of p47phox in human neutrophils [58, 59]. Two-dimensional gel electrophoresis analyses suggested that p47phox was phosphorylated on several sites as several forms of p47phox were detected in stimulated neutrophils [55, 56]. Phosphorylation studies showed that p47phox is a good substrate for protein kinase C (PKC), protein kinase A 156

216 226

SH3

286 297

SH3

340 360-368 390

AIR

PRR

COOH

379

SSIRNAHSIHQRSRKRLSQDAYRRNSVRFLQQ RRRQARPGPQSPGSPLEEERQTQRSKPQPAVPPRPSADLILNRCS

Fig. 15.2 Structure of human p47phox. The human p47phox protein has 390 amino acids, organized into multiple domains. A phox domain (PX) (amino acids 4–128), two Src homology 3 (SH3) domains [SH3A (amino acids 156–216) and SH3B (amino acids 226–286)], a proline-

rich region (PRR) (amino acids 360–367), and an autoinhibitory region (AIR) (amino acids 297–341). The COOH-terminal sequence, including the AIR domain, contains several serines that can potentially be phosphorylated

252

(PKA) and other protein kinases [57, 60–63]. The phosphorylated sites were then localized in the AIR domain at the carboxy-terminal region of the protein between Ser303 and Ser379 [64, 65]. Most of the phosphorylation sites are localized in an arginine-rich region with the consensus phosphorylation sequence “RRXSXR”, which is targeted by PKC and a proline-rich region with the consensus phosphorylation sequence “PXSP”, which is a substrate for the mitogen activated protein kinases (MAPK) [64, 65].

2.3.1

Phosphorylation of p47phox Is Required for NADPH Oxidase/NOX2 Activation PMA, the direct activator of PKC, induces a strong activation of NOX2 in neutrophils, a process correlating with the phosphorylation of p47phox [60, 66]. In addition, the use of PKC selective inhibitors suggested that the PKC-dependent phosphorylation of p47phox was involved in NOX2 activation [67]. The proof that the phosphorylation of p47phox was required for NOX2 activation in intact cells was provided by site-directed mutagenesis studies. These studies clearly showed that p47phox phosphorylation is required for PMA-, fMLF- and IgG-mediated activation of the NADPH oxidase in EBV-transformed lymphocytes, B lymphocytes and COS-7 cells stably expressing gp91phox, p22phox, p47phox and p67phox (COSphox cells) [68–70]. Individual mutation of each identified serine showed that only Ser379 is essential for oxidase activation [68–70], while double mutation analysis showed that two pairs of phosphorylated serines, Ser(303 + 304) and Ser(359 + 370), are necessary for optimal NADPH oxidase activation [71, 72]. The development of a cell-free system based on the phosphorylation of p47phox by PKC, provided further evidence of a direct role for p47phox phosphorylation in the activation of the NADPH oxidase. The involvement of a cytosolic PKC in the activation of the NADPH oxidase was first shown in a cell-free system by Tauber and colleagues [73]. McPhail et al. showed that phosphatidic acid and diacylglycerol were able to activate the NADPH oxidase in a cell-free system in a phosphorylation-dependent manner [74]. Babior and colleagues developed a PKC-dependent cell-free system using recombinant p47phox and showed that mutation of p47phox Ser379 to Ala inhibited the activation [75, 76]. Interestingly, Babior’s group subsequently showed that p47phox phosphorylated by PKC could activate the NADPH oxidase not only in a cell-free system containing neutrophil membrane and cytosol, but also in a system in which the cytosol was replaced by p67phox, Rac2 and phosphorylated p47phox recombinant proteins, suggesting that the neutrophil membrane plus those three cytosolic proteins are both necessary and sufficient for NADPH oxidase activation [77]. In addition to PKC, the protein kinase Akt (protein kinase B), which depends on phosphatidylinositol 3-kinase for activation, was

P. M.-C. Dang and J. El-Benna

shown to phosphorylate p47phox on serines Ser304 and Ser328 and activate the NADPH oxidase in a cell free system [78].

2.3.2

Phosphorylation of p47phox Regulates NADPH Oxidase/NOX2 Priming Depending on the environment where neutrophils reside, the phagocyte NADPH oxidase can either be in a “resting/dormant” state, in a “primed/pre-activated” state, or in a fully “activated” state, [6, 79]. Resting NADPH oxidase is mainly present in circulating neutrophils, while primed NADPH oxidase is found in adherent neutrophils and in neutrophils in contact with pro-inflammatory agents such as the pro-inflammatory cytokines, tumor necrosis factor alpha (TNFα), granulocyte/macrophage-colony stimulating factor (GM-CSF), interleukin-8 (IL-8) and the toll-like recptors (TLR) agonists, lipopolysaccharide (LPS) and CL097, the highly water-soluble derivative of the imidazoquinoline compound Resiquimod [6, 79]. These pro-inflammatory priming agents do not activate the NADPH oxidase, but they prime it for full activation in response to a second stimulus such as the bacterial peptide fMLF. Priming agents have been shown to induce partial phosphorylation of p47phox on Ser345 [80– 82] and translocation of the cytochrome b558 to the plasma membrane [83, 84], which are critical for the priming of ROS production in neutrophils. Ser345 is located within a MAPK consensus phosphorylation site, and its phosphorylation involves p38MAPK under TNFα stimulation and the extracellular-regulated kinases 1 and 2 (ERK1/2) under GM-CSF stimulation [80]. Importantly, phosphoSer-345 is a binding site for the proline isomerase Pin1 [85], an enzyme that recognizes and catalyzes the cis-trans isomerization of phospho-Ser/Thr-Pro peptide bonds [86]. Pin1 is likely to play a major role in the activation of the NADPH oxidase. Pin1 induces p47phox conformational changes and NADPH oxidase hyper-activation through its activation by the priming agents, TNFα, CL097, LPS and fMLF in intact neutrophils. Furthermore, inhibition of Pin1 activation prevents the priming process [81, 82, 85]. Activated Pin1 binds to p47phox when it is phosphorylated on Ser-345, then catalyses the conformational change of p47phox necessary for the subsequent phosphorylation of p47phox on other sites by PKC, thereby allowing the hyper-activation of the NADPH oxidase [85]. 2.3.3

Phosphorylation of p47phox Is Regulated by p67phox and p40phox In a recent study, our team used phospho-specific antibodies that we developed against five major p47phox-phosphorylated sites (phospho-Ser304, -Ser315, -Ser320, -Ser328 and -Ser345), and found that phosphorylation of p47phox on these serine residues was dramatically reduced in neutrophils

15

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1)

isolated from p67phox-deficient CGD patients (p67phox-/-) [87]. This finding was confirmed in Epstein-Barr virus (EBV)transformed B- lymphocytes from p67phox-/- CGD patients and in COSphox cells transfected with all the NADPH oxidase components, except for p67phox. In vitro studies showed that recombinant p47phox was phosphorylated on Ser304, Ser315, Ser320 and Ser328 by different PKC isoforms and the addition of recombinant p67phox alone or in combination with p40phox potentiated this process [87]. These new data demonstrated that p67phox and p40phox are required for optimal p47phox phosphorylation on Ser304, Ser315, Ser320 and Ser328 in intact cells. Therefore, p67phox and p40phox are novel regulators of p47phox-phosphorylation.

2.3.4

Phosphorylation of p47phox Induces p47phox Conformational Changes and its Interaction with p22phox Cloning of the p47phox gene revealed the presence of two SH3 domains (amino acids 159–214 (SH3A) and amino acids 229–284 (SH3B)) [51, 52]. The SH3 domain is a protein sequence of about 60 amino acids, first identified in the Src protein tyrosine kinases and later in other proteins involved in signal transduction [88], which mediates proteinprotein interaction by binding to PRR sequences [89]. In resting cells, 100% of the p47phox are located in the cytosol and are not phosphorylated. In this form, the p47phox is in a closed, auto-inhibited state [90, 91]. X-ray structure of the auto-inhibited form of p47phox reveals that the tandem SH3 domains share an interface that forms a shallow groove,

253

which constitutes the peptide binding surface of the p47phox AIR C-terminal domain [92–95]. Furthermore, in the cytosol, p47phox is associated with the p67phoxp40phox complex, where the p47phox-PRR interacts with the p67phox C-terminal SH3 domain [90, 96]. During NADPH oxidase activation in intact cells, approximately 10–20% of p47phox migrate to the cell membranes after being phosphorylated, while 80–90% of p47phox remain in the cytosol [97, 98]. Binding of p47phox to the cytochrome b558 (gp91phox/NOX2 and p22phox) is required for activation of the NADPH oxidase complex as translocation of p47phox to the plasma membrane is impaired in neutrophils from gp91phox- or p22phoxdeficient CGD patients [16, 18]. The p47phox-p22phox interaction is the best-known interaction in the NOX field. Sumimoto et al. [90] and Leto et al. [91] used recombinant p47phox-SH3 domains to show that in vitro, they interact with the cytosolic tail of p22phox via the PRR sequence (amino acids 151–160: PPSNPPPRPP). The requirement of the p47phox-SH3 domains for p47phox translocation and NADPH oxidase activation was later confirmed by deletion analysis of p47phox in whole cells [99]. Native, non-phosphorylated p47phox does not interact with p22phox; however, phosphorylation of p47phox in its C-terminal region introduces negative charges that inhibit the intra-protein interaction and open the closed conformation allowing the two SH3 domains of p47phox to then interact with the p22phox-PRR [90–92] (Fig. 15.3). In addition to the p47phox-p22phox interaction, it was shown that

gp91phox /NOX2

p22phox

PX

PRR

N ter

SH3 A AIR pp N-ter

PX

SH3B

Protein Kinases

SH3A AIR

PRR

C-ter

p47phox (Closed conformation) Fig. 15.3 Closed and open conformations of p47phox. In the native non-phosphorylated state, the p47phox-SH3 tandem domains interact with the AIR domain, keeping p47phox in a closed, inactive conformation. Phosphorylation of p47phox on multiple residues in its AIR

PRR

p

p

p

Cter

Phosphorylated-p47phox (Open Conformation) domain inhibits the two-SH3/AIR intra-molecular interaction, resulting in the p47phox-SH3 domains to be able to bind to the p22phox PRR sequence upon translocation. The p47phox-PX domain interacts with membrane phospholipids

254

P. M.-C. Dang and J. El-Benna

p47phox interacts with some specific sequences of the gp91phox/NOX2 (amino acids (86–93), (450–457), (494–498) and (554–564)) [100, 101]. However the direct interaction between the native p47phox and gp91phox has not been shown yet. The p47phox PX domain (a sequence of about 125 amino acids) is known to bind to phosphatidylinositol 3,4bisphosphate (PtdIn(3,4)-P2) and phosphatidic acid [102, 103]. Some studies had proposed that the phosphorylation of the C-terminal region of p47phox could release the PX domain, allowing its binding to (PtdIn(3,4)-P2) and phosphatidic acid [104–106]. This PX-phospholipid interaction might play a role in the translocation of the cytosolic complex to the membranes, as the proximity of p67phox and p40phox to gp91phox/NOX2 allows for the p67phox AD to promote gp91phox enzymatic activity. Most of the studies on the p47phox conformational changes were obtained using in vitro systems with phosphorylated recombinant p47phox protein or in the presence of SDS or arachidonic acid [107–109]. Swain et al. used tryptophan fluorescence and circular dichroism spectroscopy to show that SDS and phosphorylation induced p47phox conformational changes [107]. Park and colleagues used a cysteine labeling approach to show that PKC-mediated phosphorylation and arachidonic acid induced p47phox conformational changes [108, 109]. These data were confirmed by Marcoux et al. using mass spectrometry coupled to hydrogen/ deuterium exchange and limited proteolysis [110]. Interestingly, Shiose and Sumimoto showed that arachidonic acid at low concentrations synergises with phosphorylation of p47phox to induce the conformational change necessary for the interaction with p22phox and the activation of the NADPH oxidase [111]. They also showed that mutation of known phosphorylated serines to alanines resulted in inhibition of PKC-dependent NADPH oxidase activation in vitro [111, 112].

3

NOXO1, the NOX1 Organizer and Regulator

3.1

History of NOXO1 Discovery

NOX1, the homologue of gp91phox/NOX2, was first cloned from the human adenocarcinoma epithelial cell line Caco-2 in 1999 [30]. In normal human tissues, its messenger was found to be mostly expressed in the gastrointestinal tract, specifically in the colon [30]. As ectopic expression of NOX1 alone resulted in minimal superoxide anion generation, it was hypothesized that NOX1, like gp91phox, might require regulatory subunits to be fully active. Banfi et al. demonstrated that NOX1 was able to produce superoxide anion when cells expressing NOX1 were co-transfected with p47phox and

p67phox [40], indicating a functional similarity between NOX1 and NOX2. The expression of p47phox and p67phox being largely restricted to myeloid cells, it was hypothesized that homologues of p47phox and p67phox could be found in cells and tissues expressing NOX1. The cloning of human and mouse cDNAs encoding for proteins homologous to p47phox and p67phox was quickly performed by three independent teams [40–42]. These novel proteins were named NOX Organizer 1 (NOXO1) and NOX Activator 1 (NOXA1) and are the homologues of p47phox and p67phox, respectively. The mRNAs of NOXO1 and NOXA1 were predominantly found in the colon and more precisely in the epithelium [40, 41]. It was then found that co-transfection of NOXO1, NOXA1 and NOX1 in human embryonic kidney 293 (HEK293), COS-7, Chinese hamster ovary (CHO) or HeLa cells resulted in low constitutive activity of the NOX1 complex and that the absence of one of the proteins was sufficient to prevent its activity [40, 113]. Nevertheless, studies with human proteins showed that the constitutive production of superoxide by the NOX1 complex could be increased by pharmacological agents such as PMA [41, 42]. NOXO1 can also operate in a complex with NOX3 in transfected cells. Notably, Cheng et al. showed that NOX3 could be strongly activated by NOXO1 in the absence of NOXA1 [114]. However, the activation of NOX3 induced by NOXO1 is less efficient than that of NOX1, and NOXO1 remains the primary partner of NOX1 [115, 116].

3.2

Structure of NOXO1

Human NOXO1, which is encoded by a gene located on chromosome 16 (16p13.3) and composed of eight exons, is a protein of 370 amino acids that migrates at almost 41 kDa by SDS-polyacrylamide gel electrophoresis [116]. Although the overall amino acids sequence of human NOXO1 has only 27% identity with p47phox, the domain arrangements are well conserved. As for p47phox, NOXO1 N-terminal amino acid sequence has a PX domain (amino acids 2–132) which is involved in the binding of membrane phosphoinositides [117]. However, while the PX domain of p47phox binds to PtdIns (3,4)-P2, which is only produced upon cell activation, studies using phosphatidylinositol arrays showed that the PX domain of NOXO1 binds mainly monophosphorylated phosphatidylinositols such as PtdIns 4-P and PtdIns 5-P, which are present in plasma membranes of non-activated cells [113]. However, further studies using surface plasmon resonance identified PtdIns (4,5)-P2 and PtdIns (3,4,5)-P3 as targets of NOXO1-PX [118]. Arginine 40 (R40) is important for binding to membrane phosphinositides as its mutation to glutamine is sufficient to interfere with lipid binding [113]. However, R40 is largely buried in NOXO1 as shown by nuclear magnetic resonance spectroscopy, and it has been

15

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1)

suggested that the R40Q mutation is more likely to disrupt protein/lipid binding by destabilizing the PX domain rather than by abrogating the specific arginine/phosphatidylinositol interaction [118]. The center of NOXO1 contains two SH3 domains (amino acids 157–218 (SH3A) and 228–291 (SH3B), which are involved in the binding to p22phox [42]. Contrary to p47phox, the C-terminal sequence of NOXO1 does not contain an AIR domain, but it has a PRR sequence (amino acids 320–332: PPPTVPTRPSP) [41, 42] (Fig. 15.3), which could be involved in the interaction with NOXA1 [42] (Fig. 15.4). Four structural variants of human NOXO1 resulting from alternative splicing of both ends of exon 3 encoding the PX domain have been described, e.g., α, β, δ and γ [41, 116, 119, 120]. Lysine 50 is deleted in the α and δ forms, and five additional amino acids are found in the N-terminal PX domain of the δ form. The β and γ forms have no deletion of lysine 50, and γ has also a 5 amino acid insertion in the N-terminal PX domain [116]. The mRNAs encoding NOXO1α and NOXO1δ are found in low abundance in tissues and cells, and when expressed in E. Coli, the PX domains of the respective proteins show low expression, suggesting poor translation or unstability [116]. It is thefore unlikely that these isoforms have a significant biological role. In contrast, the mRNAs encoding NOXO1β and NOXO1γ are found in high abundance in the carcinoma T84 colon epithelial cell line and in testis, respectively, and their PX domains show stable expression in E. Coli. When expressed in HEK293 cells, these variants show differential subcellular localization as NOXO1α and NOXO1δ are found in intracellular vesicles or cytoplasmic aggregates, while NOXO1β is prominent in the plasma membrane, and NOXO1γ is detected in both the plasma membrane and the nucleus [120]. The NOXO1α and NOXO1δ isoforms support low NOX1 activity, whereas NOXO1β and NOXO1γ support high NOX1 activation, especially upon stimulation with PMA. Without PMA, NOXO1γ ability to support NOX1 activity is weaker than NOXO1β. Interestingly, purified PX domains of NOXO1β and NOXO1γ bind with the same affinity to membrane phosphoinositides [116, 119, 120].

4

p47phox

PX

NH2 NH2

Fig. 15.4 Comparaison of the structures of human p47phox and human NOXO1. The human NOXO1 protein has 370 amino acids, organized into multiple domains. Like p47phox, it has a phox domain (PX) (amino

156

157

286 297

216 226

SH3

SH3 132

PX

Phosphorylation of NOXO1

Because NOXO1 is lacking the AIR domain containing the phosphorylation sites for p47phox, it was suggested that NOXO1, unlike p47phox, did not exist in an autoinhibited conformation in resting cells, thereby allowing its constitutive association with p22phox and membrane phosphoinositides [42, 113]. This structural feature could explain the constitutive localization of NOXO1 to the membrane as well as the constitutive activity of the NOX1 complex in transfected cells [41, 42, 113]. Therefore, it was commonly accepted that NOX1 activation was not regulated by the phosphorylation of NOXO1. However, in later studies, it was demonstrated by pull-down and isothermal titration calorimetry approaches with different constructs of NOXO1 that an intramolecular interaction involving the SH3 tandem and the PRR domain existed within NOXO1, despite the absence of the AIR domain, and that disruption of this interaction facilitated NOXO1 binding to p22phox [121, 122]. Interestingly the binding between NOXO1 and NOXA1 was found to be very weak as compared to that of their phagocytic counterparts, p47phox and p67phox (100-fold lower affinities), probably due to the fact that the PRR region of NOXO1 is engaged in the intramolecular interaction and is not available to NOXA1 [122]. The existence of an intramolecular interaction within NOXO1 that could prevent p22phox binding and the observation that PMA could increase NOX1 activity, strongly suggested that posttranslational modifications could regulate NOX1 activation. Debbabi et al. demonstrated that PMA increased NOXO1 phosphorylation in 32P labeled transfected HEK-293 epithelial cells via PKC and identified Ser-154 by phosphopeptide mapping as the major phosphorylated site [123]. Furthermore, PMA-induced phosphorylation on Ser-154 was shown to enhance NOXO1 binding to NOXA1 and to p22phox, allowing optimal ROS production by NOX1 in intact cells [123]. Ser-154 of NOXO1 has also been shown to be targeted by PKA in vitro, although this PKA-induced phosphorylation has not been demonstrated in intact cells [124]. Moreover, it has been shown that PKC could also phosphorylate NOXO1 on Thr-341 in vitro and in CHO-tranfected cells [124]. Substitution of this threonine with alanine decreased

128

2

NOXO1

3.3

255

218 228

SH3

340 360-368 390

AIR 291

SH3

COOH

PRR 370

PRR

COOH

acids 2–132), two Src homology 3 (SH3) domains [SH3A (amino acids 157–218) and SH3B (amino acids 228–291)] and a proline rich region (PRR) (amino acids 320–329), but no AIR domain

256

P. M.-C. Dang and J. El-Benna

PMA-dependent phosphorylation of NOXO1 in these cells, as assessed using the phosphoprotein dye Phos-Tag, and reduced PMA-induced activity [124]. Phosphorylation of NOXO1 on Thr341 seems to be required for its interaction with NOXA1 as shown by in vitro experiments using truncated NOXO1-(154–371) and purified NOXA1-SH3 domain [124].

3.4

NOXO1 Expression

Cellular and tissular expression of NOXO1 has been mostly documented at the mRNA level. Except for its predominant expression in the colon, NOXO1 mRNA has also been detected in the kidney, liver, pancreas, uterus, testis, inner ear, and the vasculature [40, 42, 115, 125, 126]. Increased expression of NOXO1 mRNA was described in experimental gastric tumor development and in human colon cancers as compared with adjacent normal bowel mucosa [127, 128]. NOXO1 protein expression most often requires a triggering signal such as inflammatory mediators or growth factors. Therefore, pro-inflammatory cytokines including interferon gamma (INFγ), interleukin (IL)-1β, TNFα and IL-17 [129–131], and pathogens such as Escherichia coli LF82 can all increase NOXO1 protein expression along with ROS production in intestinal epithelial cells [132]. In addition, the prototype anti-inflammatory cytokine, IL-10, prevented increased expression of the NOXO1 protein induced by TNFα and INFγ in the colonic epithelial cell line [133]. Interestingly, increased protein expression of NOXO1 has been observed in colon biopsies of patients with Crohn disease in the inflamed and non-inflamed areas, as compared with healthy controls [131] and in TNFα-induced colitis in mice [134]. Proteasomal degradation of NOXO1 could also regulate the levels of the NOXO1 protein. The ubiquitination, i.e., the ligation of ubiquitin molecules, to proteins plays an essential role in their specific degradation by the proteasome. This process is catalyzed by three main groups of enzymes, namely the E1 (ubiquitin-activating enzyme), E2 (ubiquitin-conjugating enzyme), and E3 (ubiquitin–protein ligase). Ubiquitin is activated by E1 then transferred to the carrier E2 enzyme, which in turn conjugates ubiquitin to substrate proteins with the help of a specific E3 ligase [135]. In this regard, it has been shown in HCT116 colon cancer cells that NOXO1 interacted with Grb2, which in turn recruited the Casitas B-lineage lymphoma (Cbl)-E3 ligase, leading to ubiquitination and degradation of NOXO1. Epidermal growth factor (EGF)-mediated phosphorylation of NOXO1 on Ser-154, induced its dissociation from Grb2/Cbl, thereby inhibiting the rapid degradation of NOXO1 [136]. Consequently, NOXO1 association with NOXA1 facilitated the stimulation of ROS generation. Interestingly, the expression and stability of NOXO1 were shown to

signicantly increase in human colon cancer tissues compared with normal colon [136]. More recently, Cylindromatosis tumor suppressor protein (CYLD), a deubiquitinase, best known as an essential negative regulator of the NFkB pathway, was also identified as a binding partner of NOXO1 and was suggested to act as a potential tumor suppressor by decreasing the stability of the NOXO1 protein and suppressing excessive ROS generation [137].

4

Conclusions

Superoxide production by the phagocyte NADPH oxidase (NOX2) and the epithelial NOX1 is essential for many physiological functions, including host defense. However, excessive superoxide production by these enzymes can induce tissue injury resulting in inflammatory diseases. It is clear today that the phosphorylation of p47phox and the phosphorylation of NOXO1 are crucial for NOX2 and NOX1 regulation; however, the upstream pathways involved in their phosphorylation are still not fully identified. Furthermore, little is known about the dephosphorylation of phosphop47phox and phospho-NOXO1 by phosphatases, which could be involved in limiting NOX1 and NOX2 activation in cells and in the resolution of inflammation. Future studies are required to determine these pathways in neutrophils, monocytes/macrophages and epithelial cells. Identification of the pathways involved in p47phox and NOXO1 phosphorylation will help to develop new strategies to limit ROS production by phagocytes in inflammatory diseases. Acknowledgements This work was supported by grants from Institut National de la Santé et de la Recherche Médicale (INSERM) and Centre National de la Recherche Scientifique (CNRS), Université Paris Diderot, le Labex Inflamex, et l’Association Vaincre la Mucoviscidose (VLM). The authors wish to thank Martine Torres, Ph.D. for her editorial assistance. Conflict of Interest Statement Authors declare that they don’t have any conflict of interest.

References 1. Mantovani A, Cassatella MA, Costantini C et al (2011) Neutrophils in the activation and regulation of innate and adaptive immunity. Nat Rev Immunol 11:519–531. https://doi.org/10.1038/nri3024 2. Nauseef WM, Borregaard N (2014) Neutrophils at work. Nat Immunol 15:602–611. https://doi.org/10.1038/ni.2921 3. Mócsai A (2013) Diverse novel functions of neutrophils in immunity, inflammation, and beyond. J Exp Med 210:1283–1299. https://doi.org/10.1084/jem.20122220 4. Roos D, van Bruggen R, Meischl C (2003) Oxidative killing of microbes by neutrophils. Microbes Infect 5:1307–1315. https://doi. org/10.1016/j.micinf.2003.09.009

15

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1)

5. Nauseef WM (2007) How human neutrophils kill and degrade microbes: an integrated view. Immunol Rev 219:88–102. https:// doi.org/10.1111/j.1600-065X.2007.00550.x 6. El-Benna J, Hurtado-Nedelec M, Marzaioli V et al (2016) Priming of the neutrophil respiratory burst: role in host defense and inflammation. Immunol Rev 273(1):180–193. https://doi.org/10.1111/ imr.12447 7. Baldridge CW, Gerard RW (1932) The extra respiration of phagocytosis. Am J Physiol Legacy Content 103(1):235–236 8. Iyer GYN, Islam MF, Quastel JH (1961) Biochemical aspects of phagocytosis. Nature 192:535–541 9. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. J Clin Invest 52(3):741–744. https://doi.org/ 10.1172/JCI107236 10. Babior BM, Curnutte JT, Kipnes BS (1995) Pyridine nucleotidedependent superoxide production by a cell-free system from human granulocytes. J Clin Invest 56(4):1035–1042. https://doi.org/10. 1172/JCI108150 11. Vignais PV (2002) The superoxide-generating NADPH oxidase: structural aspects and activation mechanism. Cell Mol Life Sci 59: 1428–1459. https://doi.org/10.1007/s00018-002-8520-9 12. Groemping Y, Rittinger K (2005) Activation and assembly of the NADPH oxidase: a structural perspective. Biochem J 386:401– 416. https://doi.org/10.1042/BJ20041835 13. Rotrosen D, Yeung CL, Leto TL et al (1992) Cytochrome b558: the flavin-binding component of the phagocyte NADPH oxidase. Science 256(5062):1459–1462. https://doi.org/10.1126/science. 1318579 14. Koshkin V, Pick E (1993) Generation of superoxide by purified and relipidated cytochrome b559 in the absence of cytosolic activators. FEBS Lett 327(1):57–62. https://doi.org/10.1016/ 0014-5793(93)81039-3 15. Yu L, Zhen L, Dinauer MC (1997) Biosynthesis of the phagocyte NADPH oxidase cytochrome b558. Role of heme incorporation and heterodimer formation in maturation and stability of gp 91phox and p22phox subunits. J Biol Chem 272(43):27288–27294. https:// doi.org/10.1074/jbc.272.43.27288 16. Leusen JH, Bolscher BG, Hilarius PM et al (1994) 156Pro-->Gln substitution in the light chain of cytochrome b558 of the human NADPH oxidase (p22-phox) leads to defective translocation of the cytosolic proteins p47-phox and p67-phox. J Exp Med 180(6): 2329–2334. https://doi.org/10.1084/jem.180.6.2329 17. Park JW, Benna JE, Scott KE et al (1994) Isolation of a complex of respiratory burst oxidase components from resting neutrophil cytosol. Biochemistry 33(10):2907–2911. https://doi.org/10.1021/ bi00176a021 18. Heyworth PG, Curnutte JT, Nauseef WM et al (1991) Neutrophil nicotinamide adenine dinucleotide phosphate oxidase assembly. Translocation of p47-phox and p67-phox requires interaction between p47-phox and cytochrome b558. J Clin Invest 87:352– 356. https://doi.org/10.1172/JCI114993 19. Dang PMC, Cross AR, Babior BM (2001) Assembly of the neutrophil respiratory burst oxidase: a direct interaction between p67phox and cytochrome b558. Proc Natl Acad Sci U S A 98:3001–3005. https://doi.org/10.1073/pnas.061029698 20. Han CH, Freeman JL, Lee T et al (1998) Regulation of the neutrophil respiratory burst oxidase. Identification of an activation domain in p67 (phox). J Biol Chem 273(27):16663–16668. https://doi.org/10.1074/jbc.273.27.16663 21. Abo A, Pick E, Hall A et al (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1. Nature 353(6345):668–670. https://doi.org/10.1038/353668a0 22. Knaus UG, Heyworth PG, Evans T et al (1991) Regulation of phagocyte oxygen radical production by the GTP-binding protein

257

Rac 2. Science 254(5037):1512–1515. https://doi.org/10.1126/ science.1660188 23. Tsunawaki S, Kagara S, Yoshikawa K et al (1996) Involvement of p40phox in activation of phagocyte NADPH oxidase through association of its carboxyl-terminal, but not its amino-terminal, with p67phox. J Exp Med 184(3):893–902. https://doi.org/10. 1084/jem.184.3.893 24. Kuribayashi F, Nunoi H, Wakamatsu K et al (2002) The adaptor protein p40(phox) as a positive regulator of the superoxideproducing phagocyte oxidase. EMBO J 21(23):6312–6320. https://doi.org/10.1093/emboj/cdf642 25. Roos D, Kuhns DB, Maddalena A et al (2010) Hematologically important mutations: X-linked chronic granulomatous disease (third update). Blood Cells Mol Dis 45(3):246–265. https://doi. org/10.1016/j.bcmd.2010.01.009 26. Marciano BE, Spalding C, Fitzgerald A et al (2015) Common severe infections in chronic granulomatous disease. Clin Infect Dis 60:1176–1183. https://doi.org/10.1093/cid/ciu1154 27. Rosen GM, Freeman BA (1984) Detection of superoxide generated by endothelial cells. Proc Natl Acad Sci U S A 81(23):7269–7273. https://doi.org/10.1073/pnas.81.23.7269 28. Meier B, Radeke HH, Selle S et al (1989) Human fibroblasts release reactive oxygen species in response to interleukin-1 or tumour necrosis factor-alpha. Biochem J 263(2):539–545. https:// doi.org/10.1042/bj2630539 29. Griendling KK, Minieri CA, Ollerenshaw JD et al (1994) Angiotensin II stimulates NADH and NADPH oxidase activity in cultured vascular smooth muscle cells. Circ Res 74(6): 1141–1148. https://doi.org/10.1161/01.res.74.6.1141 30. Suh YA, Arnold RS, Lassegue B et al (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401(6748): 79–82. https://doi.org/10.1038/43459 31. Cheng G, Cao Z, Xu X et al (2001) Homologs of gp91phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269(1–2):131–140. https://doi.org/10.1016/s0378-1119(01) 00449-8 32. Geiszt M, Kopp JB, Várnai P et al (2000) Identification of renox, an NAD(P)H oxidase in kidney. Proc Natl Acad Sci U S A 97(14): 8010–8014. https://doi.org/10.1073/pnas.130135897 33. Shiose A, Kuroda J, Tsuruya K et al (2001) A novel superoxideproducing NAD(P)H oxidase in kidney. J Biol Chem 276(2): 1417–1423. https://doi.org/10.1074/jbc.M007597200 34. Yang S, Madyastha P, Bingel S et al (2001) A new superoxidegenerating oxidase in murine osteoclasts. J Biol Chem 276(8): 5452–5458. https://doi.org/10.1074/jbc.M001004200 35. Bánfi B, Molnár G, Maturana A et al (2001) A Ca(2+)-activated NADPH oxidase in testis, spleen, and lymph nodes. J Biol Chem 276(40):37594–37601. https://doi.org/10.1074/jbc.M103034200 36. Dupuy C, Ohayon R, Valent A et al (1999) Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. Cloning of the porcine and human cdnas. J Biol Chem 274(52):37265–37269. https://doi.org/10.1074/jbc.274.52.37265 37. De Deken X, Wang D, Many MC et al (2000) Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275(30):23227–23233. https://doi. org/10.1074/jbc.M000916200 38. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87(1):245–313. https://doi.org/10.1152/physrev.00044.2005 39. Dang PM, Rolas L, El-Benna J (2020) The dual role of reactive oxygen species-generating nicotinamide adenine dinucleotide phosphate oxidases in gastrointestinal inflammation and therapeutic perspectives. Antioxid Redox Signal 33(5):354–373. https://doi. org/10.1089/ars.2020.8018 40. Bánfi B, Clark RA, Steger K et al (2003) Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J

258 Biol Chem 278(6):3510–3513. https://doi.org/10.1074/jbc. C200613200 41. Geiszt M, Lekstrom K, Witta J et al (2003) Proteins homologous to p47phox and p67phox support superoxide production by NAD (P)H oxidase 1 in colon epithelial cells. J Biol Chem 278(22): 20006–20012. https://doi.org/10.1074/jbc.M301289200 42. Takeya R, Ueno N, Kami K et al (2003) Novel human homologues of p47phox and p67phox participate in activation of superoxideproducing NADPH oxidases. J Biol Chem 278(27):25234–25246. https://doi.org/10.1074/jbc.M212856200 43. Bromberg Y, Pick E (1984) Unsaturated fatty acids stimulate NADPH-dependent superoxide production by cell-free system derived from macrophages. Cell Immunol 88(1):213–221. https:// doi.org/10.1016/0008-8749(84)90066-2 44. Heyneman RA, Vercauteren RE (1984) Activation of a NADPH oxidase from horse polymorphonuclear leukocytes in a cell-free system. J Leukoc Biol 36(6):751–7569. https://doi.org/10.1002/jlb. 36.6.751 45. Bromberg Y, Pick E (1985) Activation of NADPH-dependent superoxide production in a cell-free system by sodium dodecyl sulfate. J Biol Chem 260(25):13539–13545 46. Curnutte JT (1985) Activation of human neutrophil nicotinamide adenine dinucleotide phosphate, reduced (triphosphopyridine nucleotide, reduced) oxidase by arachidonic acid in a cell-free system. J Clin Invest 75(5):1740–1743. https://doi.org/10.1172/ JCI111885 47. McPhail LC, Shirley PS, Clayton CC et al (1988) Activation of the respiratory burst enzyme from human neutrophils in a cell-free system. Evidence for a soluble cofactor. J Clin Invest May 75(5): 1735–1739. https://doi.org/10.1172/JCI111884 48. Volpp BD, Nauseef WM, Clark RA (1988) Two cytosolic neutrophil oxidase components absent in autosomal chronic granulomatous disease. Science 242:1295–1297. https://doi.org/10.1126/ science.2848318 49. Pick E, Kroizman T, Abo A (1989) Activation of the superoxideforming NADPH oxidase of macrophages requires two cytosolic components--one of them is also present in certain nonphagocytic cells. J Immunol 143(12):4180–4107 50. Segal AW, Heyworth PG, Cockcroft S et al (1985) Stimulated neutrophils from patients with autosomal recessive chronic granulomatous disease fail to phosphorylate a Mr-44 000 protein. Nature 316:547–549. https://doi.org/10.1038/316547a0 51. Lomax KJ, Leto TL, Nunoi H et al (1989) Recombinant 47-kilodalton cytosol factor restores NADPH oxidase in chronic granulomatous disease. Science 245:409–412. Erratum in: Science 1989, 246 (4933):987. https://doi.org/10.1126/science.2547247 52. Volpp BD, Nauseef WM, Donelson JE et al (1989) Cloning of the cDNA and functional expression of the 47-kilodalton cytosolic component of human neutrophil respiratory burst oxidase. Proc Natl Acad Sci U S A 86:7195–7199. Erratum in: Proc Natl Acad Sci U S A 86, 9563. https://doi.org/10.1073/pnas.86.18.7195 53. Leto TL, Garrett MC, Fujii H et al (1991) Characterization of neutrophil NADPH oxidase factors p47-phox and p67-phox from recombinant baculoviruses. J Biol Chem 266:19812–11988 54. Jouan A, Pilloud-Dagher MC, Fuchs A et al (1993) A generally applicable ELISA for the detection and quantitation of the cytosolic factors of NADPH-oxidase activation in neutrophils. Anal Biochem 214:252–259. https://doi.org/10.1006/abio.1993.1485 55. Okamura N, Curnutte JT, Roberts RL et al (1988) Relationship of protein phosphorylation to the activation of the respiratory burst in human neutrophils. Defects in the phosphorylation of a group of closely related 48-kDa proteins in two forms of chronic granulomatous disease. J Biol Chem 263(14):6777–67782 56. Rotrosen D, Leto TL (1990) Phosphorylation of neutrophil 47-kDa cytosolic oxidase factor. Translocation to membrane is associated

P. M.-C. Dang and J. El-Benna with distinct phosphorylation events. J Biol Chem 265:19910– 19925 57. El-Benna J, Dang PM, Gougerot-Pocidalo MA et al (2009) p47phox, the phagocyte NADPH oxidase/NOX2 organizer: structure, phosphorylation and implication in diseases. Exp Mol Med 41:217–225. https://doi.org/10.3858/emm.2009.41.4.058 58. Badwey JA, Heyworth PG, Karnovsky ML (1989) Phosphorylation of both 47 and 49kDa proteins accompanies superoxide release by neutrophils. Biochem Biophys Res Commun 158(3): 1029–1035. https://doi.org/10.1016/0006-291x(89)92825-8 59. Heyworth PG, Badwey JA (1990) Continuous phosphorylation of both the 47 and the 49 kDa proteins occurs during superoxide production by neutrophils. Biochim Biophys Acta 1052(2): 299–305. https://doi.org/10.1016/0167-4889(90)90225-3 60. Majumdar S, Kane LH, Rossi MW (1993) Protein kinase C isotypes and signal-transduction in human neutrophils: selective substrate specificity of calcium-dependent beta-PKC and novel calcium-independent nPKC. Biochim Biophys Acta 1176(3): 276–286. https://doi.org/10.1016/0167-4889(93)90056-u 61. Kramer IM, van der Bend RL, Verhoeven AJ et al (1988) The 47-kDa protein involved in the NADPH:O2 oxidoreductase activity of human neutrophils is phosphorylated by cyclic AMP-dependent protein kinase without induction of a respiratory burst. Biochim Biophys Acta 971:189–196. https://doi.org/10. 1016/0167-4889(88)90191-7 62. Dang PM, Fontayne A, Hakim J et al (2001) Protein kinase C zeta phosphorylates a subset of selective sites of the NADPH oxidase component p47phox, and participates in formyl peptide-mediated neutrophil respiratory burst. J Immunol 166:1206–1213. https:// doi.org/10.4049/jimmunol.166.2.1206 63. Fontayne A, Dang PM, Gougerot-Pocidalo MA et al (2002) Phosphorylation of p47phox sites by PKC alpha, beta II, delta, and zeta: effect on binding to p22phox and on NADPH oxidase activation. Biochemistry 41:7743–7750. https://doi.org/10.1021/bi011953s 64. El Benna J, Faust LP, Babior BM (1994) The phosphorylation of the respiratory burst oxidase component p47phox during neutrophil activation. Phosphorylation of sites recognized by protein kinase C and by proline-directed kinases. J Biol Chem 269: 23431–23436 65. Belambri SA, Rolas L, Raad H et al (2018) NADPH oxidase activation in neutrophils: role of the phosphorylation of its subunits. Eur J Clin Investig 48(Suppl 2):e12951. https://doi.org/ 10.1111/eci.12951 66. Wolfson M, McPhail LC, Nasrallah VN et al (1985) Phorbol myristate acetate mediates redistribution of protein kinase C in human neutrophils: potential role in the activation of the respiratory burst enzyme. J Immunol 135(3):2057–2062 67. Robinson JM, Heyworth PG, Badwey JA (1990) Utility of staurosporine in uncovering differences in the signal transduction pathways for superoxide production in neutrophils. Biochim Biophys Acta 1055(1):55–62. https://doi.org/10.1016/0167-4889 (90)90090-z 68. Faust LP, El Benna J, Babior BM et al (1995) The phosphorylation targets of p47 phox a subunit of the respiratory burst oxidase. Functions of the individual target serines as evaluated by sitedirected mutagenesis. J Clin Invest 96:1499–1505 69. Cheng N, He R, Tian J et al (2007) A critical role of protein kinase C delta activation loop phosphorylation in formyl-methionylleucyl-phenylalanine-induced phosphorylation of p47 (phox) and rapid activation of nicotinamide adenine dinucleotide phosphate oxidase. J Immunol 179(11):7720–7728. https://doi.org/10.4049/ jimmunol.179.11.7720 70. Belambri SA, Hurtado-Nedelec M, Senator A et al (2012) Phosphorylation of p47phox is required for receptor-mediated NADPH oxidase/NOX2 activation in Epstein-Barr virus-transformed human B lymphocytes. Am J Blood Res 2(3):187–193

15

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1)

71. Inanami O, Johnson JL, McAdara JK et al (1998) Activation of the leukocyte NADPH oxidase by phorbol ester requires the phosphorylation of p47PHOX on serine 303 or 304. J Biol Chem 273(16): 9539–9543. https://doi.org/10.1074/jbc.273.16.9539 72. Johnson JL, Park JW, Benna JE et al (1998) Activation of p47 (PHOX), a cytosolic subunit of the leukocyte NADPH oxidase. Phosphorylation of ser-359 or ser-370 precedes phosphorylation at other sites and is required for activity. J Biol Chem 273(52): 35147–35152. https://doi.org/10.1074/jbc.273.52.35147 73. Cox JA, Jeng AY, Sharkey NA et al (1985) Activation of the human neutrophil nicotinamide adenine dinucleotide phosphate (NADPH)-oxidase by protein kinase C. J Clin Invest 76(5): 1932–1938. https://doi.org/10.1172/JCI112190 74. McPhail LC, Qualliotine-Mann D, Waite KA (1995) Cell-free activation of neutrophil NADPH oxidase by a phosphatidic acidregulated protein kinase. Proc Natl Acad Sci U S A 92(17): 7931–7935. https://doi.org/10.1073/pnas.92.17.7931 75. El Benna J, Park JW, Ruedi JM et al (1995) Cell-free activation of the respiratory burst oxidase by protein kinase C. Blood Cells Mol Dis 21(3):201–206. https://doi.org/10.1006/bcmd.1995.0023 76. Park JW, Hoyal CR, Benna JE et al (1997) Kinase-dependent activation of the leukocyte NADPH oxidase in a cell-free system. Phosphorylation of membranes and p47(PHOX) during oxidase activation. J Biol Chem 272(17):11035–11043. https://doi.org/10. 1074/jbc.272.17.11035 77. Lopes LR, Hoyal CR, Knaus UG et al (1999) Activation of the leukocyte NADPH oxidase by protein kinase C in a partially recombinant cell-free system. J Biol Chem 274(22): 15533–15537. https://doi.org/10.1074/jbc.274.22.15533 78. Hoyal CR, Gutierrez A, Young BM et al (2003) Modulation of p47PHOX activity by site-specific phosphorylation: Akt-dependent activation of the NADPH oxidase. Proc Natl Acad Sci U S A 100(9):5130–5135. https://doi.org/10.1073/pnas. 1031526100 79. El-Benna J, Dang PM, Gougerot-Pocidalo MA (2008) Priming of the neutrophil NADPH oxidase activation: role of p47phox phosphorylation and NOX2 mobilization to the plasma membrane. Semin Immunopathol 30:279–289. https://doi.org/10.1007/ s00281-008-0118-3 80. Dang PM, Stensballe A, Boussetta T et al (2006) A specific p47phox serine phosphorylated by convergent MAPKs mediates neutrophil NADPH oxidase priming at inflammatory sites. J Clin Invest 116(7):2033–2043. https://doi.org/10.1172/JCI27544 81. Makni-Maalej K, Boussetta T, Hurtado-Nedelec M et al (2012) The TLR7/8 agonist CL097 primes N-formyl-methionyl-leucylphenylalanine-stimulated NADPH oxidase activation in human neutrophils: critical role of p47phox phosphorylation and the proline isomerase Pin 1. J Immunol 189:4657–4665. https://doi.org/ 10.4049/jimmunol.1201007 82. Liu M, Bedouhene S, Hurtado-Nedelec M et al (2019) The prolyl isomerase Pin1 controls lipopolysaccharide-induced priming of NADPH oxidase in human neutrophils. Front Immunol 10:2567. https://doi.org/10.3389/fimmu.2019.02567 83. DeLeo FR, Renee J, McCormick S et al (1998) Neutrophils exposed to bacterial lipopolysaccharide upregulate NADPH oxidase assembly. J Clin Invest 101(2):455–463. https://doi.org/10. 1172/JCI949 84. Ward RA, Nakamura M, McLeish KR (2000) Priming of the neutrophil respiratory burst involves p 38 mitogen-activated protein kinase-dependent exocytosis of flavocytochrome b558containing granules. J Biol Chem 275(47):36713–33679. https:// doi.org/10.1074/jbc.M003017200 85. Boussetta T, Gougerot-Pocidalo MA, Hayem G et al (2010) The prolyl isomerase Pin1 acts as a novel molecular switch for TNFalpha-induced priming of the NADPH oxidase in human

259

neutrophils. Blood 116(26):5795–5802. https://doi.org/10.1182/ blood-2010-03-273094 86. Liou YC, Zhou XZ, Lu KP (2011) Prolyl isomerase Pin1 as a molecular switch to determine the fate of phosphoproteins. Trends Biochem Sci 36(10):501–514. https://doi.org/10.1016/j.tibs.2011. 07.001 87. Belambri SA, Marzailoli V, Hurtado-Nedelec M, Pintard C, Liang S, Liu Y, Boussetta T, Gougerot-Pocidalo MA, Ye RD, Dang PM, El Benna J (2022) Impaired p47phox phosphorylation in neutrophils from p67phox-deficient chronic granulomatous disease patients. Blood 139(16):2512–2522. https://doi.org/10.1182/ blood.2021011134 88. Musacchio A, Gibson T, Lehto VP et al (1992) SH3--an abundant protein domain in search of a function. FEBS Lett 307(1):55–61. https://doi.org/10.1016/0014-5793(92)80901-r 89. Ren R, Mayer BJ, Cicchetti P et al (1993) Identification of a ten-amino acid proline-rich SH3 binding site. Science 259(5098): 1157–1161. https://doi.org/10.1126/science.8438166 90. Sumimoto H, Kage Y, Nunoi H et al (1994) Role of Src homology 3 domains in assembly and activation of the phagocyte NADPH oxidase. Proc Natl Acad Sci U S A 91:5345–5349. https://doi.org/ 10.1073/pnas.91.12.5345 91. Leto TL, Adams AG, de Mendez I (1994) Assembly of the phagocyte NADPH oxidase: binding of Src homology 3 domains to proline-rich targets. Proc Natl Acad Sci U S A 91:10650–10654. https://doi.org/10.1073/pnas.91.22.10650 92. Groemping Y, Lapouge K, Smerdon SJ et al (2003) Molecular basis of phosphorylation-induced activation of the NADPH oxidase. Cell 113:343–355. https://doi.org/10.1016/s0092-8674(03) 00314-3 93. Yuzawa S, Ogura K, Horiuchi M et al (2004) Solution structure of the tandem Src homology 3 domains of p47phox in an autoinhibited form. J Biol Chem 279:29752–29760. https://doi. org/10.1074/jbc.M401457200 94. Yuzawa S, Suzuki NN, Fujioka Y et al (2004) A molecular mechanism for autoinhibition of the tandem SH3 domains of p47phox, the regulatory subunit of the phagocyte NADPH oxidase. Genes Cells 9:443–456. Erratum in: Genes Cells 2004,9: 609. https://doi. org/10.1111/j.1356-9597.2004.00733.x 95. Durand D, Cannella D, Dubosclard V et al (2006) Small-angle X-ray scattering reveals an extended organization for the autoinhibitory resting state of the p47 (phox) modular protein. Biochemistry 45(23):7185–7193. https://doi.org/10.1021/ bi060274k 96. Wientjes FB, Panayotou G, Reeves E (1996) Interactions between cytosolic components of the NADPH oxidase: p40phox interacts with both p67phox and p47phox. Biochemist J317(Pt 3):919–924. https://doi.org/10.1042/bj3170919 97. Clark RA, Volpp BD, Leidal KG et al (1990) Two cytosolic components of the human neutrophil respiratory burst oxidase translocate to the plasma membrane during cell activation. J Clin Invest 85:714–721. https://doi.org/10.1172/JCI114496 98. El Benna J, Ruedi JM, Babior BM (1994) Cytosolic guanine nucleotide-binding protein Rac 2 operates in vivo as a component of the neutrophil respiratory burst oxidase. Transfer of Rac2 and the cytosolic oxidase components p47phox and p67phox to the submembranous actin cytoskeleton during oxidase activation. J Biol Chem 269(9):6729–6734 99. de Mendez I, Adams AG, Sokolic RA et al (1996) Multiple SH3 domain interactions regulate NADPH oxidase assembly in whole cells. EMBO J 15(6):1211–1220 100. DeLeo FR, Nauseef WM, Jesaitis AJ et al (1995) A domain of p47phox that interacts with human neutrophil flavocytochrome b558. J Biol Chem 270(44):26246–26251. https://doi.org/10. 1074/jbc.270.44.26246

260 101. DeLeo FR, Yu L, Burritt JB et al (1995) Mapping sites of interaction of p47-phox and flavocytochrome b with random-sequence peptide phage display libraries. Proc Natl Acad Sci U S A 92(15): 7110–7114. https://doi.org/10.1073/pnas.92.15.7110 102. Kanai F, Liu H, Field SJ (2001) The PX domains of p47phox and p40phox bind to lipid products of PI (3)K. Nat Cell Biol 3(7): 675–678. https://doi.org/10.1038/35083070 103. Zhan Y, Virbasius JV, Song X (2002) The p40phox and p47phox PX domains of NADPH oxidase target cell membranes via direct and indirect recruitment by phosphoinositides. J Biol Chem 277: 4512–4518. https://doi.org/10.1074/jbc.M109520200 104. Ago T, Kuribayashi F, Hiroaki H (2003) Phosphorylation of p47phox directs phox homology domain from SH3 domain toward phosphoinositides, leading to phagocyte NADPH oxidase activation. Proc Natl Acad Sci U S A 100:4474–4479. https://doi.org/10. 1073/pnas.0735712100 105. Karathanassis D, Stahelin RV, Bravo J (2002) Binding of the PX domain of p47 (phox) to phosphatidylinositol 3,4-bisphosphate and phosphatidic acid is masked by an intramolecular interaction. EMBO J21:5057–5068. https://doi.org/10.1093/emboj/cdf519 106. Marcoux J, Man P, Petit-Haertlein I et al (2010) p47phox molecular activation for assembly of the neutrophil NADPH oxidase complex. J Biol Chem 285(37):28980–28990. https://doi.org/10. 1074/jbc.M110.139824 107. Swain SD, Helgerson SL, Davis AR (1997) Analysis of activationinduced conformational changes in p47phox using tryptophan fluorescence spectroscopy. J Biol Chem 272:29502–29509. https://doi.org/10.1074/jbc.272.47.29502 108. Park JW, Babior BM (1997) Activation of the leukocyte NADPH oxidase subunit p47phox by protein kinase C. A phos- phorylationdependent change in the conformation of the C-terminal end of p47phox. Biochemistry 36:7474–7480. https://doi.org/10.1021/ bi9700936 109. Park HS, Park JW (1998) Fluorescent labeling of the leukocyte NADPH oxidase subunit p47(phox): evidence for amphiphileinduced conformational changes. Arch Biochem Biophys 360(2): 165–172. https://doi.org/10.1006/abbi.1998.0938 110. Marcoux J, Man P, Castellan M et al (2009) Conformational changes in p47(phox) upon activation highlighted by mass spectrometry coupled to hydrogen/deuterium exchange and limited proteolysis. FEBS Lett 583(4):835–840. https://doi.org/10.1016/j. febslet.2009.01.046 111. Shiose A, Sumimoto H (2000) Arachidonic acid and phosphorylation synergistically induce a conformational change of p47phox to activate the phagocyte NADPH oxidase. J Biol Chem 275:13793– 13801. https://doi.org/10.1074/jbc.275.18.13793 112. Ago T, Nunoi H, Ito T et al (1999) Mechanism for phosphorylation-induced activation of the phagocyte NADPH oxidase protein p47(phox). Triple replacement of serines 303, 304, and 328 with aspartates disrupts the SH3 domain-mediated intramolecular interaction in p47(phox), thereby activating the oxidase. J Biol Chem 274(47):33644–33653. https://doi.org/10.1074/jbc. 274.47.33644 113. Cheng G, Lambeth JD (2004) NOXO1, regulation of lipid binding, localization, and activation of Nox 1 by the Phox homology (PX) domain. J Biol Chem 279(6):4737–4742. https://doi.org/10. 1074/jbc.M305968200 114. Cheng G, Ritsick D, Lambeth JD (2004) Nox3 regulation by NOXO1, p47phox, and p67phox. J Biol Chem 279(33): 34250–34255. https://doi.org/10.1074/jbc.M400660200 115. Bánfi B, Malgrange B, Knisz J (2004) NOX3, a superoxidegenerating NADPH oxidase of the inner ear. J Biol Chem 279(44):46065–46072. https://doi.org/10.1074/jbc.M403046200 116. Cheng G, Lambeth JD (2005) Alternative mRNA splice forms of NOXO1: differential tissue expression and regulation of NOX1

P. M.-C. Dang and J. El-Benna and NOX3. Gene 356:118–126. https://doi.org/10.1016/j.gene. 2005.03.008 117. Sato TK, Overduin M, Emr SD (2001) Location, location, location: membrane targeting directed by PX domains. Science 294(5548): 1881–1885. https://doi.org/10.1126/science.1065763 118. Davis NY, McPhail LC, Horita DA (2012) The NOXO1β PX domain preferentially targets PtdIns (4,5)P2 and PtdIns (3,4,5)P3. J Mol Biol 417(5):440–453. https://doi.org/10.1016/j.jmb.2012. 01.058 119. Takeya R, Taura M, Yamasaki T (2006) Expression and function of Noxo1gamma, an alternative splicing form of the NADPH oxidase organizer 1. FEBS J 273(16):3663–3677. https://doi.org/ 10.1111/j.1742-4658.2006.05371.x 120. Ueyama T, Lekstrom K, Tsujibe S (2007) Subcellular localization and function of alternatively spliced Noxo1 isoforms. Free Radic Biol Med 42(2):180–190. https://doi.org/10.1016/j.freeradbiomed. 2006.08.024 121. Yamamoto A, Kami K, Takeya R (2007) Interaction between the SH3 domains and C-terminal proline-rich region in NADPH oxidase organizer 1 (Noxo1). Biochem Biophys Res Commun 352(2): 560–565. https://doi.org/10.1016/j.bbrc.2006.11.060 122. Dutta S, Rittinger K (2010) Regulation of NOXO1 activity through reversible interactions with p22 and NOXA1. PLoS One 5(5): e10478. https://doi.org/10.1371/journal.pone.0010478 123. Debbabi M, Kroviarski Y, Bournier O et al (2013) NOXO1 phosphorylation on serine 154 is critical for optimal NADPH oxidase 1 assembly and activation. FASEB J 27(4):1733–1748. https://doi. org/10.1096/fj.12-216432 124. Yamamoto A, Takeya R, Matsumoto M et al (2013) Phosphorylation of Noxo1 at threonine 341 regulates its interaction with Noxa1 and the superoxide-producing activity of Nox1. FEBS J280(20): 5145–5159. https://doi.org/10.1111/febs.12489 125. Kiss PJ, Knisz J, Zhang Y et al (2006) Inactivation of NADPH oxidase organizer 1 results in severe imbalance. Curr Biol l16(2): 208–213. https://doi.org/10.1016/j.cub.2005.12.025 126. Brandes RP, Harenkamp S, Schürmann C et al (2016) The cytosolic NADPH oxidase subunit NoxO1 promotes an endothelial stalk cell phenotype. Arterioscler Thromb Vasc Biol 36(8): 1558–1565. https://doi.org/10.1161/ATVBAHA.116.307132 127. Oshima H, Ishikawa T, Yoshida GJ et al (2014) TNF-α/TNFR1 signaling promotes gastric tumorigenesis through induction of Noxo1 and Gna14 in tumor cells. Oncogene 33(29):3820–3829. https://doi.org/10.1038/onc.2013.356 128. Juhasz A, Ge Y, Markel S et al (2009) Expression of NADPH oxidase homologues and accessory genes in human cancer cell lines, tumours and adjacent normal tissues. Free Radic Res 43(6): 523–532. https://doi.org/10.1080/10715760902918683 129. Kuwano Y, Kawahara T, Yamamoto H et al (2006) Interferongamma activates transcription of NADPH oxidase 1 gene and upregulates production of superoxide anion by human large intestinal epithelial cells. Am J Physiol Cell Physiol 290:C433–C443. https://doi.org/10.1152/ajpcell.00135.2005 130. Kuwano Y, Tominaga K, Kawahara T et al (2008) Tumor necrosis factor alpha activates transcription of the NADPH oxidase organizer 1 (NOXO1) gene and upregulates superoxide production in colon epithelial cells. Free Radic Biol Med 45:1642–1652. https:// doi.org/10.1016/j.freeradbiomed.2008.08.033 131. Makhezer N, Ben Khemis M, Liu D et al (2019) NOX1-derived ROS drive the expression of Lipocalin-2 in colonic epithelial cells in inflammatory conditions. Mucosal Immunol 12:117–131. https://doi.org/10.1038/s41385-018-0086-4 132. Elatrech I, Marzaioli V, Boukemara H et al (2015) Escherichia coli LF82 differentially regulates ROS production and mucin expression in intestinal epithelial T84 cells: implication of NOX1. Inflamm Bowel Dis 21:1018–1026. https://doi.org/10.1097/MIB. 0000000000000365

15

p47phox and NOXO1, the Organizer Subunits of the NADPH Oxidase 2 (Nox2) and NADPH Oxidase 1 (Nox1)

133. Kamizato M, Nishida K, Masuda K et al (2009) Interleukin 10 inhibits interferon gamma- and tumor necrosis factor alphastimulated activation of NADPH oxidase 1 in human colonic epithelial cells and the mouse colon. J Gastroenterol 44:1172– 1184. https://doi.org/10.1007/s00535-009-0119-6 134. Mouzaoui S, Djerdjouri B, Makhezer N et al (2014) Tumor necrosis factor-α-induced colitis increases NADPH oxidase 1 expression, oxidative stress, and neutrophil recruitment in the colon: preventive effect of apocynin. Mediat Inflamm 2014:312484. https://doi.org/ 10.1155/2014/312484

261

135. Nandi D, Tahiliani P, Kumar A et al (2006) The ubiquitinproteasome system. J Biosci 31:137–155. https://doi.org/10.1007/ BF02705243 136. Joo JH, Oh H, Kim M et al (2016) NADPH oxidase 1 activity and ROS generation are regulated by Grb2/Cbl-mediated proteasomal degradation of NoxO1 in colon cancer cells. Cancer Res 76(4): 855–865. https://doi.org/10.1158/0008-5472.CAN-15-1512 137. Haq S, Sarodaya N, Karapurkar JK et al (2022) CYLD destabilizes NoxO1 protein by promoting ubiquitination and regulates prostate cancer progression. Cancer Lett 525:146–157. https://doi.org/10. 1016/j.canlet.2021.10.032

The NADPH Oxidase Activator p67phox and Its Related Proteins

16

Hideki Sumimoto, Akira Kohda, Junya Hayase, and Sachiko Kamakura

Abstract

The Nox family NADPH oxidases can be divided into two groups based on the presence or absence of the Ca2+binding EF-hand motif in the N-terminal cytoplasmic region. Members of the former group can be activated via direct interaction with Ca2+, the cytosolic concentration of which is elevated upon cell stimulation. On the other hand, activation machineries for EF-hand motifdeficient oxidases, which have evolved in animals and fungi, consist of the small GTPase Rac and its binding partner such as p67phox and its homologous proteins. The EF-hand-free phagocyte oxidase Nox2, dormant in resting cells, becomes activated during phagocytosis to produce superoxide, a precursor of microbicidal oxidants, thereby playing a crucial role in host defense. In this process, the essential Nox2 activator p67phox translocates via collaboration with its constitutively-binding proteins p47phox and p40phox from the cytosol to the phagosomal membrane, and interacts there with independently-recruited, GTP-bound Rac to induce a conformational change of Nox2 for superoxide production. Similarly, the p67phoxrelated proteins NoxA1 and NoxR associate with Rac– GTP to activate the non-phagocytic oxidase Nox1 and the fungal oxidases NoxA and NoxB, respectively. Here we describe how p67phox as well as its homologues functions in activation of EF-hand-independent oxidases at the molecular level. Keywords

p67phox · NoxA1 · NoxR · Rac · TPR motif · Activation domain · SH3 domain · PB1 domain · p47phox · p40phox

H. Sumimoto (✉) · A. Kohda · J. Hayase · S. Kamakura Department of Biochemistry, Kyushu University Graduate School of Medical Sciences, Fukuoka, Japan e-mail: [email protected]

1

Introduction

The Nox family of superoxide-producing NADPH oxidases is widely present in eukaryotes (reviewed in [1–4]) and also found in prokaryotes [5–7]. The human genome contains seven oxidase genes, which encode Nox1 through Nox5 and the dual oxidases Duox1 and Duox2 (reviewed in [1– 4]). These oxidases possess two distinct (inner and outer) hemes in six transmembrane segments, followed by a ferredoxin–NADP+ reductase (FNR)-like dehydrogenase domain that harbors FAD- and NADPH-binding sites; as a result, they constitute the complete electron transporting apparatus from cytoplasmic NADPH via FAD and two hemes to extracellular O2 in the single polypeptide. Nox1, Nox2, Nox3, and Nox5 release superoxide as a primary product; on the other hand, Nox4, Duox1, and Duox2 mainly liberate hydrogen peroxide, which is formed from superoxide by effective dismutation at a site near the outer heme [8– 10]. In addition to the common oxidase structure, Nox5, Duox1, and Duox2 carry an N-terminal extension that contains Ca2+-binding EF-hand motifs. Nox5 contains four EF-hand motifs in the N-terminus, and the Duox oxidases feature an N-terminal peroxidase-like ectodomain, which precedes an additional transmembrane segment and two EF-hands in the cytoplasm. The EF-hand-deficient oxidases Nox1, Nox2, Nox3, and Nox4 form a stable heterodimer with the membrane-integrated protein p22phox (reviewed in [1–4]). Because reactive oxygen species (ROS) such as superoxide and hydrogen peroxide have not only deleterious effects but also beneficial roles, appropriate amounts of ROS must be produced at the right place by Nox-family oxidases. Nox4 is constitutively active, and thus Nox4-catalyzed ROS formation is primarily controlled by its expression at the protein level (reviewed in [1–4]) (See Chap. 12 by L. Hecker, K. Kato, and K. Griendling). Nox3 appears to be in a partially active state capable of generating superoxide to some extent, although the production is significantly promoted by activators and organizers also working with Nox1 and

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_16

263

264

Nox2 [11, 12] (See Chap. 11 by Y. Nakano and B. Bánfi). The catalytic activity of the other members is much more strictly regulated; they do not produce ROS under resting conditions. Nox5, Duox1, and Duox2 are inactive in resting cells but become activated via their own N-terminal EF hands, whose conformational change is induced by elevation of the cytoplasmic Ca2+ concentration in stimulated cells (See Chap. 14 by F. Miot and X. De Deken, and Chap. 13 by L. L. Camargo, F. Rios, A. Montezano, and R. Touyz). Fungi have three distinct subfamilies of NADPH oxidases: NoxC (Nox3) contains a single EF-hand in the N-terminal extension, although NoxA (Nox1) and NoxB (Nox2) lack the Ca2+binding motif (reviewed in [13, 14]) (See Chap. 25 by D. Takemoto and B. Scott). The plant NADPH oxidase homologues (Rboh) harbor two EF hands in the N-terminal region (See Chap. 26 by G. Miller and R. Mittler). The widespread presence of the EF-hand motif may suggest its identification as an original mechanism for Nox activation (reviewed in [15, 16]). These EF-hand-harboring oxidases also undergo stimulus-induced phosphorylation, which controls ROS production in animals and plants (reviewed in [17–19]). In the Opisthokonta, a major eukaryotic supergroup that includes animals and fungi, an EF-hand-independent machinery for Nox activation has evolved [3]. The machinery is constituted of the small GTPase Rac and its binding soluble partner that induces a conformational change of Nox to elicit electron transfer for superoxide production. The Rac partners include NoxA1 (Nox activator 1), p67phox (also known as NoxA2), and NoxR. p67phox is an essential activator of the mammalian phagocyte oxidase Nox2, which is dormant in resting cells but activated during phagocytosis to produce superoxide, a precursor of microbicidal oxidants (reviewed in [1–4, 19, 20]). In stimulated cells, p67phox translocates from the cytosol to the phagosomal membrane, enriched in the Nox2–p22phox dimer, and thus interacts with GTP-bound Rac to activate Nox2. NoxA1, encoded by a gene paralogous to that for p67phox (NoxA2), is required for Rac-dependent activation of Nox1, a p22phox-complexed oxidase in non-phagocytic cells (reviewed in [1–4]). The fungal p67phox-related protein NoxR functions together with RacA, a GTPase homologous to mammalian Rac, to activate NoxA and NoxB but not the EF-hand-containing oxidase NoxC (reviewed in [3, 14, 21]). On the other hand, plants do not have a clear homologue of p67phox, as expected from the presence of EF-hand motifs in all the Rboh oxidases. Here we describe the structure and function of these Rac-interacting oxidase activators as an essential component of the EF-hand-independent machinery for Nox activation.

H. Sumimoto et al.

2

p67phox (NoxA2)

The significance of p67phox in activation of the phagocyte oxidase Nox2 (a.k.a. gp91phox) is evident because genetic defects in p67phox, as well as those in gp91phox, p22phox, or p47phox, cause chronic granulomatous disease (CGD). Recurrent and life-threatening infections occur in CGD patients, whose neutrophils fail to generate superoxide as a precursor of microbicidal ROS during phagocytosis [22]. In contrast to the crucial role of ROS in host defence, as indicated in excessive inflammation characterized by granuloma formation in CGD, low production of ROS also appears to induce hyperinflammation and thus may drive autoimmune diseases. Indeed, mutations leading to a low function of p67phox, encoding by the gene NCF2 [23–25], as well as those of the NCF1-encoded protein p47phox [26, 27] are known to be associated with systemic lupus erythematosus (SLE), the prototypic systemic autoimmune disease, and with veryearly-onset inflammatory bowel disease [28–30]. The Nox2 activator p67phox forms a ternary complex with p47phox and p40phox with 1:1:1 stoichiometry in the cytosol of resting phagocytes [31, 32] (Fig. 16.1). On the other hand, GDP-bound Rac resides in the cytosol as a heterodimer with Rho GDP-dissociation inhibitor (RhoGDI). The p67phox– p47phox–p40phox complex translocates en bloc to the membrane during phagocytosis as well as in response to soluble stimulants such as the chemoattractant N-formyl-methionylleucyl-phenylalanine (fMLF), arachidonic acid (AA), and phorbol 12-myristate 13-acetate (PMA), a potent activator of protein kinase C (PKC). Rac is independently targeted to the membrane, following dissociation from RhoGDI and conversion into the GTP-bound state upon cell stimulation. At the membrane, p67phox interacts with Rac–GTP and elicits Nox2-catalyzed superoxide production. In contrast to fMLF and PMA, AA is also capable of activating the phagocyte oxidase in a cell-free system [33], which is reconstituted with the membrane-bound cytochrome b558, composed of Nox2 and p22phox, and the cytosolic proteins p67phox, p47phox, and Rac–GTP. In the presence of p67phox and Rac–GTP at much higher concentrations [34, 35] or a p67phox–Rac chimeric protein [36], p47phox is dispensable for cell-free activation of Nox2, indicating that Rac-bound p67phox, but not p47phox, directly induces a conformational change of Nox2. Human p67phox of 526 amino acids comprises the following four protein–protein interaction modules: an N-terminal Rac-binding domain composed of four tetratricopeptide repeat (TPR) motifs; a central SH3 domain, whose target has remained yet-to-be unidentified; a PB1 (Phox and Bem1) domain for stable association with p40phox; and a C-terminal SH3 domain that interacts with p47phox

16

The NADPH Oxidase Activator p67phox and Its Related Proteins

265

Fig. 16.1 The p67phox–p47phox–p40phox complex in phagocytes. The oxidase activator p67phox forms a ternary complex with p47phox and p40phox in the cytosol of resting phagocytes. The C-terminal SH3 domain of p67phox directly binds to the proline-rich region (PRR) of p47phox, whereas p67phox stably associates with p40phox via an interaction between PB1 domains. The N-terminal domain of p67phox comprises four tetratricopeptide repeat (TPR) motifs (T1–T4), followed by the activation domain (AD). The bis-SH3 domain of p47phox is

normally masked via an intramolecular interaction with the autoinhibitory region (AIR), but becomes accessible to p22phox upon cell stimulation that leads to phosphorylation of specific serine residues in the AIR disengaging the intramolecular bonds with the bis-SH3 domain. The PX domain of p40phox is capable of specifically binding to phosphatidylinositol 3-phosphate (PI3P), which is produced in the phagosomal membrane

(Fig. 16.2). p67phox adopts an elongated conformation with no or a few apparent intramolecular interactions between these four globular domains in a resting state [31, 32, 37, 38]. The flexible structure allows p67phox to simultaneously associate with p47phox and p40phox for formation of the stable ternary complex in the cytosol of resting cells [31, 32] (Fig. 16.1). The TPR domain in full-length p67phox is also in a state accessible to Rac, as indicated by the finding that

Rac–GTP binds to the full-length protein to the same extent as that to an isolated N-terminal domain [39, 40]. Although intramolecular interactions between the four globular domains (the TPR, N-terminal SH3, PB1, and C-terminal SH3 domain) is almost absent in a resting state, conformational changes within a domain and/or novel associations/ dissociations between a domain and a liker region would occur in p67phox during activation.

Fig. 16.2 p67phox-mediated activation of the phagocyte oxidase Nox2. In simulated phagocytes, p67phox translocates to the plasma membrane via p47phox, whose bis-SH3 domain and PX domain bind to the transmembrane protein p22phox and phosphoinositides (PIPs), respectively. Targeting of p67phox to the phagosomal membrane requires specific interaction of the p40phox PX domain with phosphatidylinositol 3-phosphate (PI3P), a phagosome-specific phospholipid. Independently of the p67phox–p47phox–p40phox complex, Rac is recruited to the membrane and converted into the GTP-bound state. At the membrane, p67phox interacts with Rac–GTP via the N-terminal TPR domain (amino acids 1–186), which allows the activation domain (AD) to mediate

Nox-catalyzed superoxide production. Solid two-headed arrows show individual associations that participate in membrane translocation of p67phox, whereas a dotted two-headed arrow denotes an interaction that is likely involved in enzymatic activation of Nox2. In the AD of human p67phox (amino acids 190–210), amino acid residues that are completely conserved throughout evolution are shown in red, and the invariant tyrosine residue in p67phox, which is substituted with phenylalanine in NoxA1, is shown in blue. Val204, which is not completely conserved but plays an important role in Nox2 activation, is shown in green

266

The N-terminal region of p67phox (amino acids 1–186) specifically binds to Rac–GTP albeit with a low affinity (Kd = 2–3 μM) [40]. Human genome contains three Rac-encoding genes: Rac1 is ubiquitously expressed, Rac2 expression is mostly restricted to cells of the hematopoietic lineage such as neutrophils, and Rac3 is an isoform abundant in the developing brain. Nox2 can be activated by any of the three Rac GTPases [41], although their closely-related GTPase Cdc42 neither binds to p67phox nor activates oxidases [41, 42]. Rac2 is responsible to Nox2 activation in human neutrophils and also involved in a number of leukocyte functions [43]. Accordingly, mutations in the Rac2 gene not only manifest as CGD but also induce other neutrophil function defects [22]. The Rac-binding site in p67phox is formed by the β-hairpin-insertion-containing region (amino acids 100–119) between the third and fourth TPR motifs (TPR3 and TPR4), together with the loops that connect TPR1 with TPR2 and TPR2 with TPR3 [40] (Fig. 16.3). Rac–GTP makes direct contacts with the side chains of the four conserved residues in p67phox: Arg102 and Asn104 between TPR3 and the β-hairpin insertion; Ser37 in the TPR1–TPR2 loop; and Asp67 in the TPR2–TPR3 loop [40]. Among them, Arg102 plays a key role in complex formation; indeed, glutamate substitution for Arg102 results in a complete loss of both interaction with Rac and activation of Nox2 [39]. Although the TPR domain of p67phox (amino acids 1–186) is essential and sufficient to interact with Rac–GTP, the domain by itself is incapable of activating Nox2 [44, 45]. Nox activation also requires a region C-terminal to the Rac-binding domain: p67phox-(1–212) and p67phox(1–210) are fully active in cell-free activation of Nox2, but p67phox-(1–199) and p67phox-(1–198) fail to activate Nox2 [44, 45]. Furthermore, alanine-scanning mutagenesis of amino acids 201–210 in p67phox-(1–210) revealed that alanine substitution for Val204 results in an almost complete loss of Nox2 activation [44]. Based on these findings, the region spanning amino acids 199–210 is designated as an “activation domain” of p67phox, although the effect of alanine substitution for Leu199 and Gly200 was not tested at that time [44]. On the other hand, the fragment N-terminal to the “activation domain” (amino acids 190–198) is also evolutionarily well conserved in p67phox and its related proteins [3]. A later study using alanine-scanning mutagenesis of amino acids 190–205 in full-length p67phox clarified the crucial roles of Val190, Leu193, Asp197, Tyr198, Leu199, and Gly200 in Nox2 activation in the whole-cell system, in addition to the significance of Val204 [46]. Unexpectedly and intriguingly, alanine substitution for Gln192 results in a three-to-four-fold enhancement in Nox2-catalyzed superoxide production, and Nox2 activation is completely abrogated by replacement of Tyr198 with Ala but rather facilitated by that with Phe or Trp [46]. Thus, the region of amino acids

H. Sumimoto et al.

190–210 appears to play an important role and thus is referred hereafter as the expanded activation domain (expanded AD) (Fig. 16.2), whose consensus sequence is ØXXLXXXD(Y/F)LGKXXX(V/I/L)(V/I)(A/S)(S/A)ØX(D/P) (where Ø indicates a residue with an aliphatic side chain, and X denotes any residue; completely conserved residues are bolded). In this activation domain, Leu193, Asp197, Leu199, and Gly200, and Lys201 are completely conserved in p67phox and its related proteins; Tyr198 is an invariant residue in p67phox but replaced by Phe in the p67phoxparalogous protein NoxA1, which is useful for discrimination between p67phox and NoxA1. The expanded activation domain (amino acids 190–210) appear to participate not only in superoxide production [46] but also in direct interaction of Rac-bound p67phox with Nox2 [47]. However, it remains unclear how p67phox, complexed with Rac–GTP, drives Nox2-catalyzed superoxide production. Rac binding may bring p67phox into a correct position to interact productively with Nox2 and/or induce a conformational change in p67phox to drive electron transport for superoxide production in Nox2. The region C-terminal to the TPR domain (amino acids 187–193) forms a short α-helix in the absence of Rac [48] but appears to become disordered upon binding of Rac–GTP [40] (Fig. 16.3). Of note, this short α-helix includes a part of the expanded activation domain consisting of amino acids 190–210. It is thus tempting to postulate that Rac binding to the TPR domain of p67phox increases flexibility in the expanded activation domain, which may be required for productive interaction with Nox2. In addition, it seems likely that a region other than the expanded activation domain also plays an important role in Nox2 activation. It has recently been shown that Rac binding disengages the β-hairpin from the C-terminal region (amino acids 181–186) in the TPR domain, which unmasks a previously hidden Nox2-binding site in p67phox [49, 50]. On the other hand, there is no evidence for direct interaction of Rac with Nox oxidases; for example, the insert α-helix of Rac (amino acids 124–135), a region unique to the Rho family among the Ras-related superfamily of small GTPases (Fig. 16.3), is dispensable to activation of Nox1, Nox2, and Nox3 [41, 51]. In contrast to the direct role of p67phox as activator of Nox2, p47phox and p40phox in the cytosolic phox ternary complex function as carriers for p67phox to the proximity of Nox2 in the membrane. In general, the CGD subtype, with mutations in the p47phox gene (NCF1), proceeds clinically more benignly than the p67phox gene (NCF2)-mutated CGD [22]. The oxidase organizer p47phox directly binds upon cell activation to p22phox, thereby playing a central role in the interaction of the p67phox–p47phox–p40phox complex with the Nox2–p22phox heterodimer on the membrane. Human p47phox is a 390-amino-acid protein that has several interaction modules: an N-terminal PX domain, which supports

16

The NADPH Oxidase Activator p67phox and Its Related Proteins

267

Fig. 16.3 Structure of the p67phox TPR domain in complex with Rac– GTP. The N-terminal domain of p67phox (amino acids 1–186) contains four tetratricopeptide repeat (TPR) motifs. Each repeat folds into two antiparallel α-helices that pack against one another: amino acids 3–30 in TPR1; amino acids 37–66 in TPR2; amino acids 71–99 in TPR3; and amino acids 120–149 in TPR4. In addition, the capping α-helix (amino acids 154–167) packs against the preceding helix of TPR4. The nine helical bundle exhibits a right-handed superhelical twist that produces a groove, which is occupied in an intramolecular interaction with amino acids 169–186. The intervening region between TPR3 and TPR4 contains a β-hairpin insertion (amino acids 106–119) that forms two

short β-strands and a 310 helical turn. The region C-terminal to the TPR domain (amino acids 187–193) forms a short α-helix in the absence of Rac but becomes disordered in the presence of Rac–GTP. It should be noted that the short α-helix includes a part of the activation domain (AD) of amino acids 190–210. The structure of the p67phox TPR domain in complex with GTP-bound Rac1, is drawn using the PyMOL software (http://www.pymol.org) and the pdb coordinates 1E96 [40]. The β-hairpin insertion in p67phox is highlighted in blue; and the insert helix, the switch I region, and GTP in Rac1 are highlighted in orange, red, and magenta, respectively

membrane recruitment via its ability to interact with phosphoinositides (this PX domain has a relatively weak phosphoinositide-binding activity with a poor specificity, although it prefers phosphatidylinositol 3,4-bisphosphate); two SH3 domains in tandem (bis-SH3 domain) for direct binding to p22phox; an autoinhibitory region (AIR), which prevents the bis-SH3 domain from interacting with p22phox in

resting cells; and a proline-rich region (PRR) for association with p67phox (Fig. 16.1). The C-terminal SH3 domain of p67phox stably associates with the C-terminus of p47phox with a high affinity by binding not only to the poly-proline II (PPII) helix of the PRR (amino acids 362–369) but also to its C-terminally flanking region that comprises two α helices (amino acids 372–386) [52] (PPII helix exhibits a PXXP

268

motif and an extended left-handed helical structure). Upon cell activation, p47phox undergoes phosphorylation at Ser303, Ser304, and Ser328 in the AIR, which allows the bis-SH3 domain to directly interact with p22phox, leading to p47phoxmediated localization of p67phox in the vicinity of Nox2. The p47phox–p67phox association constitutively occurs in resting neutrophils, but it seems possible that p47phox partially dissociates from p67phox after cell stimulation: Ser-379 in the flanking region becomes phosphorylated in stimulated cells, which likely attenuates p47phox binding to p67phox [53, 54] and thus negatively regulates phagocyte oxidase activation [53]. In contrast to the C-terminal SH3 domain, the central SH3 domain is dispensable for the superoxideproducing activity of Nox2 [37, 50, 55], although it is the most evolutionarily-conserved region in p67phox [3]. p67phox is fully recruited to the plasma membrane via the function of p47phox, but its targeting to the phagosomal membrane is promoted by p40phox. Inherited p40phox deficiency resembles a mild, atypical CGD; in vitro, phagocytosisinduced oxidase activity is severely impaired in neutrophils obtained from these patients, whereas fMLF- or PMAelicited superoxide production is not affected in these cells [56, 57]. Human p40phox of 340 amino acids associates with p67phox via stable interaction between their PB1 domains [58]. The PB1 domains comprise about 80 amino acids and are grouped into three types: type I, type II, and type I/II (reviewed in [59]). p40phox and p67phox harbor type I and II PB1 domains, respectively. Heterodimeric assembly occurs between type I and II PB1 domains via specific electrostatic interactions between a conserved acidic DX(D/E)GD region of the OPCA motif of a type I PB1 domain (Asp289-Asp293 in human p40phox) and an invariant lysine residue from a type II PB1 domain (Lys355 in human p67phox). The OPCA motif stands for a sequence variously named OPR (octicosapeptide repeat), PC (Phox and Cdc24), and AID (atypical protein kinase C (PKC)-interaction domain) [59, 60]. In addition to the C-terminal PB1 domain, p40phox harbors a N-terminal PX domain, which specifically and strongly binds to phosphatidylinositol 3-phosphate, a phospholipid enriched in the phagosomal membrane [61, 62]. During phagocytosis by neutrophils, p67phox is recruited via the PX domain of p40phox to the phagosomal membrane, where the Nox2– p22phox complex is delivered from the specific granules. The PB1 domain of p67phox is also able to interact with Vav, a guanine nucleotide exchange factor for Rac, which interaction likely facilitates Nox2-catalyzed superoxide production via activation of Rac [24].

H. Sumimoto et al.

3

NoxA1, a Vertebrate p67phox-Homologous Oxidase Activator

The p67phox-related oxidase activator NoxA1 plays an essential role in activation of the non-phagocytic oxidase Nox1, which is abundant in colon epithelial cells and also present in various cells including vascular smooth muscle cells [63– 65]. NoxA1 appears to participate as Nox1 activator in atherosclerotic plaque formation by driving proliferation and migration of vascular smooth muscle cells and by promoting the transition of these cells to a proinflammatory, macrophage-like phenotype [66]. Both Nox1 and NoxA1 are found from fishes to mammals but not in the urochordate or the cephalochordate, suggesting that they simultaneously emerged in the vertebrates and that NoxA1 has evolved for regulation of Nox1 activity [3]. Indeed, NoxA1 activates Nox1 more efficiently than does p67phox, whereas NoxA1 is a poor activator of Nox2 compared with p67phox [65]. NoxA1, as well as p67phox, is also able to enhance superoxide production by Nox3 [11, 12], an oxidase that is required for otoconia formation in the murine inner ear [67, 68] and also involved in sensorineural hearing loss [68]. On the other hand, the constitutively active oxidase Nox4 is not regulated by NoxA1 or p67phox. Human NoxA1 of 476 amino acids is a soluble protein showing a domain architecture comprising an N-terminal Rac-binding TPR domain, an activation domain, and a C-terminal SH3 domain (Fig. 16.4). Unlike p67phox, however, human NoxA1 lacks a central SH3 domain and a functional PB1 domain (Fig. 16.4). The latter module may be unnecessary for the non-phagocytic oxidase activator NoxA1, since the PB1 domain of p67phox is essential for stable association with p40phox, an adaptor protein that is responsible to phagosomal recruitment of p67phox. The absence of the central SH3 domain in NoxA1 also may be explicable at least partially by the finding that truncation of this SH3 domain in p67phox does not affect activation of Nox1 and Nox3 at all, although it decreases Nox2-dependent superoxide production [55]. NoxA1 activates Nox1 and also enhances Nox3-catalyzed superoxide production in a manner dependent on TPR domain-mediated interaction with GTP-bound Rac: the invariant residue Arg103 in the β-hairpin insertion of the TPR domain, corresponding to Arg102 in p67phox, plays a crucial role in interaction with Rac and subsequent activation of Nox1 as well as Nox3 [12, 65]. In addition to Arg103, Asp67 in the TPR2–TPR3 loop of p67phox, a residue with a side chain that directly interacts with Rac, is conserved in NoxA1 as Asp68, although Ser between TPR1 and TPR2 is replaced by Ala38 in NoxA1; Asn104 in the β-hairpin insertion of p67phox is conserved in mouse NoxA1 but not in

16

The NADPH Oxidase Activator p67phox and Its Related Proteins

Fig. 16.4 NoxA1-mediated activation of the non-phagocytic oxidase Nox1. The non-phagocytic oxidase activator NoxA1 localizes to the plasma membrane via association via its C-terminal SH3 domain with the oxidase organizer NoxO1, which simultaneously binds to the transmembrane protein p22phox, complexed with Nox1, and membrane phosphoinositides (PIPs). At the plasma membrane, active, GTP-bound Rac interacts with the N-terminal domain of NoxA1 (amino acids 1–187), containing for four TPRs (T1–T4), and thus the activation domain (AD; amino acids 191–211) induces a conformational change of Nox1, leading to superoxide production. Solid two-headed arrows show individual associations that participate in membrane translocation of NoxA1, whereas a dotted two-headed arrow denotes an interaction that is likely involved in enzymatic activation of Nox1. In the AD of human NoxA1, amino acid residues that are completely conserved throughout evolution are shown in red, and the invariant phenylalanine residue in NoxA1, which is substituted with tyrosine in p67phox, is shown in blue. Val205, which is not completely conserved but positively regulates Nox1 activation, is shown in green

human NoxA1. NoxA1 binds to Rac1 and Rac2 but not to Cdc42 [65], and Rac1, Rac2, and Rac3 are all able to activate Nox1 in a NoxA1-dependent manner [41]. It has been reported that NoxA1 undergoes phosphorylation at Ser-171 in the TPR domain, which is involved in negative regulation of Nox1 activation [69]. The core sequence LXXXD(Y/F)LGKXXXV in the activation domain is completely conserved in NoxA1. As the second residue in the core D(Y/F)LGK motif, Tyr is invariably present in p67phox (Tyr198 in human p67phox), but is replaced by Phe in all NoxA1 proteins from various species (Phe199 in human NoxA1) [3, 46]. Indeed, alanine scanning experiments have revealed that the conserved residues, including Phe199, participate in NoxA1-dependent activation of Nox1 [46]. In addition, it has been reported that the synthetic NoxA1-derived 11-mer peptide (amino acids 195–205), containing the core D(Y/F)LGK motif with the F199A substitution, effectively inhibits NoxA1-mediated Nox1 activation but not p67phox-dependent Nox2 activation [70], which further confirms the significance of the expanded activation domain defined here (amino acids 190–210 in

269

human p67phox). The distance (i.e., the number of amino acids) from the TPR domain to the activation domain in NoxA1 is identical to that in p67phox, supporting the idea that Rac binding to the TPR domain changes a conformation of the activation domain in a similar manner. Targeting of the soluble protein NoxA1 to the Nox1– p22phox complex on the membrane is dependent on its interaction with NoxO1, an organizer that directly binds via the bis-SH3 domain to p22phox (reviewed in [1–4]). Similar to the p67phox–p47phox interaction, the C-terminal SH3 domain of NoxA1 appears to simultaneously make contacts with the PRR [71] and its flanking region [72] in the C-terminus of NoxO1, which contributes to incorporation of NoxA1 into the Nox1-based oxidase complex via the p22phox-binding protein NoxO1. The tail-to-tail association is facilitated by AA and by PKC-catalyzed phosphorylation of Thr341 in the C-terminal region of NoxO1, leading to enhancement of Nox1 activation [71, 73]. The SH3 domain of NoxA1 is also capable of interacting with peroxiredoxin 6 (Prdx6), an atypical thiol peroxidase with a phospholipase A2 activity, which interaction appears to stabilize NoxA1 and to upregulate Nox1-catalyzed superoxide production [74]. Although Prdx6 is able to associate with p67phox [75, 76] and to enhance Nox2 activation [75, 77], it remains unknown which region of p67phox is involved in the association; furthermore, Prdx6 binding to p67phox per se rather inhibits its phospholipase A2 activity [76], which is required for Prdx6-mediated enhancement of Nox2 activation [77, 78]. Thus, the effect of Prdx6 on p67phox seems different from that on NoxA1.

4

NoxR, a Fungal p67phox-Related Protein that Regulates NoxA and NoxB

The Opisthokonta, a major supergroup in eukaryotes, includes animals and fungi but not plants: eukaryotes are presently considered to form the two domains Diaphoretickes and Amorphea; plants belong to subgroups in the Diaphoretickes, whereas the Amorphea is composed of the subgroups Amoebozoa and Obazoa, the latter of which includes the clade Opisthokonta [79]. NADPH oxidases in fungi vary in number from none to three genes: NoxA (Nox1), NoxB (Nox2), and NoxC (Nox3) (See Chap. 25 by D. Takemoto and B. Scott). Fungal NoxA and NoxB regulate a variety of functions, such as fruiting body differentiation, ascospore germination, sclerotia formation, and symbiotic or pathogenic interaction with plants [13, 14]. NoxA and NoxB are phylogenetically close to each other: NoxA has a compact oxidase structure similar to mammalian Nox2, but NoxB contains an additional short N-terminal extension. The distantly-related oxidase NoxC harbors an N-terminal

270

H. Sumimoto et al.

Fig. 16.5 NoxR-mediated activation of the fungal oxidase NoxA. The fungal oxidase activator NoxR plays a crucial role in activation of NoxA, which forms a complex with NoxD, a tetramembrane-spanning protein that is distantly related to mammalian p22phox. NoxR interacts via its C-terminal PB1 domain with the cell polarity protein Cdc24, which interaction is likely involved in plasma membrane localization. GTP-bound Rac interacts with the N-terminal domain of NoxR (amino acids 1–183), containing four TPRs (T1–T4), and thus the activation domain (AD; amino acids 187–207) induces a conformational change of

NoxA, leading to superoxide production. Solid two-headed arrows show individual associations that participate in membrane translocation of NoxR, whereas a dotted two-headed arrow denotes an interaction that is likely involved in enzymatic activation of NoxA. In the AD of NoxR in the fungal endophyte Epichloë festucae, amino acid residues that are completely conserved throughout evolution are shown in red, and the invariant tyrosine residue in NoxR and p67phox, which is substituted with phenylalanine in NoxA1, is shown in blue

putative Ca2+-binding EF-hand motif, which is absent in NoxA and NoxB. NoxC is less common, and its role remains largely unknown. Similar to mammalian p22phox-complexed oxidases (Nox1 through Nox4), which do not have EF-hand motifs, the fungal oxidase NoxA functions as a heterodimer with NoxD (also known as Pro41), a protein that has four transmembrane segments and shares a weak homology with p22phox [80, 81] (Fig. 16.5). On the other hand, NoxB may interact very weakly with another membrane-spanning protein Pls1 (PlsA), which is structurally unrelated to p22phox [82–84]. In contrast to p22phox, NoxD (as well as Pls1) lacks the PRR, which directly interacts with the bis-SH3 domain of p47phox and its paralogous-protein NoxO1 to play a crucial role in recruitment of p67phox and NoxA1 for mammalian Nox activation (reviewed in [1–4]). Consistent with the absence of the PRR, no gene for a p47phox-related protein has been found in fungal genomes. NoxA/B-containing fungi possesses a gene encoding a p67phox-related protein, designated NoxR (Nox regulator) [3, 14, 21, 85] (Fig. 16.5). NoxR is required for activation of NoxA and NoxB but not for that of the EF-handcontaining oxidase NoxC [14, 21]. RacA, encoded in a fungal homologue of the mammalian Rac genes, is also essential for superoxide production by NoxA and NoxB [14, 21] and functions via direct binding to the N-terminal domain of NoxR. This domain comprises four TPR motifs and a

β-hairpin insertion between the third and fourth motif and the p67phox residues directly involved in binding to Rac are well conserved [3, 14, 85]. Disruption of RacA binding to NoxR results in an impaired activation of NoxA and NoxB. The activation domain of NoxR is also evolutionarily well conserved especially in its core region LXXXD(Y/F) LGKXXXV, with the exception that Leu replaces the last residue Val, which corresponds to Val204 in human p67phox [46]. In addition, NoxR harbors Tyr in the invariant D(Y/F) LGK motif of the activation domain, as in p67phox but not in NoxA1 [46]. It should be noted that adaptor proteins that contain the Rac-binding TPR domain but lack the activation domain are distributed in the Opisthokonta and in the Amoebozoa [3, 14], suggesting that this type of the TPR domain is widely used not only in the Nox complex but also in other Rac-regulated systems. Intracellular localization of NoxR is regulated by regions other than the TPR and activation domains. NoxR lacks both SH3 domains present in p67phox but contains a PB1 domain in the C-terminus. This domain appears to belong to type I/II, because it contains not only a conserved Lys (present in type II PB1 member) but also an atypical OPCA motif (found in type I PB1 member). The OPCA motif in NoxR PB1 domain is an atypical type I/II; in NoxR of the fungal endophyte Epichloë festucae, Met replaces the first Asp in the canonical OPCA sequence DX(D/E)GDX8(E/D), where X is any amino acid [86]. Although type I PB1 domains interact with type II

16

The NADPH Oxidase Activator p67phox and Its Related Proteins

PB1 domains and vice versa, type I/II PB1 domains have a potential to bind to both type I and type II members via their invariant Lys and via their OPCA motif, respectively [59]. Indeed, the PB1 domain of NoxR is capable of interacting with the type I PB1 domain of Cdc24, a fungal protein crucial for cell polarization, which interaction appears to enable NoxR to localize to the correct site for activation of NoxA [86]. The Cdc24-binding protein Bem1, harboring two SH3, one PX, and one PB1 domains, also interacts with NoxR albeit in a PB1-independent manner [86]. In addition to the potential to associate with both type I and type II domain, type I/II PB1 domains are able to engage in homotypic interactions to form homo-oligomers [59, 87– 89]. For instance, the selective autophagy receptor p62 (a.k.a. SQSTM1) assembles via its type I/II PB1 domain to form filamentous polymers and thus plays a critical role in mitophagy and autophagic degradation of protein aggregates [90, 91]. Interestingly, it has been shown that the type I/II PB1 domain of NoxR is able to form a homodimer [86], but the role of the PB1-mediated self-assembly remains unknown at present. Thus, similar to p67phox, NoxR likely activates the fungal oxidases NoxA and NoxB via the RacA-binding TPR domain and the conserved activation domain in the N terminus. The C-terminal PB1 domain participates in correct localization of NoxR via its association with the polarity protein Cdc24 [86]. NoxR also appears to be regulated by phosphorylation [92]. The fungal oxidase Nox2 (NoxB) and the PKC-homologous protein Pkc1 are both required for a process of the entry of the blast fungus Magnaporthe oryzae to its host plant; the infection causes rice blast—the most widespread and serious disease of rice. Intriguingly, Pkc1 interacts with Nox2 and NoxR and directly phosphorylates NoxR, suggesting its significance in activation of the oxidase complex [92].

5

Epilogue

p67phox and its related proteins NoxA1 and NoxR play a crucial role in activation of NADPH oxidases without the EF-hand. The essential modules in these oxidase activators are the Rac-binding TPR domain and the activation domain. Interaction with GTP-bound Rac is considered to render the activation domain in a state accessible to Nox to drive electron transfer for superoxide production. The precise mechanism for the activator-driven transition to a productive conformation in Nox, however, remains presently unknown. The other modules in the activators participate in interactions with adaptor proteins for recruitment to the correct site in stimulated cells. In the case of p67phox, the C-terminal SH3 domain stably interacts with p47phox, which binds upon cell stimulation to p22phox to target p67phox to the vicinity of

271

Nox2; the PB1 domain associates with p40phox, which delivers p67phox to the phagosome by virtue of the exceptionally high affinity of the p40phox PX domain for phosphatidylinositol 3-phosphate (PI3P) formed on the phagosomal membrane. The association with p40phox stabilizes p67phox at the protein level in Epstein-Barr virus-transformed B cells, but not in neutrophils [57]; the reason for this discrepancy is presently unknown. The absence of the EF-hand solely indicates that Ca2+ does not interact directly with this type of NADPH oxidase. In activation of EF-hand-deficient oxidases, Ca2+ plays a major regulatory role in an indirect manner. For instance, it is well established that extracellular Ca2+ entry exerts a signal that ensures a full activation of Nox2 during phagocytosis or in fMLF-stimulated neutrophils [93]. However, it has long remained obscure how the Ca2+ signal is transduced into the regulation of Nox2. The transduction possibly involves Ca2+-dependent PKCs (conventional PKCs), because PKC is able to phosphorylate p47phox and thus induce the interaction of p47phox with p22phox, a crucial event that allows p67phox to activate Nox2, as described here. PKC may also regulate oxidase activity by phosphorylating NoxA1 and NoxR, as well as NoxO1, Nox1, Nox2, and p22phox [69, 71, 93, 94]. Interestingly, p67phox is required for optimal PKC-catalyzed phosphorylation of p47phox [95], although it remains to be elucidated how p67phox accelerates the reaction. The activator p67phox, thus, appears on occasions to serve also as an oxidase organizer. Further studies on the molecular mechanism whereby p67phox and its related proteins function may facilitate the development of novel therapies for Nox-associated diseases. Acknowledgements This work was supported in part by KAKENHI Grant (21H02698) and Grant-in-Aid for Transformative Research Areas [A] (21H05267) from JSPS (Japan Society for the Promotion of Science).

References 1. Lambeth JD (2004) NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol 4:181–189 2. Bedard K, Krause K-H (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313 3. Sumimoto H (2008) Structure, regulation and evolution of Nox-family NADPH oxidases that produce reactive oxygen species. FEBS J 275:3249–3277 4. Leto TL, Morand S, Hurt D, Ueyama T (2009) Targeting and regulation of reactive oxygen species generation by Nox family NADPH oxidases. Antioxid Redox Signal 11:2607–2619 5. Magnani F, Nenci S, Millana Fananas E, Ceccon M, Romero E, Fraaije MW, Mattevi A (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci U S A 114:6764–6769 6. Hajjar C, Cherrier MV, Mirandela GD, Petit-Hartlein I, Stasia MJ, Fontecilla-Camps JC, Fieschi F, Dupuy J (2017) The NOX family of proteins is also present in bacteria. MBio 8:e01487–e01417

272 7. Miles JA, Egan JL, Fowler JA, Machattou P, Millard AD, Perry CJ, Scanlan DJ, Taylor PC (2021) The evolutionary origins of peroxynitrite signalling. Biochem Biophys Res Commun 580:107–112 8. Takac I, Schröder K, Zhang L, Lardy B, Anilkumar N, Lambeth JD, Shah AM, Morel F, Brandes RP (2011) The E-loop is involved in hydrogen peroxide formation by the NADPH oxidase Nox4. J Biol Chem 286:13304–13313 9. Ueyama T, Sakuma M, Ninoyu Y, Hamada T, Dupuy C, Geiszt M, Leto TL, Saito N (2015) The extracellular A-loop of dual oxidases affects the specificity of reactive oxygen species release. J Biol Chem 290:6495–6506 10. Sun J (2020) Structure of mouse DUOX1-DUOXA1 provide mechanistic insights into enzyme activation and regulation. Nat Struct Mol Biol 27:1086–1093 11. Ueno N, Takeya R, Miyano K, Kikuchi H, Sumimoto H (2005) The NADPH oxidase Nox3 constitutively produces superoxide in a p22phox-dependent manner: its regulation by oxidase organizers and activators. J Biol Chem 280:23328–23339 12. Ueyama T, Geiszt M, Leto TL (2006) Involvement of Rac1 in activation of multicomponent Nox1- and Nox3-based NADPH oxidases. Mol Cell Biol 26:2160–2174 13. Aguirre J, Ríos-Momberg M, Hewitt D, Hansberg W (2005) Reactive oxygen species and development in microbial eukaryotes. Trends Microbiol 13:111–118 14. Takemoto D, Tanaka A, Scott B (2007) NADPH oxidases in fungi: diverse roles of reactive oxygen species in fungal cellular differentiation. Fungal Genet Biol 44:1065–1076 15. Torres MA, Dangl JL (2005) Functions of the respiratory burst oxidase in biotic interactions, abiotic stress and development. Curr Opin Plant Biol 8:397–403 16. Adachi H, Yoshioka H (2015) Kinase-mediated orchestration of NADPH oxidase in plant immunity. Brief Funct Genomics 14: 253–259 17. Fulton DJR (2019) The molecular regulation and functional roles of NOX5. In: Knaus UG, Leto TL (eds) NADPH oxidases: methods and protocols. Springer Science+Bussiness Media, New York, pp 353–375 18. Rigutto S, Hoste C, Grasberger H, Milenkovic M, Communi D, Dumont JE, Corvilain B, Miot F, De Deken X (2009) Activation of dual oxidases Duox1 and Duox2: differential regulation mediated by cAMP-dependent protein kinase and protein kinase C-dependent phosphorylation. J Biol Chem 284:6725–6734 19. Nauseef WM (2019) The phagocyte NOX2 NADPH oxidase in microbial killing and cell signaling. Curr Opin Immunol 60:130–140 20. Winterbourn CC, Kettle AJ, Hampton MB (2016) Reactive oxygen species and neutrophil function. Annu Rev Biochem 85:765–792 21. Choi J, Détry N, Kim K-T, Asiegbu FO, Valkonen JP, Lee Y-H (2014) fPoxDB: fungal peroxidase database for comparative genomics. BMC Microbiol 14:117 22. Roos D, van Leeuwen K, Hsu AP, Priel DL, Begtrup A, Brandon R, Rawat A, Vignesh P, Madkaikar M, Stasia MJ, Bakri FG, de Boer M, Roesler J, Köker N, Köker MY, Jakobsen M, Bustamante J, Garcia-Morato MB, Shephard JLV, Cagdas D, Tezcan I, Sherkat R, Mortaz E, Fayezi A, Shahrooei M, Wolach B, Blancas-Galicia L, Kanegane H, Kawai T, Condino-Neto A, Vihinen M, Zerbe CS, Holland SM, Malech HL, Gallin JI, Kuhns DB (2021) Hematologically important mutations: the autosomal forms of chronic granulomatous disease (third update). Blood Cells Mol Dis 92:102596 23. Cunninghame Graham DS, Morris DL, Bhangale TR, Criswell LA, Syvänen A-C, Rönnblom L, Behrens TW, Graham RR, Vyse TJ (2011) Association of NCF2, IKZF1, IRF8, IFIH1, and TYK2 with systemic lupus erythematosus. PLoS Genet 7:e1002341 24. Jacob CO, Eisenstein M, Dinauer MC, Ming W, Liu Q, John S, Quismorio FP Jr, Reiff A, Myones BL, Kaufman KM, McCurdy D,

H. Sumimoto et al. Harley JB, Silverman E, Kimberly RP, Vyse TJ, Gaffney PM, Moser KL, Klein-Gitelman M, Wagner-Weiner L, Langefeld CD, Armstrong DL, Zidovetzki R (2012) Lupus-associated causal mutation in neutrophil cytosolic factor 2 (NCF2) brings unique insights to the structure and function of NADPH oxidase. Proc Natl Acad Sci U S A 109:E59–E67 25. Kim-Howard X, Sun C, Molineros JE, Maiti AK, Chandru H, Adler A, Wiley GB, Kaufman KM, Kottyan L, Guthridge JM, Rasmussen A, Kelly J, Sánchez E, Raj P, Li Q-Z, Bang S-Y, Lee H-S, Kim T-H, Kang YM, Suh C-H, Chung WT, Park Y-B, Choe J-Y, Shim SC, Lee S-S, Han B-G, Olsen NJ, Karp DR, Moser K, Pons-Estel BA, Wakeland EK, James JA, Harley JB, Bae S-C, Gaffney PM, Alarcón-Riquelme M, GENLES, Looger LL, Nath SK (2014) Allelic heterogeneity in NCF2 associated with systemic lupus erythematosus (SLE) susceptibility across four ethnic populations. Hum Mol Genet 23:1656–1668 26. Zhao J, Ma J, Deng Y, Kelly JA, Kim K, Bang S-Y, Lee H-S, Li QZ, Wakeland EK, Qiu R, Liu M, Guo J, Li Z, Tan W, Rasmussen A, Lessard CJ, Sivils KL, Hahn BH, Grossman JM, Kamen DL, Gilkeson GS, Bae S-C, Gaffney PM, Shen N, Tsao BP (2017) A missense variant in NCF1 is associated with susceptibility to multiple autoimmune diseases. Nat Genet 49:433–437 27. Urbonaviciute V, Luo H, Sjöwall C, Bengtsson A, Holmdahl R (2019) Low production of reactive oxygen species drives systemic lupus erythematosus. Trends Mol Med 25:826–835 28. Denson LA, Jurickova I, Karns R, Shaw KA, Cutler DJ, Okou DT, Dodd A, Quinn K, Mondal K, Aronow BJ, Haberman Y, Linn A, Price A, Bezold R, Lake K, Jackson K, Walters TD, Griffiths A, Baldassano RN, Noe JD, Hyams JS, Crandall WV, Kirschner BS, Heyman MB, Snapper S, Guthery SL, Dubinsky MC, Leleiko NS, Otley AR, Xavier RJ, Stevens C, Daly MJ, Zwick ME, Kugathasan S (2018) Clinical and genomic correlates of neutrophil reactive oxygen species production in pediatric patients with Crohn’s disease. Gastroenterology 154:2097–2110 29. Dhillon SS, Fattouh R, Elkadri A, Xu W, Murchie R, Walters T, Guo C, Mack D, Huynh HQ, Baksh S, Silverberg MS, Griffiths AM, Snapper SB, Brumell JH, Muise AM (2014) Variants in nicotinamide adenine dinucleotide phosphate oxidase complex components determine susceptibility to very early onset inflammatory bowel disease. Gastroenterology 147:680–689 30. Muise AM, Xu W, Guo C-H, Walters TD, Wolters VM, Fattouh R, Lam GY, Hu P, Murchie R, Sherlock M, Gana JC, NEOPICS, Russell RK, Glogauer M, Duerr RH, Cho JH, Lees CW, Satsangi J, Wilson DC, Paterson AD, Griffiths AM, Silverberg MS, Brumell JH (2012) NADPH oxidase complex and IBD candidate gene studies: identification of a rare variant in NCF2 that results in reduced binding to RAC2. Gut 61:1028–1035 31. Lapouge K, Smith SJM, Groemping Y, Rittinger K (2002) Architecture of the p40-p47-p67phox complex in the resting state of the NADPH oxidase: a central role for p67phox. J Biol Chem 277:10121–10128 32. Ziegler CS, Bouchab L, Tramier M, Durand D, Fieschi F, DupréCrochet S, Mérola F, Nüße O, Erard M (2019) Quantitative live-cell imaging and 3D modeling reveal critical functional features in the cytosolic complex of phagocyte NADPH oxidase. J Biol Chem 294:3824–3836 33. Pick E (2020) Cell-free NADPH oxidase activation assays: a triumph of reductionism. In: Quinn MT, DeLeo FR (eds) Neutrophil: methods and protocols. Springer Science+Bussiness Media, New York, pp 325–411 34. Freeman JL, Lambeth JD (1996) NADPH oxidase activity is independent of p47phox in vitro. J Biol Chem 271:22578–22582 35. Koshkin V, Lotan O, Pick E (1996) The cytosolic component p47phox is not a sine qua non participant in the activation of NADPH oxidase but is required for optimal superoxide production. J Biol Chem 271:30326–30329

16

The NADPH Oxidase Activator p67phox and Its Related Proteins

36. Gorzalczany Y, Alloul N, Sigal N, Weinbaum C, Pick E (2002) A prenylated p67phox-Rac1 chimera elicits NADPH-dependent superoxide production by phagocyte membranes in the absence of an activator and of p47phox: conversion of a pagan NADPH oxidase to monotheism. J Biol Chem 277:18605–18610 37. Yuzawa S, Miyano K, Honbou K, Inagaki F, Sumimoto H (2009) The domain organization of p67phox, a protein required for activation of the superoxide-producing NADPH oxidase in phagocytes. J Innate Immun 1:543–555 38. Durand D, Vivès C, Cannella D, Pérez J, Pebay-Peyroula E, Vachette P, Fieschi F (2010) NADPH oxidase activator p67phox behaves in solution as a multidomain protein with semi-flexible linkers. J Struct Biol 169:45–53 39. Koga H, Terasawa H, Nunoi H, Takeshige K, Inagaki F, Sumimoto H (1999) Tetratricopeptide repeat (TPR) motifs of p67phox participate in interaction with the small GTPase Rac and activation of the phagocyte NADPH oxidase. J Biol Chem 274:25051–25060 40. Lapouge K, Smith SJ, Walker PA, Gamblin SJ, Smerdon SJ, Rittinger K (2000) Structure of the TPR domain of p67phox in complex with Rac
GTP. Mol Cell 6:899–907 41. Miyano K, Koga H, Minakami R, Sumimoto H (2009) The insert region of the Rac GTPases is dispensable for activation of superoxide-producing NADPH oxidases. Biochem J 422:373–382 42. Kwong CH, Adams AG, Leto TL (1995) Characterization of the effector-specifying domain of Rac involved in NADPH oxidase activation. J Biol Chem 270:19868–19872 43. Nauseef WM, Borregaard N (2014) Neutrophils at work. Nat Immunol 15:602–611 44. Han C-H, Freeman JL, Lee T, Motalebi SA, Lambeth JD (1998) Regulation of the neutrophil respiratory burst oxidase. Identification of an activation domain in p67phox. J Biol Chem 273:16663–16668 45. Hata K, Takeshige K, Sumimoto H (1997) Roles for proline-rich regions of p47phox and p67phox in the phagocyte NADPH oxidase activation in vitro. Biochem Biophys Res Commun 241:226–231 46. Maehara Y, Miyano K, Yuzawa S, Akimoto R, Takeya R, Sumimoto H (2010) A conserved region between the TPR and activation domains of p67phox participates in activation of the phagocyte NADPH oxidase. J Biol Chem 285:31435–31445 47. Matono R, Miyano K, Kiyohara T, Sumimoto H (2014) Arachidonic acid induces direct interaction of the p67phox–Rac complex with the phagocyte oxidase Nox2, leading to superoxide production. J Biol Chem 289:24874–24884 48. Grizot S, Fieschi F, Dagher MC, Pebay-Peyroula E (2001) The active N-terminal region of p67phox. Structure at 1.8 Å resolution and biochemical characterizations of the A128V mutant implicated in chronic granulomatous disease. J Biol Chem 276:21627–21631 49. Bechor E, Zahavi A, Berdichevsky Y, Pick E (2021) The molecular basis of Rac-GTP action: promoting binding of p67phox to Nox2 by disengaging the β hairpin from downstream residues. J Leukoc Biol 110:219–237 50. Bechor E, Zahavi A, Amichay M, Fradin T, Federman A, Berdichevsky Y, Pick E (2020) p67phox binds to a newly identified site in Nox2 following the disengagement of an intramolecular bond – Canaan sighted ? J Leukoc Biol 107:509–528 51. Berdichevsky Y, Mizrahi A, Ugolev Y, Molshanski-Mor S, Pick E (2007) Tripartite chimeras comprising functional domains derived from the cytosolic NADPH oxidase components p47phox, p67phox, and Rac1 elicit activator-independent superoxide production by phagocyte membranes: an essential role for anionic membrane phospholipids. J Biol Chem 282:22122–22139 52. Kami K, Takeya R, Sumimoto H, Kohda D (2002) Diverse recognition of non-PxxP peptide ligands by the SH3 domains from p67phox, Grb2 and Pex13p. EMBO J 21:4268–4276 53. Mizuki K, Takeya R, Kuribayashi F, Nobuhisa I, Kohda D, Nunoi H, Takeshige K, Sumimoto H (2005) A region C-terminal to the proline-rich core of p47phox regulates activation of the

273 phagocyte NADPH oxidase by interacting with the C-terminal SH3 domain of p67phox. Arch Biochem Biophys 444:185–194 54. Massenet C, Chenavas S, Cohen-Addad C, Dagher M-C, Brandolin G, Pebay-Peyroula E, Fieschi F (2005) Effects of p47phox C terminus phosphorylations on binding interactions with p40phox and p67phox. Structural and functional comparison of p40phox and p67phox SH3 domains. J Biol Chem 280:13752–13761 55. Maehara Y, Miyano K, Sumimoto H (2009) Role for the first SH3 domain of p67phox in activation of superoxide-producing NADPH oxidases. Biochem Biophys Res Commun 379:589–593 56. Matute JD, Arias AA, Wright NA, Wrobel I, Waterhouse CC, Li XJ, Marchal CC, Stull ND, Lewis DB, Steele M, Kellner JD, Yu W, Meroueh SO, Nauseef WM, Dinauer MC (2009) A new genetic subgroup of chronic granulomatous disease with autosomal recessive mutations in p40phox and selective defects in neutrophil NADPH oxidase activity. Blood 114:3309–3315 57. van de Geer A, Nieto-Patlán A, Kuhns DB, Tool AT, Arias AA, Bouaziz M, de Boer M, Franco JL, Gazendam RP, van Hamme JL, van Houdt M, van Leeuwen K, Verkuijlen PJ, van den Berg TK, Alzate JF, Arango-Franco CA, Batura V, Bernasconi AR, Boardman B, Booth C, Burns SO, Cabarcas F, Bensussan NC, Charbit-Henrion F, Corveleyn A, Deswarte C, Azcoiti ME, Foell D, Gallin JI, Garcés C, Guedes M, Hinze CH, Holland SM, Hughes SM, Ibañez P, Malech HL, Meyts I, Moncada-Velez M, Moriya K, Neves E, Oleastro M, Perez L, Rattina V, Oleaga-Quintas C, Warner N, Muise AM, López JS, Trindade E, Vasconcelos J, Vermeire S, Wittkowski H, Worth A, Abel L, Dinauer MC, Arkwright PD, Roos D, Casanova J-L, Kuijpers TW, Bustamante J (2018) Inherited p40phox deficiency differs from classic chronic granulomatous disease. J Clin Invest 128:3957–3975 58. Kuribayashi F, Nunoi H, Wakamatsu K, Tsunawaki S, Sato K, Ito T, Sumimoto H (2002) The adaptor protein p40phox as a positive regulator of the superoxide-producing phagocyte oxidase. EMBO J 21:6312–6320 59. Sumimoto H, Kamakura S, Ito T (2007) Structure and function of the PB1 domain, a protein interaction module conserved in animals, fungi, amoebas, and plants. Sci STKE 2007:re6 60. Ponting CP, Ito T, Moscat J, Diaz-Meco MT, Inagaki F, Sumimoto H (2002) OPR, PC and AID: all in the PB1 family. Trends Biochem Sci 27:10 61. Suh C-I, Stull ND, Li XJ, Tian W, Price MO, Grinstein S, Yaffe MB, Atkinson S, Dinauer MC (2006) The phosphoinositidebinding protein p40phox activates the NADPH oxidase during FcγIIA receptorinduced phagocytosis. J Exp Med 203:1915–1925 62. Ellson CD, Davidson K, Ferguson GJ, O'Connor R, Stephens LR, Hawkins PT (2006) Neutrophils from p40phox-/- mice exhibit severe defects in NADPH oxidase regulation and oxidant-dependent bacterial killing. J Exp Med 203:1927–1937 63. Bánfi B, Clark RA, Steger K, Krause K-H (2003) Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 278:3510–3513 64. Geiszt M, Lekstrom K, Witta J, Leto TL (2003) Proteins homologous to p47phox and p67phox support superoxide production by NAD (P)H oxidase 1 in colon epithelial cells. J Biol Chem 278:20006– 20012 65. Takeya R, Ueno N, Kami K, Taura M, Kohjima M, Izaki T, Nunoi H, Sumimoto H (2003) Novel human homologues of p47phox and p67phox participate in activation of superoxideproducing NADPH oxidases. J Biol Chem 278:25234–25246 66. Vendrov AE, Sumida A, Canugovi C, Lozhkin A, Hayami T, Madamanchi NR, Runge MS (2019) NOXA1-dependent NADPH oxidase regulates redox signaling and phenotype of vascular smooth muscle cell during atherogenesis. Redox Biol 21:101063 67. Paffenholz R, Bergstrom RA, Pasutto F, Wabnitz P, Munroe RJ, Jagla W, Heinzmann U, Marquardt A, Bareiss A, Laufs J, Russ A, Stumm G, Schimenti JC, Bergstrom DE (2004) Vestibular defects in

274 head-tilt mice result from mutations in Nox3, encoding an NADPH oxidase. Genes Dev 18:486–491 68. Mohri H, Ninoyu Y, Sakaguchi H, Hirano S, Saito N, Ueyama T (2021) Nox3-derived superoxide in cochleae induces sensorineural hearing loss. J Neurosci 41:4716–4731 69. Kroviarski Y, Debbabi M, Bachoual R, Périanin A, GougerotPocidalo M-A, El-Benna J, Dang PM (2010) Phosphorylation of NADPH oxidase activator 1 (NOXA1) on serine 282 by MAP kinases and on serine 172 by protein kinase C and protein kinase A prevents NOX1 hyperactivation. FASEB J 24:2077–2092 70. Ranayhossaini DJ, Rodriguez AI, Sahoo S, Chen BB, Mallampalli RK, Kelley EE, Csanyi G, Gladwin MT, Romero G, Pagano PJ (2013) Selective recapitulation of conserved and nonconserved regions of putative NOXA1 protein activation domain confers isoform-specific inhibition of Nox1 oxidase and attenuation of endothelial cell migration. J Biol Chem 288:36437–36450 71. Yamamoto A, Kami K, Takeya R, Sumimoto H (2007) Interaction between the SH3 domains and C-terminal proline-rich region in NADPH oxidase organizer 1 (Noxo1). Biochem Biophys Res Commun 352:560–565 72. Dutta S, Rittinger K (2010) Regulation of NOXO1 activity through reversible interactions with p22phox and NOXA1. PLoS One 5: e10478 73. Yamamoto A, Takeya R, Matsumoto M, Nakayama KI, Sumimoto H (2013) Phosphorylation of Noxo1 at threonine 341 regulates its interaction with Noxa1 and the superoxide-producing activity of Nox1. FEBS J 280:5145–5159 74. Kwon J, Wang A, Burke DJ, Boudreau HE, Lekstrom KJ, Korzeniowska A, Sugamata R, Kim Y-S, Yi L, Ersoy I, Jaeger S, Palaniappan K, Ambruso DR, Jackson SH, Leto TL (2016) Peroxiredoxin 6 (Prdx6) supports NADPH oxidase 1 (Nox1)-based superoxide generation and cell migration. Free Radic Biol Med 96:99–115 75. Leavey PJ, Gonzalez-Aller C, Thurman G, Kleinberg M, Rinckel L, Ambruso DW, Freeman S, Kuypers FA, Ambruso DR (2002) A 29-kDa protein associated with p67phox expresses both peroxiredoxin and phospholipase A2 activity and enhances superoxide anion production by a cell-free system of NADPH oxidase activity. J Biol Chem 277:45181–45187 76. Krishnaiah SY, Dodia C, Feinstein SI, Fisher AB (2013) p67phox terminates the phospholipase A2-derived signal for activation of NADPH oxidase (NOX2). FASEB J 27:2066–2073 77. Chatterjee S, Feinstein SI, Dodia C, Sorokina E, Lien Y-C, Nguyen S, Debolt K, Speicher D, Fisher AB (2011) Peroxiredoxin 6 phosphorylation and subsequent phospholipase A2 activity are required for agonist-mediated activation of NADPH oxidase in mouse pulmonary microvascular endothelium and alveolar macrophages. J Biol Chem 286:11696–11706 78. Ellison MA, Thurman GW, Ambruso DR (2012) Phox activity of differentiated PLB-985 cells is enhanced, in an agonist specific manner, by the PLA2 activity of Prdx6-PLA2. Eur J Immunol 42: 1609–1617 79. Adl SM, Bass D, Lane CE, Lukeš J, Schoch CL, Smirnov A, Agatha S, Berney C, Brown MW, Burki F, Cárdenas P, Čepička I, Chistyakova L, Del Campo J, Dunthorn M, Edvardsen B, Eglit Y, Guillou L, Hampl V, Heiss AA, Hoppenrath M, James TY, Karnkowska A, Karpov S, Kim E, Kolisko M, Kudryavtsev A, Lahr DJG, Lara E, Le Gall L, Lynn DH, Mann DG, Massana R, Mitchell EAD, Morrow C, Park JS, Pawlowski JW, Powell MJ,

H. Sumimoto et al. Richter DJ, Rueckert S, Shadwick L, Shimano S, Spiegel FW, Torruella G, Youssef N, Zlatogursky V, Zhang Q (2019) Revisions to the classification, nomenclature, and diversity of eukaryotes. J Eukaryot Microbiol 66:4–119 80. Lacaze I, Lalucque H, Siegmund U, Silar P, Brun S (2015) Identification of NoxD/Pro41 as the homologue of the p22phox NADPH oxidase subunit in fungi. Mol Microbiol 95:1006–1024 81. Siegmund U, Marschall R, Tudzynski P (2015) BcNoxD, a putative ER protein, is a new component of the NADPH oxidase complex in Botrytis cinerea. Mol Microbiol 95:988–1005 82. Zhao Y-L, Zhou T-T, Guo H-S (2016) Hyphopodium-specific VdNoxB/VdPls1-dependent ROS-Ca2+-signaling is required for plant infection by Verticillium dahliae. PLoS Pathog 12:e1005793 83. Marschall R, Siegmund U, Burbank J, Tudzynski P (2016) Update on Nox function, site of action and regulation in Botrytis cinerea. Fungal Biol Biotechnol 3:8 84. Green KA, Eaton CJ, Savoian MS, Scott B (2019) A homologue of the fungal tetraspanin Pls1 is required for Epichloë festucae expressorium formation and establishment of a mutualistic interaction with Lolium perenne. Mol Plant Pathol 20:961–975 85. Takemoto D, Tanaka A, Scott B (2006) A p67Phox-like regulator is recruited to control hyphal branching in a fungal-grass mutualistic symbiosis. Plant Cell 18:2807–2821 86. Takemoto D, Kamakura S, Saikia S, Becker Y, Wrenn R, Tanaka A, Sumimoto H, Scott B (2011) Polarity proteins Bem1 and Cdc24 are components of the filamentous fungal NADPH oxidase complex. Proc Natl Acad Sci U S A 108:2861–2866 87. Noda Y, Kohjima M, Izaki T, Ota K, Yoshinaga S, Inagaki F, Ito T, Sumimoto H (2003) Molecular recognition in dimerization between PB1 domains. J Biol Chem 278:43516–43524 88. Wilson MI, Gill DJ, Perisic O, Quinn MT, Williams RL (2003) PB1 domain-mediated heterodimerization in NADPH oxidase and signaling complexes of atypical protein kinase C with Par6 and p62. Mol Cell 12:39–50 89. Lamark T, Perander M, Outzen H, Kristiansen K, Øvervatn A, Michaelsen E, Bjørkøy G, Johansen T (2003) Interaction codes within the family of mammalian Phox and Bem1p domaincontaining proteins. J Biol Chem 278:34568–34581 90. Evans TD, Sergin I, Zhang X, Razani B (2017) Target acquired: selective autophagy in cardiometabolic disease. Sci Signal 10: eaag2298 91. Johansen T, Lamark T (2020) Selective autophagy: ATG8 family proteins, LIR motifs and cargo receptors. J Mol Biol 432:80–103 92. Ryder LS, Dagdas YF, Kershaw MJ, Venkataraman C, Madzvamuse A, Yan X, Cruz-Mireles N, Soanes DM, OsesRuiz M, Styles V, Sklenar J, Menke FLH, Talbot NJ (2019) A sensor kinase controls turgor-driven plant infection by the rice blast fungus. Nature 574:423–427 93. Hann J, Bueb J-L, Tolle F, Bréchard S (2020) Calcium signaling and regulation of neutrophil functions: still a long way to go. J Leukoc Biol 107:285–297 94. Brandes RP, Schröder K (2014) NOXious phosphorylation: smooth muscle reactive oxygen species production is facilitated by direct activation of the NADPH oxidase Nox1. Circ Res 115:898–900 95. Belambri SA, Marzaioli V, Hurtado-Nedelec M, Pintard C, Liang S, Liu Y, Boussetta T, Gougerot-Pocidalo M-A, Ye RD, Dang PM, El-Benna J (2022) Impaired p47phox phosphorylation in neutrophils from patients with p67phox-deficient chronic granulomatous disease. Blood 139:2512–2522

p40phox: Composition, Function and Consequences of Its Absence

17

Taco W. Kuijpers and Dirk Roos

Abstract

The NOX2 protein complex in human leukocytes consists of five proteins: gp91phox (NOX2) and p22phox (together called flavocytochrome b558) embedded in membranes, and p47phox, p67phox and p40phox in the cytosol. During cell activation (e.g. during phagocytosis of bacteria or fungi) the cytosolic components move to the membranes and form a complex with the flavocytochrome. This complex formation initiates the enzymatic activity of NOX2, i.e. accepting electrons from NADPH in the cytosol and transporting these to molecular oxygen on the apical side of the membrane, thus generating superoxide (O2–). This process is essential for killing pathogens. P40phox has a special function in this process: it directs the NOX2 complex to the membranes of phagosomes, the vacuoles that contain ingested microorganisms, thus ensuring superoxide formation in close vicinity to the pathogenic targets. Individuals with mutations in one of the genes that encode the NOX2 components suffer from a disease called chronic granulomatous disease (CGD). In general, these patients suffer from recurrent life-threatening infections with bacteria and fungi. However, patients with mutations in NCF4, encoding p40phox, rarely present with infections but instead with autoimmune and hyperinflammatory symptoms. The explanation may lie in disturbed suppression of autoreactivity against RNA antigens in B cells. Keywords

p40phox · NADPH oxidase · Chronic granulomatous disease · Inflammation · Autoimmunity · Hyperreactivity T. W. Kuijpers (✉) Department of Pediatrics, Amsterdam University Medical Center, location AMC, Amsterdam, The Netherlands e-mail: [email protected] D. Roos Sanquin Research, and Landsteiner Laboratory, Amsterdam University Medical Center, location AMC, Amsterdam, The Netherlands e-mail: [email protected]

1

Introduction

P40phox was discovered as a component of the leukocyte NADPH oxidase complex by Wientjes et al. in 1993 [1]. These authors found this protein of about 40 kD to be present in a heterotrimeric complex with p47phox and p67phox in the cytosol of neutrophils from healthy human donors. Upon activation of human neutrophils with PMA, all three phox proteins translocated from the cytosol to the membrane fraction. This process required the presence of flavocytochrome b558 in the membranes, thus suggesting the joint involvement of all three cytosolic phox proteins in NADPH oxidase activation. The association of p40phox with p67phox seemed to be stronger than with p47phox, given the fact that cytosol from p67phox-deficient CGD patients contained significantly less p40phox than that of normal donors or that of p47phox-deficient CGD patients [1]. (However, the reverse is not true: p40phox-deficient neutrophils contain normal amounts of p67phox [2]). Moreover, pull down of p67phox from neutrophil cytosol with a monoclonal antibody also precipitated p40phox, but addition of a monoclonal against p47phox did not result in any co-precipitation. These results were confirmed by Tsunawaki et al. [3]. The PB1 domain-mediated interaction between p40phox and p67phox has been studied in the crystallized PB1 heterodimer, which revealed a front-to-back arrangement through electrostatic interactions between an acidic OPCA motif (octicosapeptide repeat [OPR], PC motif [phox and cdc24p], and AID motif [atypical protein kinase C-interaction domain]) in p40phox with basic residues in p67phox [4]. Partial amino acid (aa) sequencing of the purified protein and searching of a cDNA library from HL60 cells with a PCR probe from the largest sequenced peptide identified the cDNA sequence of p40phox [1]. It contains an open reading frame of 1017 nucleotides, encoding a 339 amino acidcontaining protein with a predicted molecular mass of 37 kD. The N-terminal part of p40phox was found to be

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_17

275

276

T. W. Kuijpers and D. Roos

67% similar to that of p47phox. In addition, the C-terminal part of p40phox contains an SH3 domain (aa 173–228), similar to the situation in the C-termini of p47phox and p67phox. Subsequently, the structure of the NCF4 gene that encodes p40phox has been elucidated, as well as its chromosomal localization [5]. NCF4 spans about 18 kb and contains 10 exons. It is localized on chromosome 22 at 22q13.1. The start site of transcription was mapped at 196 nucleotides upstream of the ATG translation initiation codon. mRNA of p40phox was detected in human neutrophils, monocytes, B lymphocytes and eosinophils, all known to express NADPH oxidase components, but also in mast cells, basophils and megakaryoblasts, which do not express such components [5]. This raises the possibility of p40phox to be involved in additional cellular structures or processes beyond NADPH oxidase activation (see the remark about polymorphisms in p40phox at the end of the section “Role in inflammatory reactions” in this chapter).

2

Role in NADPH Oxidase Activation

Since it was known that p40phox is not needed for oxidase activity in a cell-free system composed of recombinant p67phox, p47phox and RAC, together with pure flavocytochrome b558 [6], a search was started as to what the role of p40phox might be in intact cells. First, Fuchs et al. in a yeast two-hybrid system, analyzed the various proteinprotein interactions between regions of p40phox, p47phox and p67phox [7]. Interaction between p47phox and p67phox appeared to be restricted to the C-terminal SH3 domain of p67phox and the proline-rich region of p47phox. This last region was also found to mediate binding to the SH3 domain of p40phox, as well as an intramolecular interaction within p47phox. Interaction between p67phox and p40phox was localized in a 150 amino-acid stretch between the two SH3 domains of p67phox and the C-terminal tail of p40phox. As this last region contains neither an SH3 domain nor a proline-rich region, a new kind of protein-protein interaction was suspected. This was later identified as interaction between the PB1 (phox and Bem1) domains in the C-termini of p67phox and p40phox (at aa 285–306) [4, 8]. A schematic picture of these interactions is shown in Fig. 17.1. In its N-terminus p40phox also contains a so-called PX (phox homology) domain, at aa 19–140, able to interact with phospholipids in the membrane. The p40phox PX domain has a high specificity for phosphatidyl-inositol-3-phosphate [PI(3)P] [10, 11]. This PX domain in p40phox, bound to PI(3) P, has been crystallized and studied by 1.7-Å resolution X-ray diffraction [12]. This showed a positively charged pocket in the protein that specifically bound PI(3)P. This phospholipid is generated during phagocytosis in the phagosomal membrane and on early endosomes by the

Class III PI3 kinase Vps34 [13–15]. However, in the resting state of the phagocyte, p40phox is unable to bind to PI(3)P, because the PX domain is then shielded by the PB1 domain in p40phox, which involves the side of the PB1 domain that is opposite of the face that interacts with p67phox [16]. This autoinhibitory PX-PB1 interaction is disrupted by reactive oxygen compounds in the cytoplasm and by p40phox targeting to the membrane of phagosomes [17]. The binding of PI(3)P to p40phox is critical for its positive regulation of NADPH oxidase activity when the oxidase is assembled on intracellular membranes such as phagosomes. To become fully active, all phox proteins need to be phosphorylated at serine and/or threonine residues during phagocyte activation. For a recent review of this subject see [18]. Phosphorylation of p40phox takes place primarily at Thr154 and Ser315, through action of a protein kinase C [19]. This reaction increases the association between recombinant p67phox and p47phox with purified flavocytochrome b558 in a cell-free system, possibly by inducing a conformational change in p67phox that allows it to fully interact with the flavocytochrome [20]. In reconstituted mouse p40phox-/neutrophils, phosphorylation of p40phox at Thr154 but not at Ser315 proved essential for p47phox translocation and NADPH activation induced by fMLF, serum-opsonized S. aureus or IgG-SRBCs, but not for the translocation and activation induced by PMA [21]. In contrast to the situation in p47phox, phosphorylation of other phox proteins induces largely unknown structural effects in these proteins [18]. Thus, also in p40phox, it is not known how phosphorylation of this protein induces the oxidase activity-increasing effects. The third reaction required for NADPH oxidase activation, is GDP-GTP exchange on a RAC protein (RAC2 in human neutrophils, RAC1 and RAC2 in human macrophages) [22]). This nucleotide exchange releases RAC from its Rho-GDI inhibitor in resting phagocytes and enables association of RAC with p67phox (in its four tetratricopeptide [TPR] motifs), with gp91phox (in an aa 419–430 cytosolic stretch) [23, 24] and with the plasma membrane (via RAC’s prenylated C-terminus) [9, 25]. However, since p40phox has no direct role in this chain of events, this subject will not discussed any further in this chapter. For more details, the reader is referred to the Chap. 1 by E. Pick and Chap. 18 by Y. Lin and Y. Zheng). In resting cells, p40phox, together with p67phox, as well as p47phox by itself, are attached to the cytoskeleton, in casu to coronin, an actin-binding protein [26, 27]. This attachment may also be mediated through binding of p40phox and p47phox to moesin, another actin-binding protein. Indeed, the PX domains of p40phox and p47phox are able to bind the N-terminal part of moesin in a PI-dependent manner [28, 29]. However, X-linked moesin deficiency (X-MAID) does not present with CGD-like symptoms, rendering the role

17

p40phox: Composition, Function and Consequences of Its Absence

Fig. 17.1 Phagocyte NADPH oxidase. The domain architecture of the NADPH oxidase subunits is shown schematically. Blue dotted lines, interactions between phox subunits prior to activation. Red dotted lines, additional interactions that mediate NADPH oxidase assembly, localization and activity. Membrane components include gp91phox (NOX2) and p22phox. Three soluble phox subunits, p47phox, p67phox and p40phox, are linked in a trimeric complex in the cytosol prior to cellular activation. p67phox is constitutively associated with p40phox via complementary PB1 motifs in their C-termini. p67phox is also tethered to p47phox via a tail-to-tail interaction involving a SH3 domain and a proline-rich region (PRR) in the respective C-termini of these subunits. p67phox contains four tetratricopeptide (TPR) motifs that create a binding surface for Rac-GTP. p47phox contains tandem SH3 domains for binding its target proline-rich PxxP motif in p22phox as well as a regulatory PX domain that binds phosphoinositides, including PI(3,4)P2 and, less strongly, other phosphoinositides, as well as phosphatidylserine (PS) and phosphatidic acid (PA). The p47phox PX and SH3 domains are

of moesin in NADPH oxidase component stabilization questionable. Direct binding of p40phox via its PX domain to F-actin has also been reported [27]. This interaction with the cytoskeleton may constitute a means of safely keeping in check the cytosolic NADPH oxidase subunits in resting cells, as well as indicate involvement of these phox proteins in the reorganization of the submembranous cytoskeleton during phagosome formation. During phagocyte activation, p47phox associates with the p67phox-p40phox complex, and the whole trimer then translocates to the membrane [1, 30–34]. The stoichiometry of the three subunits in the complex is 1:1:1, with almost all subunits complexed in living cells [35].

277

masked in the resting state by an adjacent polybasic autoinhibitory domain. The PX domain in p40phox subunit, in contrast, is highly specific for PI(3)P, but is also unable to access its target in resting cells. GTP-bound Rac is an essential subunit of the NADPH oxidase. In unactivated cells, Rac-GDP resides in the cytoplasm because its isoprenylated CAAX motif is masked by Rho-GDI. Following cellular activation, phosphorylation of the p47phox autoinhibitory domain exposes its SH3 motifs, leading to translocation of the trimeric phox complex to flavocytochrome b558, where an activation domain (AD) in p67phox promotes electron transfer through gp91phox. Activation of GEFs results in exchange of GTP for GDP on Rac as well as removal of Rho-GDI, allowing Rac to bind to the membrane via its isoprenyl group and adjacent polybasic domain. Interactions between the effector domain of Rac-GTP with p67phox and gp91phox are also required for NADPH oxidase activity. PX domains in p47phox and p40phox additionally regulate the enzyme on plasma and phagosome membranes, respectively. Reproduced with permission from Nunes et al. [9]

3

Role in NADPH Oxidase Localization

The role of p40phox in localization of NADPH oxidase activity in phagocytes depends purely on its PX domain, with specificity for PI(3)P. This phospho-inositide is generated almost exclusively in phagosomal membranes during particle uptake [14, 36, 37]. However, one has to realize that translocation of p40phox from the cytosol to the membrane as such does not mean that it has a role there in enhancing NADPH oxidase activity. Such translocation can also be the result of co-translocation with p67phox and p47phox. Without the

278

T. W. Kuijpers and D. Roos

presence of PI(3)P, p40phox will not modulate the ROS-producing activity of the complex. During phagocytosis, be it via IgG opsonins to Fcγ receptors, C3b/iC3b opsonins to the β2-integrin receptor CR3, bacterial or fungal β-glucans to CR3 and the dectin-1 receptor or via CD36 scavenger receptor on phagocytes, a number of different signal transduction pathways are activated. In general, immunoreceptor tyrosine activation motifs (ITAM) are activated by Scr family kinase –mediated phosphorylation. This process takes place either on the receptors themselves or on Dap12 and Fcγ receptor adaptors. This induces recruitment of the Syk tyrosine kinase and the resulting activation of downstream effectors, including PLCγ that activates PKC-δ (needed for phosphorylation of NADPH oxidase components) and the Vav family of Guanine nucleotide Exchange Factors (GEFs) needed for activation of RAC. The class III PI3 kinase Vps34 is constitutively active on Rab5-positive early endosomes, and the fusion of these endosomes with the nascent phagosome leads to generation of PI(3)P in the phagosomal membrane [9, 13]. At the same time, at least in mouse macrophages, some PI(3,4)P2 is also generated in the phagosomal membrane after Fcγ receptor engagement. This takes place by action of the inositol phosphatase SHIP-1 on PI(3,4,5)P3 and may be important for enhancing the activity of PKC-δ. More details can be found in ref. [9]. Translocation of p40phox to the phagosome membranes of neutrophils depends on the presence of p67phox and p47phox, as deduced from experiments with p67phox- or p47phox-deficient neutrophils [38]. This can be explained by the association of p40phox with p67phox and—after stimulation—also with p47phox, as well as by a diminished expression of p40phox in cells deficient of p67phox [1]. However, the reverse is not true: in the absence of p40phox, both p67phox and p47phox translocate normally [39, 40]. P40phox with a mutation in the PX domain that prevents association with PI(3)P does translocate to the membranes of phagosomes (possibly through transient interaction with the cytoskeleton) but dissociates rapidly, despite the retained presence of p67phox [41]. This indicates that the PB1-PB1 interaction between p40phox and p67phox is no longer operative under these conditions.

4

Role in NADPH Oxidase Activity Regulation

Proof that the p40phox binding to PI(3)P in the phagosomal membrane is important not only for the localization but also for the activation of the NADPH oxidase has come from patients with genetic defects in p40phox, from genetically engineered cells and from murine phagocytes [2, 40–

45]. These studies have shown that translocation of p40phox is needed for activation of oxidase activity on phagosome membranes and that this process depends on PI(3)P binding to the PX domain of p40phox. This process was found to be essential in the human situation when a CGD patient was discovered with a p40phox deficiency [41]. This patient was a compound heterozygote for two different mutations on the two NCF4 alleles, i.e. one that destroyed p40phox expression by a frameshift mutation that induced a premature termination of protein translation, and one in the PX domain that destroyed its PI (3)P-binding capacity. The neutrophils of this patient showed only transient translocation of p40phox to the membrane fraction (see previous paragraph) but also severely reduced intracellular NADPH oxidase activation by particulate stimuli (serum-opsonized zymosan or S. aureus, IgG-coated beads) but not by soluble stimuli (PMA, fMLF). Recently, a report has described another 24 p40phox-deficient CGD patients with similar abnormalities [2]. The clinical presentation of these patients is described below. Exactly how p40phox enhances oxidase activity on phagosome membranes is not fully understood. Assuming that interaction of the activation domain of p67phox with gp91phox is critical in every situation leading to initiation of oxidase activity [46], the most straightforward explanation would be the recruitment of p67phox to phagosomes by p40phox. However, p67phox translocation to phagosomes does not require the presence of p40phox [39, 40]. Therefore, it must be assumed that the presence of p40phox enables p67phox to interact efficiently with gp91phox, perhaps via conformational effects of p40phox on gp91phox [9] . This also assumes a different conformation of gp91phox in the phagosomal membrane from that in the plasma membrane, the first needing the combination of p40phox and p67phox (among other signals) to be activated and the latter only p67phox (among other signals). Regulation of NADPH oxidase activity by p40phox is even more complicated, because p40phox can also have an inhibitory effect. This was shown both in a cell-free system with PKC-phosphorylated recombinant (rec.) p47phox, rec. p67phox and neutrophil membranes [47] and in a K562 whole cell assay after co-transfection with rec. p40phox [48]. In the first mentioned cell-free assay, phosphorylation of p40phox was decisive: unphosphorylated p40phox did not inhibit oxidase activity, but phosphorylated p40phox (on Thr154) did. Possibly, this was caused by a phosphorylation-dependent change in conformation and might play a role in termination of the respiratory burst. In the second assay, it was shown that the isolated SH3 domain of p40phox was even more effective in inhibiting oxidase activity than intact p40phox. This SH3 domain binds to the same proline-rich region in p47phox that interacts with p67phox and might thus interfere with the

17

p40phox: Composition, Function and Consequences of Its Absence

interaction between the cytosolic phox proteins, rendering interpretation of the role of native cytosolic phox proteins in intact cells difficult.

5

Role in Inflammatory Reactions

It is well known that CGD patients suffer not only from recurrent bacterial and fungal infections but also from excessive inflammatory reactions and manifestations of autoimmunity [49]. The origin of this inflammatory state is not well understood. Currently, several possible explanations are considered. One of these is related to the autophagy process, another to undue T-cell activation. Autophagy is the process in which redundant or damaged cytoplasmic constituents in a cell are removed by vesicle formation around the material, fusion with lysosomes and degradation. There are strong indications that autophagy is also involved in the killing process of infectious microorganisms by phagocytes—in particular by macrophages (for a review see [50]). Since autophagy is promoted by reactive oxygen species (ROS), this process may be diminished in CGD cells. This can lead to a decreased bacterial killing as a result of insufficient phagosome maturation but also to increased release of the pro-inflammatory cytokines IL-1β and IL-18 as a result of excessive inflammasome activity [51]. In Paneth cells in the intestines autophagy proteins may be involved in proper handling of intestinal microbes [52]. If this process is impaired, this again may lead to high levels of IL-1β and IL-18 secretion and contribute to occurrence of inflammatory bowel disease, frequently seen in CGD patients [49, 50]. In goblet cells in the colon, mucus is produced that forms a barrier against intestinal microbes. In CGD mice it was found that this mucus secretion was impaired as a result of decreased autophagy [53], and thus—in the human situation—could be involved in the occurrence of Crohn’s disease in CGD patients. Another possible cause of hyperinflammation in CGD patients is a defect in the regulation of T-cell activity. It is now known that human neutrophils can act as so-called myeloid-derived suppressor cells (MDSCs) of T-cell proliferation [54]. This suppressor activity is ROS-dependent, and therefore strongly diminished in CGD patients. Similarly, murine macrophages have been found to be involved in inhibiting autoreactive T-cell responses [55]. The involvement of p40phox in susceptibility to hyperinflammation and autoimmune phenomena has been studied experimentally largely in mice. Conway et al. [56] used p40phox knock-out mice and found that these animals showed enhanced colon inflammation to both dextran sulfate sodium (DSS) and anti-CD40, excessive local generation of IFN-γ, TNF-α, IL-1β, IL-6, iNOS (inducible nitric oxide synthase) and IL-17A, and increased neutrophil recruitment

279

to the colon. These phenomena developed during the recovery phase of the inflammation, 6 days after its induction. A caveat in this study is that the p40phox KO mice have markedly reduced p67phox levels, and very low oxidase activity, so it is hard to distinguish this effect from a specialized role for p40phox. Neutrophils with intact p40phox were essential for complete recovery, perhaps by normalizing chemokine receptor-1 expression and glycan metabolism. Winter et al. [44] studied collagen-induced arthritis and mannan-induced psoriatic arthritis-like disease in mice lacking p40phox and in mice with an Arg58Ala mutation in p40phox that precludes its binding to PI(3)P. These mice displayed delayed neutrophil apoptosis and aggravated disease manifestation, with stronger clinical reactions in case of p40phox absence than in case of p40phox lacking PI(3)P binding. Finally Bagaitkar et al. [45], also studying p40phoxR58A/ R58A mice, found that these mice had increased numbers of newly recruited neutrophils and monocytes in peritoneal inflammation elicited by zymosan, monosodium urate crystals or sodium periodate. The resolution of such inflammation was retarded in the genetically engineered mice. These studies indicate a direct link between the lack of p40phox binding to PI(3)P during an inflammatory response and the severity of this response and the duration of recovery from this response. However, from patients with p40phox deficiency, we know that Crohn’s disease and other autoinflammatory manifestations are found even more frequently than in patients with other causes of CGD [2]. More details are given below. Why this is so is unclear at present. Perhaps this relates to functions of p40phox other than supporting localization and activity of the leukocyte NADPH oxidase. Indications for this idea may be derived from polymorphisms in NCF4 that do not affect the NADPH oxidase activity but are found more frequently in patients with lung disease, rheumatoid arthritis or Crohn’s disease [57–59].

6

Differences Between Mice and Men

A note of caution must be mentioned. The reactions measured in human and murine neutrophils with defects in p40phox are not always similar. Neutrophils from human patients with p40phox deficiency display a clear defect in NADPH oxidase activity after addition in vitro of IgG-opsonized particles, complement fragment C3b-opsonized S. aureus, A. fumigatus hyphae, serumopsonized zymosan or bare zymosan. In contrast, normal oxidase activity is generated in these cells with PMA or fMLF. However, in murine neutrophils with p40phoxR58A/ R58A ROS production elicited by A. fumigates hyphae, serum-opsonized zymosan or bare zymosan is completely normal, and ROS production induced by IgG-opsonized

280

T. W. Kuijpers and D. Roos

particles or serum-opsonized S. aureus is inhibited by about 70–80% [36, 45, 60, 61]. Moreover, in p40phoxR58A/R58A mouse macrophages incubated with IgG-opsonized particles a defect in plasma membrane oxidase activity was noted [45]. These contradictory results should warn us not to draw far-reaching conclusions about the human situation from murine studies.

7

Clinical Phenotype in Patients

Deficiency of p40phox is not characterized by invasive bacterial and fungal infections, as in other CGD subtypes, but mostly by auto-inflammation and peripheral infections of the skin [2, 41]. Some individuals are even asymptomatic. Among the symptomatic patients, about half presented with inflammatory lupus-like skin lesions, granulomatous gastrointestinal manifestations including oral ulcers and periodontitis, esophagitis, gastritis, Crohn-like inflammatory bowel disease and perianal abscesses/fistula. Some lesions were probably both infectious and inflammatory, as often seen in CGD. A minority of the p40phox-deficient patients was reported to suffer from pulmonary infection, which might, however, have been interstitial lung disease that responds to systemic steroids rather than antibiotics. Evaluation of autoantibodies is rarely positive [2]. Studies with human “knock-out” neutrophils from CGD patients, independent of the knock-out animal models and in-vitro cellular models, have expanded our understanding of the importance of and the interaction between ROS-mediated and ROS-independent killing mechanisms as well as the role of pattern recognition receptors (PRRs) and signaling in human microbial killing. From the studies with CGD patients it is becoming clear that neutrophils employ fundamentally distinct mechanisms to kill bacteria and fungi. CGD patients with a ‘classic’ defect in the NADPH oxidase activity (see Chap. 32 by M.J. Stasia and D. Roos) are susceptible to invasive bacterial and fungal infections, accompanied by increased rates of mortality. CGD can present any time from infancy to late adulthood, but the majority of patients are diagnosed during childhood. Several tissues and organs can be infected, such as lung, skin, lymph nodes, and liver as the sites of infection most frequently affected, followed by osteomyelitis, perianal abscesses, and gingivitis [49, 62, 63] (see Chap. 32 by M.J Stasia and D. Roos). The main pathogens causing the major infections in CGD consist of only a limited number of organisms: S. aureus, Burkholderia (Pseudomonas) cepacia, Serratia marcescens, Nocardia and Aspergillus species, but also Klebsiella pneumoniae and Salmonella species. Bacteremia is uncommon, and usually due to B. cepacia or S. marcescens. Also BCG complications following vaccination have been noted, ranging from none to self-limited localized or disseminated

BCG-itis. Increased prevalence of tuberculosis in CGD patients is found in areas where TB is endemic. With trimethoprim-sulfamethoxazole prophylaxis, the frequency of bacterial infections in CGD in general has diminished. Often antifungal therapy is prescribed as additional prophylactic measure to prevent invasive fungal infections, which are dominated by Aspergilus fumigatus and A. nidulans strains, and to a much lesser extent by Candida albicans and C. glabrata. X-CGD patients have a more severe disease course than patients with the p47phox AR-CGD, presumably because they lack all oxidase activity [64, 65]. Residual oxidase activity in neutrophils is linked to reduced disease severity, and modest production of ROS seems to confer a greater likelihood of long-term survival [64, 65]. In that respect, it may not be surprising that the residual activity in p40phox deficiency protects from the major disease manifestations in CGD. In fact, the lack of fungal infections in p40phox deficiency forms a very clear and distinctive difference with ‘classic’ X- linked and AR CGD subtypes (p22phox, p47phox, p67phox). Studying a series of p40phox deficient pedigrees, we have not noted any fungal infection in 23 genetically affected cases [2]. When testing neutrophils from p40phox-deficient patients, these cells were able to kill conidia and prevent hyphenation in vitro in Candida and Aspergillus cultures, which clearly indicates an essential difference in potential host defense mechanisms of p40phox-deficient neutrophils in comparison with cells from classic CGD. Neutrophils of classic CGD patients are unable to kill Aspergillus fumigatus under various test conditions [66, 67]. The germination of unopsonized C albicans is as efficiently prevented (in a ROS-independent way) by CGD as by control neutrophils. Only with opsonized C. albicans we noted outgrowth of hyphae in the overnight germination assay with classic CGD neutrophils, which we explained by a difference in phagosomal compartmentalization by IgG Fcγ receptors (FcγRs). In contrast to the often problematic Aspergillus infections—once these are invasive—the fact that clinical Candida infections generally are not overwhelmingly prominent in p40phox-deficient CGD patients is then explained by the fact that most Candida colonization occurs at skin and mucosa, where host defense mediates clearance in CGD patients under unopsonized conditions [66, 67] In contrast, in p40phox deficiency we find normal Candida and Aspergillus killing whatsoever, supporting the clinical findings [2]. Also surprising is the absence of severe bacterial infections in p40phox-deficient patients. Apart from skin abscesses by S. aureus, we have not observed major bacterial complications. Under experimental conditions, CGD neutrophils show a clear S. aureus killing defect, in contrast to E. coli killing. Killing of S. aureus is a ROS-dependent process, while killing of E. coli is relatively ROS-independent [68, 69]. Since in vitro S. aureus-induced

17

p40phox: Composition, Function and Consequences of Its Absence

ROS formation and killing are severely impaired in the p40phox-deficient patients to the same extent as in classic CGD patients, the absence of invasive S. aureus infections in p40phox-deficient patients is still unexplained, but may perhaps be related to the role of an intact extracellular ROS production, in contrast to the defect in the intracellular killing system. Reasoning along these lines, the localized skin abscesses are then a consequence of skin colonization, local loss of barrier function, and a niche for persisting infection without definite clearance. The lack of invasive bacterial and fungal infections contrasts with the striking presence of severe inflammatory responses in p40phox-deficient CGD patients, which can be severe to the extent that hematopoietic stem cell transplantation (HSCT) is the ultimate rescue treatment of choice [2]. In a French cohort of classic CGD patients, more than half of the patients suffered from inflammation (2.2 fold more often in X-CGD than in AR-CGD). Most commonly affecting organs in CGD are the gastrointestinal tract (in 88%) and lungs (26%). Autoimmune manifestations such as systemic lupus erythematosus (SLE), or rheumatoid arthritis (RA) were recorded in 19% and 10% of the patients [70] In the much smaller cohort of p40phox deficiency more than half of the patients showed excessive lupus-like skin lesions or inflammation of the gastrointestinal tract and sometimes of the lungs [2]. Enhanced inflammation was also noted in mice with p40phox deficiency [45, 56]. In the light of a lack of a clear infectious phenotype and the residual ROS production in neutrophils elicited by soluble stimuli and apparently normal or subnormal respiratory burst in macrophages, the clinical manifestations in p40phox deficiency are surprising. Neutrophil ROS may contribute to the clearance of apoptotic cells or cell debris, their deficiency possibly underlying inflammation or autoimmune phenomena in various ways [71–73]. When tested, the mononuclear phagocytes in p40phox deficiency were apparently normal in terms of respiratory burst, which makes it unlikely that these cells have a role in the hyperinflammation phenomena. Whereas some studies suggest gp91phox to be required for restraining the expression of the immune suppressive molecules on Tregs, including CTLA-4, CD39 and CD73 [74] we have not been able to confirm these findings and find normal Treg activity—also in p40phox-deficient patients [75]. Alternatively, inflammation in both CGD and p40phox deficiency can also be due to abnormal ROS production in B cells, as NADPH activity was defective in all p40phoxdeficient EBV-B cells tested without any residual activity, confirming that p40phox is not redundant in human B cells [2]. As shown previously, presentation of cytoplasmic and exogenous antigen to CD4+ T cells by MHC class II molecules is impaired in human p40phox-deficient B cells (i.e. EBV-transformed lymphobastoid B cell lines, B-LCLs) [76]. Reconstitution with wild-type p40phox restores both

281

intracellular ROS generation and exogenous Ag presentation, which suggests a subtle impairment in B cell-T cell interactions in CGD, as is supported by altered antibody responses in CGD [77–79]. In contrast, class II presentation of epitopes from membrane antigens was robust in p40phoxdeficient B cells, suggesting a tendency to inherent hyperreactivity. This suggested hyperresponsiveness of B cells in relation to the increased susceptibility to autoreactivity in CGD patients as well as X-linked CGD carriers may be explained by the role of TLR7 in B cells. Under normal conditions TLR7 may act as a redox sensor relevant to endosomal ROS production, cytokine production as well as antigenspecific antibody responses [80] Although it needs confirmation, the hypothesis is that suppression of TLR7 activation by endosomal NADPH oxidase activity is a mechanism that has evolved to inhibit inflammatory responsiveness against selfRNA/antigens and autoimmunity as a consequence, but results in enhanced viral pathogenicity as a downside. The TLR7 gene is located on the X-chromosome and has been identified as one of the rare alleles that are insensitive to X-inactivation [81]. This may provide an explanation for the increased risk of autoimmunity in females. The impact of a loss of ROS generation in part of the B cells (CYBB, the X-linked gene encoding gp91phox is subject to X-chromosome inactivation) may also explain the fact that female carriers of X-CGD show lupus-like autoimmune manifestations [82, 83]. Understanding the excessive inflammation in patients with CGD or p40phox deficiency requires further studies in B cells to fully unravel the actual cell type most involved in steering the prolonged and severe inflammatory responses in these disorders. Acknowledgements We thank Dr. Mary C. Dinauer for helpful suggestions concerning the text of this chapter.

References 1. Wientjes FB, Hsuan JJ, Totty NF, Segal AW (1993) p40phox, a third cytosolic component of the activation complex of the NADPH oxidase to contain src homology 3 domains. Biochem J 296:557– 561. https://doi.org/10.1042/bj2960557 2. Van de Geer A, Nieto-Patlán A, Kuhns DB, Tool AT, Arias AA, Bouaziz M, de Boer M, Franco JL, Gazendam RP, van Hamme JL, van Houdt M, van Leeuwen K, Verkuijlen PJ, van den Berg TK, Alzate JF, Arango-Franco CA, Batura V, Bernasconi AR, Boardman B, Booth C, Burns SO, Cabarcas F, Bensussan NC, Charbit-Henrion F, Corveleyn A, Deswarte C, Azcoiti ME, Foell D, Gallin JI, Garcés C, Guedes M, Hinze CH, Holland SM, Hughes SM, Ibañez P, Malech HL, Meyts I, Moncada-Velez M, Moriya K, Neves E, Oleastro M, Perez L, Rattina V, OleagaQuintas C, Warner N, Muise AM, López JS, Trindade E, Vasconcelos J, Vermeire S, Wittkowski H, Worth A, Abel L, Dinauer MC, Arkwright PD, Roos D, Casanova JL, Kuijpers TW, Bustamante J (2018) Inherited p40phox deficiency differs from

282 classic chronic granulomatous disease. J Clin Invest 128:3957– 3975. https://doi.org/10.1172/JCI97116 3. Tsunawaki S, Mizunari H, Nagata M, Tatsuzawa O, Kuratsuji T (1994) A novel cytosolic component, p40phox, of respiratory burst oxidase associates with p67phox and is absent in patients with chronic granulomatous disease who lack p67phox. Biochem Biophys Res Commun 199:1378–1387. https://doi.org/10.1006/ bbrc.1994.1383 4. Wilson MI, Gill DJ, Perisic O, Quinn MT, Williams RL (2003) PB1 domain-mediated heterodimerization in NADPH oxidase and signaling complexes of atypical protein kinase C with Par6 and p62. Mol Cell 12:39–50. https://doi.org/10.1016/s1097-2765(03) 00246-6 5. Zhan S, Vazquez N, Zhan S, Wientjes FB, Budarf ML, Schrock E, Ried T, Green ED, Chanock SJ (1996) Genomic structure, chromosomal localization, start of transcription, and tissue expression of the human p40-phox, a new component of the nicotinamide adenine dinucleotide phosphate-oxidase complex. Blood 88:2714–2721 6. Abo A, Boyhan A, West I, Thrasher AJ, Segal AW (1992) Reconstitution of neutrophil NADPH oxidase activity in the cell-free system by four components: p67-phox, p47-phox, p21rac1, and cytochrome b-245. J Biol Chem 267:16767–16770 7. Fuchs A, Dagher MC, Fauré J, Vignais PV (1996) Topological organization of the cytosolic activating complex of the superoxidegenerating NADPH-oxidase. Pinpointing the sites of interaction between p47phoz, p67phox and p40phox using the two-hybrid system. Biochim Biophys Acta 1312:39–47. https://doi.org/10. 1016/0167-4889(96)00020-1 8. Ito T, Matsui Y, Ago T, Ota K, Sumimoto H (2001) Novel modular domain PB1 recognizes PC motif to mediate functional proteinprotein interactions. EMBO J 20:3938–3946. https://doi.org/10. 1093/emboj/20.15.3938 9. Nunes P, Demaurex N, Dinauer MC (2013) Regulation of the NADPH oxidase and associated ion fluxes during phagocytosis. Traffic 14:1118–1131. https://doi.org/10.1111/tra.12115 10. Kanai F, Liu H, Field SJ, Akbary H, Matsuo T, Brown GE, Cantley LC, Yaffe MB (2001) The PX domains of p47phox and p40phox bind to lipid products of PI(3)K. Nat Cell Biol 3:675–678. https:// doi.org/10.1038/35083070 11. Ellson CD, Gobert-Gosse S, Anderson KE, Davidson K, ErdjumentBromage H, Tempst P, Thuring JW, Cooper MA, Lim ZY, Holmes AB, Gaffney PR, Coadwell J, Chilvers ER, Hawkins PT, Stephens LR (2001) PtdIns(3)P regulates the neutrophil oxidase complex by binding to the PX domain of p40(phox). Nat Cell Biol 3:679–682. https://doi.org/10.1038/35083076 12. Bravo J, Karathanassis D, Pacold CM, Pacold ME, Ellson CD, Anderson KE, Butler PJ, Lavenir I, Perisic O, Hawkins PT, Stephens L, Williams RL (2001) The crystal structure of the PX domain from p40(phox) bound to phosphatidylinositol 3-phosphate. Mol Cell 8:829–839. https://doi.org/10.1016/s1097-2765(01) 00372-0 13. Vieira OV, Botelho RJ, Rameh L, Brachmann SM, Matsuo T, Davidson HW, Schreiber A, Backer JM, Cantley LC, Grinstein S (2001) Distinct roles of class I and class III phosphatidylinositol 3-kinases in phagosome formation and maturation. J Cell Biol 155: 19–25. https://doi.org/10.1083/jcb.200107069 14. Ellson CD, Anderson KE, Morgan G, Chilvers ER, Lipp P, Stephens LR, Hawkins PT (2001) Phosphatidylinositol 3-phosphate is generated in phagosomal membranes. Curr Biol 11:1631–1635. https://doi.org/10.1016/s0960-9822(01)00447-x 15. Raiborg C, Schink KO, Stenmark H (2013) Class III phosphatidylinositol 3-kinase and its catalytic product PtdIns3P in regulation of endocytic membrane traffic. FEBS J 280:2730–2742. https://doi. org/10.1111/febs.12116 16. Honbou K, Minakami R, Yuzawa S, Takeya R, Suzuki NN, Kamakura S, Sumimoto H, Inagaki F (2007) Full-length p40phox structure suggests a basis for regulation mechanism of its membrane

T. W. Kuijpers and D. Roos binding. EMBO J 26:1176–1186. https://doi.org/10.1038/sj.emboj. 7601561 17. Ueyama T, Nakakita J, Nakamura T, Kobayashi T, Kobayashi T, Son J, Sakuma M, Sakaguchi H, Leto TL, Saito N (2011) Cooperation of p40(phox) with p47(phox) for Nox2-based NADPH oxidase activation during Fcγ receptor (FcγR)-mediated phagocytosis: mechanism for acquisition of p40(phox) phosphatidylinositol 3-phosphate (PI(3)P) binding. J Biol Chem 286(47):40693–40705. https://doi.org/10.1074/jbc.M111.237289 18. Belambri SA, Rolas L, Raad H, Hurtado-Nedelec M, Dang PM, El-Benna J (2018) NADPH oxidase activation in neutrophils: Role of the phosphorylation of its subunits. Eur J Clin Investig 48(Suppl 2):e12951. https://doi.org/10.1111/eci.12951 19. Bouin AP, Grandvaux N, Vignais PV, Fuchs A (1998) P40(phox) is phosphorylated on threonine 154 and serine 315 during activation of the phagocyte NADPH oxidase. Implication of a protein kinase c-type kinase in the phosphorylation process. J Biol Chem 273(46):30097–30103. https://doi.org/10.1074/jbc.273.46.30097 20. Dang PM, Cross AR, Quinn MT, Babior BM (2002) Assembly of the neutrophil respiratory burst oxidase: a direct interaction between p67PHOX and cytochrome b558 II. Proc Natl Acad Sci U S A 99: 4262–4265. https://doi.org/10.1073/pnas.072345299 21. Chessa TA, Anderson KE, Hu Y, Xu Q, Rausch O, Stephens LR, Hawkins PT (2010) Phosphorylation of threonine 154 in p40phox is an important physiological signal for activation of the neutrophil NADPH oxidase. Blood 116:6027–6036. https://doi.org/10.1182/ blood-2010-08-300889 22. Didsbury J, Weber RF, Bokoch GM, Evans T, Snyderman R (1989) Rac, a novel ras-related family of proteins that are botulinum toxin substrates. J Biol Chem 264:16378–16382 23. Diebold BA, Bokoch GM (2001) Molecular basis for Rac2 regulation of phagocyte NADPH oxidase. Nat Immunol 2:211–215. https://doi.org/10.1038/85259 24. Kao YY, Gianni D, Bohl B, Taylor RM, Bokoch GM (2008) Identification of a conserved Rac-binding site on NADPH oxidases supports a direct GTPase regulatory mechanism. J Biol Chem 283: 12736–12746. https://doi.org/10.1074/jbc.M801010200 25. Nauseef WM (2004) Assembly of the phagocyte NADPH oxidase. Histochem Cell Biol 122(4):277–291. https://doi.org/10.1007/ s00418-004-0679-8 26. Grogan A, Reeves E, Keep N, Wientjes F, Totty NF, Burlingame AL, Hsuan JJ, Segal AW (1997) Cytosolic phox proteins interact with and regulate the assembly of coronin in neutrophils. J Cell Sci 110(Pt 24):3071–3081 27. Shao D, Segal AW, Dekker LV (2010) Subcellular localisation of the p40phox component of NADPH oxidase involves direct interactions between the Phox homology domain and F-actin. Int J Biochem Cell Biol 42:1736–1743. https://doi.org/10.1016/j.biocel. 2010.07.009 28. Wientjes FB, Reeves EP, Soskic V, Furthmayr H, Segal AW (2001) The NADPH oxidase components p47(phox) and p40(phox) bind to moesin through their PX domain. Biochem Biophys Res Commun 289:382–388. https://doi.org/10.1006/bbrc.2001.5982 29. Zhan Y, He D, Newburger PE, Zhou GW (2004) p47(phox) PX domain of NADPH oxidase targets cell membrane via moesinmediated association with the actin cytoskeleton. J Cell Biochem 92:795–809. https://doi.org/10.1002/jcb.20084 30. Nauseef WM, Volpp BD, McCormick S, Leidal KG, Clark RA (1991) Assembly of the neutrophil respiratory burst oxidase. Protein kinase C promotes cytoskeletal and membrane association of cytosolic oxidase components. J Biol Chem 266:5911–5917 31. Woodman RC, Ruedi JM, Jesaitis AJ, Okamura N, Quinn MT, Smith RM, Curnutte JT, Babior BM (1991) Respiratory burst oxidase and three of four oxidase-related polypeptides are associated with the cytoskeleton of human neutrophils. J Clin Invest 87:1345– 1351. https://doi.org/10.1172/JCI115138

17

p40phox: Composition, Function and Consequences of Its Absence

32. El Benna J, Ruedi JM, Babior BM (1994) Cytosolic guanine nucleotide-binding protein Rac2 operates in vivo as a component of the neutrophil respiratory burst oxidase. Transfer of Rac2 and the cytosolic oxidase components p47phox and p67phox to the submembranous actin cytoskeleton during oxidase activation. J Biol Chem 269:6729–6734 33. Someya A, Nagaoka I, Nunoi H, Yamashita T (1996) Translocation of guinea pig p40-phox during activation of NADPH oxidase. Biochim Biophys Acta 1277:217–225. https://doi.org/10.1016/ s0005-2728(96)00099-0 34. El Benna J, Dang PM, Andrieu V, Vergnaud S, Dewas C, Cachia O, Fay M, Morel F, Chollet-Martin S, Hakim J, Gougerot-Pocidalo MA (1999) P40phox associates with the neutrophil Triton X-100-insoluble cytoskeletal fraction and PMA-activated membrane skeleton: a comparative study with P67phox and P47phox. J Leukoc Biol 66: 1014–1020. https://doi.org/10.1002/jlb.66.6.1014 35. Ziegler CS, Bouchab L, Tramier M, Durand D, Fieschi F, DupréCrochet S, Mérola F, Nüße O, Erard M (2019) Quantitative live-cell imaging and 3D modeling reveal critical functional features in the cytosolic complex of phagocyte NADPH oxidase. J Biol Chem 294:3824–3836. https://doi.org/10.1074/jbc.RA118.006864 36. Ellson C, Davidson K, Anderson K, Stephens LR, Hawkins PT (2006) PtdIns3P binding to the PX domain of p40phox is a physiological signal in NADPH oxidase activation. EMBO J 25:4468– 4478. https://doi.org/10.1038/sj.emboj.7601346 37. Bissonnette SA, Glazier CM, Stewart MQ, Brown GE, Ellson CD, Yaffe MB (2008) Phosphatidylinositol 3-phosphate-dependent and -independent functions of p40phox in activation of the neutrophil NADPH oxidase. J Biol Chem 283:2108–2119. https://doi.org/10. 1074/jbc.M706639200 38. Dusi S, Donini M, Rossi F (1996) Mechanisms of NADPH oxidase activation: translocation of p40phox, Rac1 and Rac2 from the cytosol to the membranes in human neutrophils lacking p47phox or p67phox. Biochem J 314(Pt 2):409–412. https://doi.org/10.1042/ bj3140409 39. Tian W, Li XJ, Stull ND, Ming W, Suh CI, Bissonnette SA, Yaffe MB, Grinstein S, Atkinson SJ, Dinauer MC (2008) Fc gamma R-stimulated activation of the NADPH oxidase: phosphoinositidebinding protein p40phox regulates NADPH oxidase activity after enzyme assembly on the phagosome. Blood 112:3867–3877. https:// doi.org/10.1182/blood-2007-11-126029 40. Suh CI, Stull ND, Li XJ, Tian W, Price MO, Grinstein S, Yaffe MB, Atkinson S, Dinauer MC (2006) The phosphoinositide-binding protein p40phox activates the NADPH oxidase during FcgammaIIA receptor-induced phagocytosis. J Exp Med 203:1915–1925. https:// doi.org/10.1084/jem.20052085 41. Matute JD, Arias AA, Wright NA, Wrobel I, Waterhouse CC, Li XJ, Marchal CC, Stull ND, Lewis DB, Steele MG, Kellner JD, Yu W, Meroueh SO, Nauseef WM, Dinauer MC (2009) A new genetic subgroup of chronic granulomatous disease with autosomal recessive mutations in p40 phox and selective defects in neutrophil NADPH oxidase activity. Blood 114:3309–3315. https://doi.org/ 10.1182/blood-2009-07-231498 42. Ellson CD, Davidson K, Ferguson GJ, O'Connor R, Stephens LR, Hawkins PT (2006) Neutrophils from p40phox-/- mice exhibit severe defects in NADPH oxidase regulation and oxidant-dependent bacterial killing. J Exp Med 203:1927–1937. https://doi.org/10. 1084/jem.20052069 43. Anderson KE, Boyle KB, Davidson K, Chessa TA, Kulkarni S, Jarvis GE, Sindrilaru A, Scharffetter-Kochanek K, Rausch O, Stephens LR, Hawkins PT (2008) CD18-dependent activation of the neutrophil NADPH oxidase during phagocytosis of Escherichia coli or Staphylococcus aureus is regulated by class III but not class I or II PI3Ks. Blood 112:5202–5211. https://doi.org/10.1182/blood2008-04-149450

283 44. Winter S, Hopkins MH, Laulund F, Holmdahl R (2016) A reduction in intracellular reactive oxygen species due to a mutation in NCF4 promotes autoimmune arthritis in mice. Antioxid Redox Signal 25: 983–996. https://doi.org/10.1089/ars.2016.6675 45. Bagaitkar J, Barbu EA, Perez-Zapata LJ, Austin A, Huang G, Pallat S, Dinauer MC (2017) PI(3)P-p40phox binding regulates NADPH oxidase activation in mouse macrophages and magnitude of inflammatory responses in vivo. J Leukoc Biol 101:449–457. https://doi.org/10.1189/jlb.3AB0316-139R 46. Han CH, Freeman JL, Lee T, Motalebi SA, Lambeth JD (1998) Regulation of the neutrophil respiratory burst oxidase. Identification of an activation domain in p67(phox). J Biol Chem 273:16663– 16668. https://doi.org/10.1074/jbc.273.27.16663 47. Lopes LR, Dagher MC, Gutierrez A, Young B, Bouin AP, Fuchs A, Babior BM (2004) Phosphorylated p40PHOX as a negative regulator of NADPH oxidase. Biochemistry 43:3723–3730. https://doi. org/10.1021/bi035636s 48. Sathyamoorthy M, de Mendez I, Adams AG, Leto TL (1997) p40 (phox) down-regulates NADPH oxidase activity through interactions with its SH3 domain. J Biol Chem 272:9141–9146. https://doi.org/10.1074/jbc.272.14.9141 49. Van den Berg JM, van Koppen E, Ahlin A, Belohradsky BH, Bernatowska E, Corbeel L, Español T, Fischer A, KurenkoDeptuch M, Mouy R, Petropoulou T, Roesler J, Seger R, Stasia MJ, Valerius NH, Weening RS, Wolach B, Roos D, Kuijpers TW (2009) Chronic granulomatous disease: the European experience. PLoS One 4:e5234. https://doi.org/10.1371/journal.pone.0005234 50. Roos D (2016) Chronic granulomatous disease. Br Med Bull 118: 50–63. https://doi.org/10.1093/bmb/ldw009 51. Harris J, Hartman M, Roche C, Zeng SG, O'Shea A, Sharp FA, Lambe EM, Creagh EM, Golenbock DT, Tschopp J, Kornfeld H, Fitzgerald KA, Lavelle EC (2011) Autophagy controls IL-1beta secretion by targeting pro-IL-1beta for degradation. J Biol Chem 286:9587–9597. https://doi.org/10.1074/jbc.M110.202911 52. Virgin HW, Levine B (2009) Autophagy genes in immunity. Nat Immunol 10:461–470 53. Patel KK, Miyoshi H, Beatty WL, Head RD, Malvin NP, Cadwell K, Guan JL, Saitoh T, Akira S, Seglen PO, Dinauer MC, Virgin HW, Stappenbeck TS (2013) Autophagy proteins control goblet cell function by potentiating reactive oxygen species production. EMBO J 32:3130–3144. https://doi.org/10.1038/emboj.2013.233 54. Aarts CE, Hiemstra IH, Béguin EP, Hoogendijk AJ, Bouchmal S, van Houdt M, Tool AT, Mul E, Jansen MH, Janssen H, van Alphen FP, de Boer JP, Zuur CL, Meijer AB, van den Berg TK, Kuijpers TW (2019) Activated neutrophils exert myeloid-derived suppressor cell activity damaging T cells beyond repair. Blood Adv 3:3562– 3574. https://doi.org/10.1182/bloodadvances.2019031609 55. Segal BH, Han W, Bushey JJ, Joo M, Bhatti Z, Feminella J, Dennis CG, Vethanayagam RR, Yull FE, Capitano M, Wallace PK, Minderman H, Christman JW, Sporn MB, Chan J, Vinh DC, Holland SM, Romani LR, Gaffen SL, Freeman ML, Blackwell TS (2010) NADPH oxidase limits innate immune responses in the lungs in mice. PLoS One 5:e9631. https://doi.org/10.1371/journal. pone.0009631 56. Conway KL, Goel G, Sokol H, Manocha M, Mizoguchi E, Terhorst C, Bhan AK, Gardet A, Xavier RJ (2012) P40phox expression regulates neutrophil recruitment and function during the resolution phase of intestinal inflammation. J Immunol 189:3631–3640. https://doi.org/10.4049/jimmunol.1103746 57. Lee PL, West C, Crain K, Wang L (2006) Genetic polymorphisms and susceptibility to lung disease. J Negat Results Biomed 5:5. https://doi.org/10.1186/1477-5751-5-5 58. Olsson LM, Lindqvist AK, Källberg H, Padyukov L, Burkhardt H, Alfredsson L, Klareskog L, Holmdahl R (2007) A case-control study of rheumatoid arthritis identifies an associated single nucleotide

284 polymorphism in the NCF4 gene, supporting a role for the NADPHoxidase complex in autoimmunity. Arthritis Res Ther 9:R98. https:// doi.org/10.1186/ar2299 59. Denson LA, Jurickova I, Karns R, Shaw KA, Cutler DJ, Okou DT, Dodd A, Quinn K, Mondal K, Aronow BJ, Haberman Y, Linn A, Price A, Bezold R, Lake K, Jackson K, Walters TD, Griffiths A, Baldassano RN, Noe JD, Hyams JS, Crandall WV, Kirschner BS, Heyman MB, Snapper S, Guthery SL, Dubinsky MC, Leleiko NS, Otley AR, Xavier RJ, Stevens C, Daly MJ, Zwick ME, Kugathasan S (2018) Clinical and genomic correlates of neutrophil reactive oxygen species production in pediatric patients with Crohn’s disease. Gastroenterology 154:2097–2110. https://doi.org/10.1053/j. gastro.2018.02.016 60. Anderson KE, Chessa TA, Davidson K, Henderson RB, Walker S, Tolmachova T, Grys K, Rausch O, Seabra MC, Tybulewicz VL, Stephens LR, Hawkins PT (2010) PtdIns3P and Rac direct the assembly of the NADPH oxidase on a novel, pre-phagosomal compartment during FcR-mediated phagocytosis in primary mouse neutrophils. Blood 116:4978–4989. https://doi.org/10.1182/blood2010-03-275602 61. Bagaitkar J, Matute JD, Austin A, Arias AA, Dinauer MC (2012) Activation of neutrophil respiratory burst by fungal particles requires phosphatidylinositol 3-phosphate binding to p40(phox) in humans but not in mice. Blood 120:3385–3387. https://doi.org/10. 1182/blood-2012-07-445619 62. Winkelstein JA, Marino MC, Johnston RB Jr, Boyle J, Curnutte J, Gallin JI, Malech HL, Holland SM, Ochs H, Quie P, Buckley RH, Foster CB, Chanock SJ, Dickler H (2000) Chronic granulomatous disease. Report on a national registry of 368 patients. Medicine (Baltimore) 79:155–169. https://doi.org/10.1097/00005792200005000-00003 63. Wolach B, Gavrieli R, de Boer M, van Leeuwen K, Berger-AchituvS, Stauber T, Ben Ari J, Rottem M, Schlesinger Y, Grisaru-Soen G, Abuzaitoun O, Marcus N, Zion Garty B, Broides A, Levy J, Stepansky P, Etzioni A, Somech R, Roos D (2017) Chronic granulomatous disease: clinical, functional, molecular, and genetic studies. The Israeli experience with 84 patients. Am J Hematol 92:28– 36. https://doi.org/10.1002/ajh.24573 64. Kuhns DB, Alvord WG, Heller T, Feld JJ, Pike KM, Marciano BE, Uzel G, DeRavin SS, Priel DA, Soule BP, Zarember KA, Malech HL, Holland SM, Gallin JI (2010) Residual NADPH oxidase and survival in chronic granulomatous disease. N Engl J Med 363:2600– 2610. https://doi.org/10.1056/NEJMoa1007097 65. Köker MY, Camc{oğlu Y, van Leeuwen K, K{l{ç SŞ, Barlan I, Y{lmaz M, Metin A, de Boer M, Avc{lar H, Pat{roğlu T, Y{ld{ran A, Yeğin O, Tezcan I, Sanal Ö, Roos D (2013) Clinical, functional, and genetic characterization of chronic granulomatous disease in 89 Turkish patients. J Allergy Clin Immunol 132:1156– 1163.e5. https://doi.org/10.1016/j.jaci.2013.05.039 66. Gazendam RP, van Hamme JL, Tool AT, van Houdt M, Verkuijlen PJ, Herbst M, Liese JG, van de Veerdonk FL, Roos D, van den Berg TK, Kuijpers TW (2014) Two independent killing mechanisms of Candida albicans by human neutrophils: evidence from innate immunity defects. Blood 124:590–597. https://doi.org/10.1182/ blood-2014-01-551473 67. Gazendam RP, van Hamme JL, Tool AT, Hoogenboezem M, van den Berg JM, Prins JM, Vitkov L, van de Veerdonk FL, van den Berg TK, Roos D, Kuijpers TW (2016) Human neutrophils use different mechanisms to kill aspergillus fumigatus conidia and hyphae: evidence from phagocyte defects. J Immunol 196:1272– 1283. https://doi.org/10.4049/jimmunol.1501811 68. Rosen H, Michel BR (1997) Redundant contribution of myeloperoxidase-dependent systems to neutrophil-mediated killing of Escherichia coli. Infect Immun 65:4173–4178

T. W. Kuijpers and D. Roos 69. Zhao XW, Gazendam RP, Drewniak A, van Houdt M, Tool AT, van Hamme JL, Kustiawan I, Meijer AB, Janssen H, Russell DG, van de Corput L, Tesselaar K, Boelens JJ, Kuhnle I, Der Werff V, Ten Bosch J, Kuijpers TW, van den Berg TK (2013) Defects in neutrophil granule mobilization and bactericidal activity in familial hemophagocytic lymphohistiocytosis type 5 (FHL-5) syndrome caused by STXBP2/Munc18-2 mutations. Blood 122:109–111 70. Magnani A, Brosselin P, Beauté J, de Vergnes N, Mouy R, Debré M, Suarez F, Hermine O, Lortholary O, Blanche S, Fischer A, Mahlaoui N (2014) Inflammatory manifestations in a single-center cohort of patients with chronic granulomatous disease. J Allergy Clin Immunol 134:655–662.e8. https://doi.org/10.1016/j.jaci.2014. 04.014 71. Fernandez-Boyanapalli RF, Frasch SC, McPhillips K, Vandivier RW, Harry BL, Riches DW, Henson PM, Bratton DL (2009) Impaired apoptotic cell clearance in CGD due to altered macrophage programming is reversed by phosphatidylserine-dependent production of IL-4. Blood 113:2047–2055 72. Greenlee-Wacker MC, Rigby KM, Kobayashi SD, Porter AR, DeLeo FR, Nauseef WM (2014) Phagocytosis of Staphylococcus aureus by human neutrophils prevents macrophage efferocytosis and induces programmed necrosis. J Immunol 192:4709–4717 73. Fernandez-Boyanapalli R, McPhillips KA, Frasch SC, Janssen WJ, Dinauer MC, Riches DW, Henson PM, Byrne A, Bratton DL (2010) Impaired phagocytosis of apoptotic cells by macrophages in chronic granulomatous disease is reversed by IFN-gamma in a nitric oxidedependent manner. J Immunol 185:4030–4041 74. Emmerson A, Trevelin SC, Mongue-Din H, Becker PD, Ortiz C, Smyth LA, Peng Q, Elgueta R, Sawyer G, Ivetic A, Lechler RI, Lombardi G, Shah AM (2018) Nox2 in regulatory T cells promotes angiotensin II–induced cardiovascular remodeling. J Clin Invest 128:3088–3101. https://doi.org/10.1172/JCI97490 75. van de Geer A, Cuadrado E, Slot MC, van Bruggen R, Amsen D, Kuijpers TW (2019) Regulatory T cell features in chronic granulomatous disease. Clin Exp Immunol 197:222–229. https://doi.org/10. 1111/cei.13300 76. Crotzer VL, Matute JD, Arias AA, Zhao H, Quilliam LA, Dinauer MC, Blum JS (2012) Cutting edge: NADPH oxidase modulates MHC class II antigen presentation by B cells. J Immunol 189: 3800–3804 77. Kuhns DB, Spalding C, Garofalo M, Dimaggio T, Estwick T, Huang CY, Fink D, Priel DL, Fleisher TA, Holland SM, Malech HL, Gallin JI (2013) B-cell activating factor (BAFF) is elevated in chronic granulomatous disease. Clin Immunol 148:258–264. https://doi. org/10.1016/j.clim.2013.05.007 78. Cotugno N, Finocchi A, Cagigi A, Di Matteo G, Chiriaco M, Di Cesare S, Rossi P, Aiuti A, Palma P, Douagi I (2015) Defective B-cell proliferation and maintenance of long-term memory in patients with chronic granulomatous disease. J Allergy Clin Immunol 135:753–761.e2. https://doi.org/10.1016/j.jaci.2014. 07.012 79. Cachat J, Deffert C, Alessandrini M, Roux-Lombard P, Le Gouellec A, Stasia MJ, Hugues S, Krause KH (2018) Altered humoral immune responses and IgG subtypes in NOX2-deficient mice and patients: a key role for NOX2 in antigen-presenting cells. Front Immunol 9:1555. https://doi.org/10.3389/fimmu.2018.01555 80. To EE, Vlahos R, Luong R, Halls ML, Reading PC, King PT, Chan C, Drummond GR, Sobey CG, Broughton BRS, Starkey MR, van der Sluis R, Lewin SR, Bozinovski S, O'Neill LAJ, Quach T, Porter CJH, Brooks DA, O'Leary JJ, Selemidis S (2017) Endosomal NOX2 oxidase exacerbates virus pathogenicity and is a target for antiviral therapy. Nat Commun 8:69. https://doi.org/10. 1038/s41467-017-00057-x

17

p40phox: Composition, Function and Consequences of Its Absence

81. Souyris M, Cenac C, Azar P, Daviaud D, Canivet A, Grunenwald S, Pienkowski C, Chaumeil J, Mejía JE, Guéry JC (2018) TLR7 escapes X chromosome inactivation in immune cells. Sci Immunol 3:eaap8855. https://doi.org/10.1126/sciimmunol.aap8855 82. Battersby AC, Braggins H, Pearce MS, Cale CM, Burns SO, Hackett S, Hughes S, Barge D, Goldblatt D, Gennery AR (2017) Inflammatory and autoimmune manifestations in X-linked carriers of chronic granulomatous disease in the United Kingdom. J Allergy Clin Immunol 140(2):628–630.e6. https://doi.org/10.1016/j.jaci. 2017.02.029

285 83. Marciano BE, Zerbe CS, Falcone EL, Ding L, DeRavin SS, Daub J, Kreuzburg S, Yockey L, Hunsberger S, Foruraghi L, Barnhart LA, Matharu K, Anderson V, Darnell DN, Frein C, Fink DL, Lau KP, Long Priel DA, Gallin JI, Malech HL, Uzel G, Freeman AF, Kuhns DB, Rosenzweig SD, Holland SM (2018) X-linked carriers of chronic granulomatous disease: illness, lyonization, and stability. J Allergy Clin Immunol 141:365–371. https://doi.org/10.1016/j.jaci. 2017.04.035

Rho Family GTPases and their Modulators

18

Yuan Lin and Yi Zheng

Abstract

The Rho family GTPases play an important role in mediating signal transduction in multiple fundamental cellular processes. Most of the Rho family members act as molecular switches and cycle between an inactive GDP-bound state and an active GTP-bound state. Upon GTP binding, Rho GTPases undergo conformational changes to interact with a wide variety of downstream effectors. The dynamic cycling of GTP loading/GTP hydrolysis is key for Rho GTPase functions and is tightly regulated by guanine nucleotide exchange factors, GTPase-activating proteins, GDP-dissociation inhibitors, as well as additional posttranslational modifications. Dysregulation of Rho GTPase signaling pathways is involved in multiple human pathological conditions including cancer, inflammation, and cardiovascular diseases. Among the Rho family members, Rac1 and Rac2 are crucial for the assembly and activation of NADPH oxidases NOX1 and NOX2. This chapter provides an overview of the structural mechanisms of the regulation and function of Rho GTPases, with an emphasis on Rac1 and Rac2 where they apply. Keywords

Rho GTPases · Rac · Signaling effectors · NOX2 · Rho GEFs · Rho GAPs · Rho GDIs

1

Introduction

The NOX family of NADPH oxidases are key producers of the reactive oxygen species (ROS) in many cells and mediate diverse functions through redox signaling. The seven Y. Lin · Y. Zheng (✉) Experimental Hematology and Cancer Biology, Cincinnati Children’s Hospital Medical Center, University of Cincinnati, Cincinnati, OH, USA e-mail: [email protected]; [email protected]

members of the NOX family, varying in tissue distribution and activation mechanisms, have been discussed in detail in other chapters (see Chap. 2 by J.T. Curnutte and A.I. Tauber, Chap. 4 by A.W. Segal, Chap. 9 by A. van der Vliet, Chap. 10 by T.L. Leto and M. Geiszt, Chap. 11 by Y. Nakano and B. Bánfi, Chap. 12 by L. Hecker et al., Chap. 13 by L.L. Camargo et al., and Chap. 14 by F. Miot and X. De Deken). Among them, NOX1 and NOX2 are parts of a multi-subunit enzyme complex, comprising both membrane and cytosolic components. Both NOX1 and NOX2 depend on the Rho GTPase Rac1 or Rac2 to respond to a wide range of stimuli in mediating their assembly and activation. Understanding the regulatory mechanism of the NOX enzymes is important in modulating or maintaining desired ROS levels produced by NOX. This chapter summarizes the structure and function of Rho GTPases and related regulators/effectors with an emphasis on Rac-NOX2 signaling and discusses the relevance of their dysregulation in human diseases.

2

Family of Rho GTPases: An Overview

The Rho GTPase family belongs to the Ras superfamily, which encompasses more than 150 members in mammals [1]. The founding member of the Rho family, named Rho for “Ras homolog”, was identified from a cDNA library from the abdominal ganglia of Aplysia in 1985 [2]. Rac (Ras-related C3 botulinum toxin substrate) 1 and 2 were subsequently isolated as substrates for ADP-ribosylation by the toxin [3]. Rho family GTPases have since expanded to include 20 members divided into eight subfamilies: Rho, Rac, Cdc42, Rnd, RhoD/RhoF, RhoU/RhoV, RhoBTB, and RhoH [4] (Fig. 18.1a). Rho GTPases comprise a core G-domain conserved in the Ras superfamily for GDP/GTPbinding and GTP-hydrolysis activity and a Rho-specific insert involved in the binding to effectors and regulators [7] (Fig. 18.1b). Some Rho members have N-terminal extensions, although the function of those regions remains

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_18

287

288

Fig. 18.1 Phylogenetic tree, G-domain structure, and sequence alignment of Rho GTPases. Protein sequences of the 20 human Rho GTPases are extracted from GenBank. The splice variants of Rac1 and Cdc42 are not included. Alignment was performed using Clustal Omega [5] and viewed by Jalview [6]. (a). The neighbor-joining phylogenetic tree is generated for Rho subfamily members by Clustal Omega. (b). A

Y. Lin and Y. Zheng

3-dimensional G-domain structure is represented by Rac1-GDP (PDB 5N6O). The bound GDP is shown in stick and the Mg2+ ion is shown in sphere. (c). Sequence alignment of the 20 Rho GTPases. The sequences of the C-termini of RhoBTB1 and RhoBTB2 are hidden (blue line). The conserved motifs are highlighted in colors, including Switch I/II, P-loop, N/TKxD, ExSAK, and the Rho-specific insert region. Three residues,

18

Rho Family GTPases and their Modulators

largely unclear (Fig. 18.1c). Rho GTPases are cytosolic proteins but most of them are associated with membranes via lipid modification of the CAAX (where C is cysteine, A is any aliphatic amino acid, and X is any amino acid) motif located in the C-terminal hypervariable region. Adjacent to the CAAX motif, many Rho GTPases have a polybasic region with multiple roles including membrane association and subcellular localization [8]. The most characterized Rho GTPases are RhoA, Rac1, and Cdc42, which are ubiquitously expressed. These proteins and their subfamilies are also known as classical or typical Rho GTPases as they behave similarly to Ras and transition between an inactive GDP-bound conformation and an active GTP-bound conformation. Once activated, Rho GTPases relocate to the plasma membrane (PM) or other cellular compartments to interact with downstream effectors. This “molecular switch” scheme was initially considered to encompass all Rho GTPases; however, the identification of Rnd1, Rnd2, and Rnd3 (also known as RhoE) reveals the existence of atypical Rho members [9, 10]. Rnd proteins, as well as RhoH [11] and RhoBTB1/2 [12], are GTPasedeficient or have limited ability to hydrolyze GTP [13]. RhoU (also known as Wrch-1), RhoV (also known as Wrch-2/Chp), RhoD, and RhoF (also known as Rif) constitute another category of atypical Rho GTPases with significantly elevated intrinsic GDP/GTP exchange activity [14]. Apart from the atypical Rho GTPases, Rac1b, an alternative splice variant of Rac1, is fast-cycling by itself due to a 19-amino acid insertion [15]. Rho family members also differ in their expression level and pattern in tissue cell types. For example, the Rac subfamily consists of four members, i.e., Rac1, Rac2, Rac3, and RhoG. Rac1 is ubiquitously expressed and is involved in several fundamental cellular functions. Rac1-knockout mice are embryonic lethal [16]. In contrast, Rac2 expression is restricted to the hematopoietic system, whereas Rac3 expression is found predominantly in the brain and testis [17]. RhoG, which shares the lowest sequence similarity with Rac1, has the highest level of expression in the brain and hematopoietic system. The respective Rac2-, Rac3-, or RhoG-knockout mice are vital and display only cell-type specific deficiencies [18–20]. Regulation of Rho GTPases at the level of gene expression or post-transcriptionally (for example, by microRNAs) is beyond the scope of this chapter and is described in excellent reviews [21, 22]. Rho family proteins are found in all eukaryotes and play essential roles in many cellular processes [23]. The variations

289

in structure and subcellular localization of Rho GTPases allow them to interact with a large number of upstream regulators and downstream targets. Rho GTPases are master modulators of actin cytoskeleton rearrangements and are crucial for cell motility and migration, cell morphology and polarity, vesical trafficking and secretion [24]. Through a dynamic network of regulators and effectors, Rho GTPases have also been implicated in signaling pathways for gene transcription, cell cycle progression, cell survival, cell growth, and differentiation [25]. These signaling pathways include the JNK/p38/MAPK cascades, PI3K pathway, NFκB pathway, serum response factors (SRF) pathway, and the assembly of NADPH oxidase complexes.

3

Regulation of Rho GTPases

The cycling of classical Rho GTPases is tightly regulated (Fig. 18.2). Guanine nucleotide exchange factors (GEFs) function as Rho protein activators by catalyzing the exchange of GDP for GTP, whereas GTPase activating proteins (GAPs) facilitate hydrolysis of GTP, leading to Rho protein inactivation. A third family of regulators, the guaninenucleotide dissociation inhibitors (GDIs), binds to the inactive pool of Rho GTPases and sequesters them in the cytosol [26]. Membrane association, achieved mainly by prenylation at the C-termini, is crucial for most Rho GTPases activation. GDIs recognize prenylated Rho proteins and control their cycling between the cytosol and membrane. In addition to prenylation, Rho GTPases possess a wide range of posttranslational modifications, providing additional regulation mechanisms.

3.1

Guanine Nucleotide Exchange Factors (GEFs)

Most Rho proteins have high binding affinity for GDP and GTP. The intrinsic rate of nucleotide release is typically slow and needs the catalysis by GEFs for efficient GTP loading and signaling. GEFs selectively bind to Rho GTPases and stimulate nucleotide exchange in a multistep process [26]. The GEF first forms a low-affinity complex with the GDP-bound GTPase and has a relatively high affinity for the nucleotide-free GTPase (Fig. 18.3a). Binding of free GTP in the cytosol replaces the GEF, resulting in GTP-bound Rho proteins. Rho GEFs are classified into two families in

ä

Fig. 18.1 (Continued) marked with asterisk (*) and highlighted in orange, are highly conserved. T35 and G60 (numbering in Rac1, same in below) interact with the g-phosphate when GTP is bound, and Q61 is the catalytic residue for GTP hydrolysis. Residue 61 is not the conserved and is glutamine in Rnd, RhoBTB, and RhoH subfamilies, accounting

for their inability to hydrolyze GTP. The cysteine residues in the C-terminal CAAX box are highlighted in yellow and they are the sites for lipid modifications. Lipid modifications are summarized. GG: geranylgeranylation; F: farnesylation; P: palmitoylation

290

Y. Lin and Y. Zheng

Fig. 18.2 An overview of Rho GTPase signaling regulation. Rho GTPases are tightly regulated in response to various upstream signals. GEFs, GAPs, and GDIs constitute the classic regulatory protein families that regulate Rho GTPase cycle. Rho GEFs promote the exchange of GDP for GTP and activate Rho GTPases, GAPs stimulate the usually slow intrinsic GTPase activity and inactivate Rho signaling, and GDIs

sequester the prenylated Rho GTPases in the cytosol and protect them from degradation. There are a wide range of posttranslational modifications on Rho GTPases, including lipid modification (prenylation and palmitoylation), phosphorylation, SUMOylation, and ubiquitination, that affect Rho GTPase signaling function

metazoans, Dbl (diffuse B-cell lymphoma) and DOCK (dedicator of cytokinesis) families, according to their structures. There are currently at least 85 Rho GEF members in human tissues, including 74 Dbl and 11 DOCK family members. Members of the Dbl family are characterized by a unique Dbl homology (DH) domain, which is almost invariably followed by a pleckstrin homology (PH) domain with a few exceptions [27]. The DH domain comprises the catalytic domain, whereas the PH domain has diverse functions [28]. Flanking the DH-PH domain, additional N-terminal and C-terminal motifs or domains confer autoregulation, subcellular localization, and responding to upstream signals [29–31]. The large number of Dbl family members strongly suggests that a specific Rho GTPase can be activated by several GEFs to potentially respond to different signals. Several Dbl members are considered to be selective towards Rac, such as T-cell lymphoma invasion and metastasis (Tiam), phosphatidylinositol 3,4,5-triphosphate [PtdIns

(3,4,5)P3]-dependent Rac exchanger (P-Rex), and the hemopoietic-specific Vav [27]. Tiam family members are generally activated by Ras in response to the receptor tyrosine kinase (RTK) [32]. They translocate to the PM by binding to PtdIns(3,4,5)P3 with a second PH domain at the N-terminus [33]. P-Rex proteins, i.e., P-Rex1, 2, and 2b, bind to PtdIns(3,4,5)P3 with their PH domain. P-Rex1 is highly expressed in leukocytes and neurons and activates Rac in response to G protein-coupled receptors (GPCRs), E-selectin, and toll-like receptor 4 (TLR4) in mouse neutrophils [34]. The Vav proteins are activated via phosphorylation by protein tyrosine kinases downstream of various receptors, including integrins, Fc receptors (FcR), GPCRs, and TLR4 [35]. They can also act as adaptor proteins and function in specific signaling contexts. DOCK proteins are comprised of two conserved domains: a DOCK homology region-1 (DHR1) domain that locates them to membranes [36–38] and a DHR2 domain for

18

Rho Family GTPases and their Modulators

291

Fig. 18.3 Representative crystal structures of Rho GTPases in complex with regulators. (a). Complex of Rac1 with Dbl-family or DOCK family Rho GEFs: Rac1-Tiam1 (PDB 1FOE) and Rac1-DOCK2 (PDB 2YIN). Rac1 is nucleotide-free in both complexes. (b). Complex of RhoA with RhoGAP: RhoA-p50RhoGAP (PDB 1TX4). RhoA is in complex with

GDP
AlF4, and Mg2+, where GDP
ALF4 is a transition-state analog. (c). Complex of Rac1 with GDI: Rac1-RhoGDI1 (PDB 1HH4). Rac1 is in complex with GDP and Mg2+. The geranylgeranyl group is shown in stick in orange. In all complexes, Rho GTPase is colored in grey with the switch regions highlighted in magenta

nucleotide exchange [39]. The DOCK proteins are classified into four subfamilies based on sequence similarity—DOCKA (DOCK1, 2, and 5), DOCK-B (DOCK3 and 4), DOCK-C (DOCK6, 7, and 8) and DOCK-D (DOCK9, 10 and 11). The DOCK family GEFs are specific for Rac and/or Cdc42, but not other Rho proteins [40]. DOCK-A/B exclusively activates Rac, whereas DOCK-C/D predominantly activates Cdc42, and DOCK6, 7, and 10 activate both. DOCK proteins play various roles in developmental processes and in the immune system [41–43]. Among them, DOCK2 is restricted to hematopoietic cells and is a Rac-specific activator critical for leukocytes trafficking, immunological synapse formation, neutrophil chemotaxis, and ROS production [44].

hydrolysis is often necessary [45]. The Rho GAP family contains a conserved catalytic GAP domain that binds to Rho GTPases and stimulates their intrinsic GTPase activity (Fig. 18.3b). The GAP domain contains a conserved arginine residue, called “arginine finger”, that inserts into the GTP-binding site of the respective Rho protein to stabilize negative charges at the transition state to catalyze the GTPase reaction [46]. This catalytic strategy is shared with other GAPs for Ras superfamily GTPases including Ras, Rab, and Arf, although different GAP families do not have sequence or structural similarities [47]. Since the discovery of the first Rho GAP, p50RhoGAP [48], more than 66 distinct human Rho GAP proteins have been identified [49]. Unlike Rho GEF domains (DH domain or DHR2 domain) which exhibit high selectivity for the Rho, Rac, or Cdc42 subfamily proteins, the Rho GAP domain itself has limited selectivity under cell-free conditions. Instead, substrate selectivity of Rho GAP proteins is assisted by domains and motifs outside the GAP domain. These domains vary in their functions and can be classified into: A) lipid- or membrane-binding domains such as Bin-

3.2

GTPase Activating Proteins (GAPs)

The intrinsic GTP-hydrolysis reaction of Rho GTPases is typically slow. For timely termination of Rho signal transduction and to return to an inactive GDP-bound state, stimulation of the GTPase activity by GAPs to accelerate GTP

292

Y. Lin and Y. Zheng

Amphiphysin-Rvs (BAR), Fes/CIP4 homology-BAR (F-BAR), PH, and Sec14 domain; B) protein-interaction domains such as coiled-coils (CC), WW, and Src-homology (SH) 2/3 domain; and C) catalytic domains with enzyme activity such as DH domain [50]. Such motifs can also be posttranslational modified by phosphorylation or ubiquitination to regulate the GAP activity in response to signaling [4].

3.3

Guanine-Nucleotide Dissociation Inhibitors (GDIs)

Rho GDIs sequester the GDP-bound form of Rho GTPases in the cytosol and prevent them from localizing to the membranes or being activated by GEFs [51]. They also act as chaperones and target small GTPases to specific intracellular membranes, protecting them from degradation [52, 53]. In contrast to the large number of Rho GEFs and GAPs, there are only three Rho GDIs in mammals— RhoGDI1, RhoGDI2, and RhoGDI3. All three share similar domain structures; however, RhoGDI3 has a unique N-terminal extension and is found to localize in both cytoplasm and Golgi [53]. RhoGDI1 is widely expressed and interacts with several Rho GTPases, including RhoA, RhoC, Rac1–3, RhoG and Cdc42 [54]. RhoGDI2 is preferentially expressed in hematopoietic cells and seems to be less active than RhoGDI1 [55–57]. RhoGDI3 is primarily expressed in the brain and pancreas and is the least characterized member of the family. A recent comprehensive coimmunoprecipitation study of Rho GDI targets reveals that while RhoGDI1 and RhoGDI2 have relatively restricted activity towards classical Rho family members, RhoGDI3 displays a broad specificity in interacting with classical (Rho and Rac, but not Cdc42) and a number of atypical small Rho GTPases [58]. The interaction between Rho GDIs and Rho GTPase involves both the N-terminal regulatory domain that binds to Rho GTPases and inhibits their nucleotide exchange and the C-terminal immunoglobulin-like domain with a hydrophobic pocket that binds to the lipid group of Rho proteins, preventing their membrane association (Fig. 18.3c). An acidic patch in the Rho GDI lipid-binding pocket may contribute to compete with acidic phospholipids for binding to the basic tail of Rho GTPases. Activation of a specific Rho GTPase by a signaling pathway requires releasing from the Rho GDI. The dissociation process is not completely understood; however, evidence has been provided that Rho GDI interaction with Rho GTPases can be regulated by several mechanisms. Early work revealed that acidic lipids could disrupt Rac1-RhoGDI1 complex and activate the cell-free NADPH oxidase [59]. Subsequent studies used liposomes containing anionic phospholipids to promote Rac1-RhoGDI1

dissociation, suggesting that this process occurs at the level of the membrane [60, 61]. For Rac1-RhoGDI1 interaction, it is known that phosphorylation of RhoGDI1 on S174 and S101 by p21-activated kinase 1 (PAK1) could release Rac1 [62]. Both serine residues are located in the lipid-binding pocket and their phosphorylation likely drives dissociation of the complex. This effect is Rac1-specific, and thus phosphorylation of S174 and S101 may also affect the interaction of Rho GDI with the polybasic region of Rac1. In addition to lipid binding and posttranslational modifications on both RhoGDIs [4] and Rho GTPases (Sect. 3.5), protein interactions that compete for RhoGDIs also regulate specific Rho-RhoGDI interactions, such as 14–3-3τ [63], FRMD7 [64], TROY [65], and EphrinB1 [66].

3.4

Lipid Modification and the Polybasic Region Dictate Rho GTPase Subcellular Localization

The CAAX motif is found in most Rho family proteins. Geranylgeranyl or farnesyl modification at the cysteine residue of the CAAX motif, a process known as prenylation, is the first and most reported posttranslational modification of Rho GTPases [67]. The addition of either farnesyl or geranylgeranyl is mainly determined by the last residue of the CAAX sequence. Once an isoprenoid moiety is added to CAAX, a Rho protein is translocated to the endoplasmic reticulum, where the AAX tripeptide tail is cleaved and the prenylated cysteine undergoes carboxymethylation [68]. Additional lipid modification by a palmitoyl moiety has been reported for RhoB, and for the Cdc42-related proteins RhoQ (TC10), RhoJ (TCL), RhoU (Wrch-1), and RhoV (Wrch-2/Chp). Lipid modifications of the 20 Rho members are summarized in Fig. 18.1c. Besides prenylation and palmitoylation, the polybasic region of unique Rho GTPases also engages in specific protein-lipid or proteinprotein interactions for RhoGDI-mediated shuttling [8]. The majority of literature considers Rho GTPases to initiate signaling at or around the PM. Prenylation and membrane association of Rho proteins are essential for effector binding, GTP loading, and activation. For example, early work using reconstituted NADPH oxidase assays demonstrated the importance of prenylation, as well as the adjacent positively charged polybasic region in Rac1 and Rac2 [69, 70]. Rac1 contains more basic residues than Rac2 in the polybasic region (KKRKRK vs. RQQKRA), and this difference has been shown to contribute to their selective membrane recruitment depending on phospholipid composition in vitro [71] and to differential subcellular distribution and ability to promote ROS production in vivo in mouse neutrophils [72, 73]. In addition to the PM, it is becoming clear that functionally active Rho GTPases, their GEFs and GAPs,

18

Rho Family GTPases and their Modulators

and their effectors can be localized to different intracellular compartments [74]. Although predominantly residing at the PM, the classical Rho GTPases RhoA, Rac1, and Cdc42 have been seen at endosome (Rac1 [75] and Cdc42 [76]), Golgi (RhoA [77] and Cdc42 [78]), mitochondria (Rac1 [79]), and nucleus (RhoA [80], Rac1 [81], and Cdc42 [82]). Several other Rho family GTPases have been found to localize to the Golgi and endomembrane, including RhoB, RhoG, RhoQ, RhoJ, RhoU/RhoV, Rnd2, and RhoD. The underlying mechanism of Rho GTPases localizations beyond the PM remains to be fully understood [83]. The best studied is RhoB, for which the type of prenylation seems to be critical for its subcellular sorting. Geranylgeranylated RhoB is localized to late endosomes, while the farnesylated form is predominantly at the PM [84]. Notably, RhoB, RhoQ, RhoJ, RhoU, and RhoV are palmitoylated, highlighting the role of palmitoylation as an additional sorting signal. Rac1 has also been reported to incorporate a palmitic acid at C178, targeted for stable association at the actin cytoskeletonlinked, ordered membrane regions [85]. In the case of Cdc42, a brain-specific alternative splicing variant was reported as a palmitoylated protein [86]. This Cdc42 variant does not interact with RhoGDI and is therefore enriched at the PM. In contrast to prenylation, S-palmitoylation is reversible and provides weaker anchorage to the membrane, allowing the proteins to rapidly shuttle between intracellular compartments [87]. Rac1 also localizes to the nucleus and its polybasic region contains a nuclear localization signal (NLS) [88]. How Rac1 polybasic region promotes nucleus localization remains controversial. Prenylation, palmitoylation, polybasic regions, and their combinations, are likely key determinants for modulating Rho GTPase subcellular localizations and specific signaling function.

3.5

Other Posttranslational Modifications

In addition to lipid modification, a wide variety of posttranslational modifications have been reported for Rho GTPases, including phosphorylation, ubiquitination, SUMOylation, and AMPylation, that must be tightly controlled for proper signaling [4, 89]. Those modifications are particularly important for regulating atypical Rho GTPases, as these proteins are mostly in the active GTP-bound forms and lack the control of the GDP/GTP cycle. Phosphorylation on classical Rho GTPases is usually inhibitory and can modulate their interaction with guanine nucleotide, GDIs, GEFs, or effectors. For example, phosphorylation of S188 on RhoA by protein kinase A (PKA) or protein kinase G (PKG) [90–92] and phosphorylation of Cdc42 S185 by PKA [93] or Y64 by Src [94] enhance their respective binding to RhoGDI. Phosphorylation of Rac1 Y64 by Src and focal adhesion kinase (FAK) reduces GTP

293

binding, reduces association with its GEF β-PIX, enhances binding to RhoGDI, and reduces binding to the effector PAK [95]. Phosphorylation of RhoA S26 by mammalian sterile 20-like kinase 3 (Mst3) inhibits RhoA activity [96]. Phosphorylation of Rac1 S71 by AKT inhibits GTP binding [97] and mediates the interaction between Rac1 and 14–3-3 proteins [98]. Phosphorylation of Rac1 T108 by extracellular-signal regulated kinase (ERK) causes Rac1 to translocate to the nucleus, leading to decreased Rac1 activity [99]. Y34 and Y66 of RhoA can also be phosphorylated to negatively affect RhoA activity [100]. Ubiquitination is best characterized for RhoA and Rac1, although most Rho GTPases are likely subject to ubiquitination regulation. Several E3 ligase complexes have been identified, most of which target RhoA or Rac1 for proteasome degradation [101–105]. The role of ubiquitination as a modulator for Rho GTPase subcellular distribution and local activity within specific compartments remains largely unknown. In contrast to ubiquitination, the only known Rho GTPase substrate for SUMOylation is Rac1. SUMOylation of Rac1 within the polybasic region in the C-terminus enhances its activity [106]. AMPylation of Y34 of RhoA or Y32 of Rac1/Cdc42, mediated by FIC (filamentation induced by cyclic AMP) domain-containing effector proteins, abolishes effector binding and signaling of the respective Rho GTPase [107].

4

Effectors of Rho GTPases

The multi-facet signaling functions of Rho GTPases are mediated by downstream effector proteins. A wide range of effectors for Rho GTPases have been identified [108, 109]. More recent Rho GTPase interactome studies using proteomic approaches, such as quantitative GTPase affinity purification (qGAP) coupled to mass spectrometry [110] and proximity biotinylation (BioID) coupled to mass spectrometry [111], help provide a better understanding of the Rho GTPase effector portfolio. To date, more than 70 effector proteins for the classical Rho GTPases subfamilies, in particular RhoA, Rac1, and Cdc42, have been identified, whereas less is known about effectors for the atypical Rho family members [109]. A typical effector protein interacts specifically with the GTP-bound form of Rho GTPases and can be shared by more than one Rho GTPase family member.

4.1

Major Effectors for Rho, Rac, and Cdc42 Subfamilies

One essential function of Rho GTPases is to regulate the assembly and organization of the actin cytoskeleton

294

[112]. Not surprisingly, effectors related to actin cytoskeleton comprise a major group and are best characterized. Among Rho subfamily (RhoA, RhoB, RhoC) effectors, Rho kinases (ROCK) 1 and 2 regulate actin myosin contraction and stress fiber formation [113]; Citron kinase is required for cytokinesis [114]; mDia1–3 are formin-like proteins for actin filament nucleation/polymerization in cooperation with ROCK [115]. While mDia1 is only activated by Rho subfamily, mDia2 and mDia3 are also activated by Rac and Cdc42 [116]. Rac and Cdc42 shares some key effectors, such as p21-ativated kinases (PAKs), IQ motif-containing GTPaseactivating proteins (IQGAPs), insulin receptor substrate p53 (IRSp53), and partitioning-defective protein 6 (Par6). PAKs are prime regulators of actin cytoskeleton by phosphorylating multiple substrates that affect cytoskeleton structure including LIM kinase (LIMK), Arpc1b, Filamin A (FLNA), cortactin, and myosin light-chain kinase (MLCK) [117, 118]. IQGAPs do not have GAP activity but function as scaffolding proteins for regulating the cytoskeleton and cell-cell contacts [119]. IRSp53 is a scaffolding protein and regulator of membrane and actin dynamics [120]. Par6, in complex with Par3 and atypical PKC (aPKC), regulates cell polarity [121]. Additional effectors for regulating cytoskeleton organization include Wishkott-Aldrich Syndrome protein (WASP), neuronal WASP (N-WASP), myotonic dystrophyrelated Cdc42-binding kinase (MRCK), formin-like protein FMNL1 and 2, inverted formin 2 (INF2) downstream of Cdc42; and WASP-family verprolin-homologous protein (WAVE) regulatory complex (WRC), formin-like protein FMNL1 and FHOD1 downstream of Rac. WASP, N-WASP, and WRC are nucleation promoting factors (NPF) for actin-nucleating proteins, including the actinrelated protein 2/3 (Arp2/3) complex and formins. MRCK kinases contribute to myosin II light chain (MLC) phosphorylation [122, 123]. Rho GTPases also regulate other cellular processes including vesicle trafficking, cell-cycle progression, apoptosis, and gene expression. RhoA, Rac1, Cdc42, and RhoG activate Kinectin (KTN1), which binds to Kinesin and enhances microtube-dependent kinesin ATPase activity [124, 125]. Protein kinase N (PKN) 1–3 downstream of RhoA have diverse functions including cell-cycle regulation, receptor trafficking, vesicle transport, and apoptosis [126]. Several effectors initiate signaling pathways influencing gene expression. For example, RhoA, Rac1, and Cdc42 can all direct interact with and enhance PKCα [127], which has diverse functions [128]. RhoA, Rac1 and Cdc42 are required for transcription activation by serum response factor (SRF) [129], releasing the SRF cofactors of the myocardin-related transcription factor (MRTF) family from monomeric G-actin when actin polymerization occurs. Additional effectors for Rho subfamily includes phospholipase C (PLC) ε [130] and diacylglycerol kinase (DGK) θ [131]. Additional Rac and

Y. Lin and Y. Zheng

Cdc42 effectors include mixed lineage kinase (MLK) 3 [132], MEKK1 and 4 [133], p70S6K [134], PI3K regulatory subunit p85 [135], and PLCβ2 [136]. Importantly, Rho GTPases also participate in enzyme assembly. For example, Rac1 and Rac2 directly interact with p67phox, a cytosolic component of the NADPH oxidase complex, and regulate NADPH oxidase in producing ROS [137].

4.2

Effector Activation Mechanisms

Despite of the diversity of Rho GTPase effectors, there are some common regulatory themes for their activation. The effectors are usually multi-domain proteins. Binding of Rho GTPases often induces a conformational change, releasing effectors from an autoinhibitory conformation to expose functional domains. Examples of this theme include regulation of PAK, ROCK, PKN, mDia, WASP, and IRSp53 [108]. Interaction with Rho GTPases may also release effectors from inhibitory binding partners, such as the WAVE regulatory complex [138]. Their full activation can be achieved by binding of Rho GTPases alone or coupled to other signals. For example, PKN needs RhoA binding, lipid association, and autophosphorylation to become fully active [139]. PAK autophosphorylates on multiple sites to control its activity [140]. There are also occasions that binding of Rho GTPases results in an inhibition of effectors, such as Cdc42EPs [141]. Another layer of effector regulation is subcellular localization. Recruiting of effectors to membranes or other proper subcellular compartments is often a part of the activation mechanism. For example, translocation of PAK, PKN, Citron, ROCK, and several other effectors for activation has been reported [109].

4.3

Rac GTPases in the Activation of NADPH Oxidases

NADPH oxidases of the NOX family are important enzymatic sources of reactive oxygen species (ROS). The phagocyte NADPH oxidase-mediated release of ROS, also termed “oxidative burst”, is key for host defense and inflammation responses [142]. In addition to the phagocyte NOX (i.e., NOX2), other NOX family enzymes responsible for superoxide production have been defined in various tissues, including NOX1, 3, 4, 5, and Duox1/2 [143]. Among them, NOX1 and NOX2 require Rac1 or Rac2 GTPase for their assembly and activation [144]. NOX1 is widely expressed in various non-hemopoietic cell types, with a high expression in colon epithelium. NOX1 activation is mediated by Rac1. NOX2 in phagocytes has been well described [143, 145– 147], and Rac1/Rac2 play an exquisite role essential in the regulation of this enzyme.

18

Rho Family GTPases and their Modulators

295

Fig. 18.4 Activation process of the phagocytic NADPH oxidase NOX2 by Rac GTPase. NOX2 (gp91phox) harbors six transmembrane helices and helices 3 and 5 chelate two heme groups, whereas the structure of p22phox is not clear despite various studies, represented by two transmembrane helices here. The domain architectures of p47phox, p67phox, and p40phox are shown schematically. In resting condition, these three subunits form trimeric complex in the cytosol. p47phox proline-rich region (PRR) interacts with p67phox C-terminal SH3 domain, and p67phox and p40phox interact through their respective phox and Bem1 (PB1) domains (shown as dotted lines). Pathogens or other stimuli trigger signaling receptors that lead to the phosphorylation of p47phox, and possibly p40phox and p67phox. After phosphorylation, the

tandem SH3 domains of p47phox, previously masked by an adjacent polybasic autoinhibitory region (AIR), interact with the PRR domain of p22phox, inducing translocation of the p47phox/p67phox/p40phox complex to the membrane bound NOX2. At the membrane, the PX domains of p40phox and p47phox bind to specific membrane lipids to stabilize the NOX2 complex. In parallel, Rac-GDP sequestered in cytosol by RhoGDI is translocated to the membrane and the bound GDP is exchanged for GTP by Rho GEFs. Rac-GTP further interacts with the tetratricopeptide repeats (TRPs) of p67phox, promoting its membrane translocation and exposing its activation domain (AD) to promote electron transfer through NOX2. PKC, PAK, and other kinases involved in this activation process are not depicted for simplicity

The phagocytic NADPH oxidase comprises six subunits that interact to form an active enzyme complex (Fig. 18.4). Two of those, NOX2 and p22phox, are integral membrane proteins that form the heterodimeric cytochrome b558. NOX2, also known as gp91phox, is an electron transferase and p22phox serves as a scaffolding protein. Under resting conditions, the regulatory subunits, p40phox, p47phox and p67phox, are all in an autoinhibitory conformation and exist

in the cytosol as a complex (see Chap. 15 by P.M.-C. Dang and J. El-Benna, Chap. 16 by H. Sumimoto, and Chap. 17 by T.W. Kuijpers and D. Roos). A 3D structural model of this trimeric complex has been proposed using fluorescence resonance energy transfer (FRET), fluorescence cross correlation spectroscopy (FCSS), and small angle X-ray scattering (SAXS) [148]. Interaction of cytochrome b558 with the cytosolic components results in oxidase activation, a process

296

requiring translocation of the cytosolic complex to the membrane-bound cytochrome. Rac1 (in macrophages) or Rac2 (in neutrophils), sequestered by GDI when not activated, were found as another cytosolic component critical for this process [149–153]. Rac1/Rac2 interacts with p67phox [137, 154] and critical regions for this interaction have been mapped on both Rac1/Rac2 [155–157] and p67phox structures [158]. In parallel to interacting with p67phox, Rac1/Rac2 has also been proposed to directly bind to cytochrome b558 subunit NOX2 [154, 159, 160], and translocation of Rac2 to neutrophil plasma membranes was seen to be dependent on cytochrome b558, but not p47phox or p67phox [161]. This proposal remains controversial, however, because of the unexpected equal binding abilities of GTP- and GDP-bound Rac and the mediation of binding by the Rac insert domain, the role of which in oxidase activation is still unclear [144]. Rac GTPases were initially thought to only function as a “carrier” to recruit the p47phox/p40phox/p67phox complex to the membrane. Prenylated Rac can recruit p67phox to the membrane independent of p47phox and an amphiphilic activator [162]. However, studies using various p67phox-Rac1 chimeric constructs and further mutagenesis work by the Pick group suggest that Rac can function more than mere recruitment of p67phox to the membrane [163–168]. It is now generally accepted that upon stimulation, there are two independent translocation events of the cytosolic components to the membrane: translocation of p47phox/p40phox/p67phox via p47phox-p22phox interaction after p47phox phosphorylation, and translocation of Rac GTPase. In addition, both p47phox and p40phox have a PX (phox homology) domain that interacts with phospholipids. At the membrane, Rac interacts with p67phox and induces a conformational change that exposes a hidden sequence to bind to NOX2, promoting superoxide production [169]. It has become clear that NOX2 regulation involves a complex network of protein-protein and protein-lipid interactions. Triggered by the activation of specific receptors, multiple levels of regulation occur during this process. Phosphorylation of several phox subunits through receptorspecific signaling pathways, including p22phox, p67phox, p40phox, and more importantly, p47phox, directly modulates translocation of the p47phox/p40phox/p67phox complex and NADPH oxidase activity [142, 146, 170]. The other independent but critical event is the translocation and activation of Rac GTPase to its GTP-bound state [144, 171]. The cycling of Rac1/Rac2 between the GDP- and the GTP-bound states is controlled by selective GEFs and GAPs. Several GEFs have been implicated in promoting NOX2 activity in neutrophils, including P-Rex (mostly P-Rex1), Vav (mostly Vav1 and Vav3), Tiam (mostly Tiam2), and DOCK (mostly DOCK2 and DOCK5) families [172]. Major GAPs in neutrophils contributing to the termination of NOX2 activity include Bcr, p190RhoGAP, and p50RhoGAP [173]. Further,

Y. Lin and Y. Zheng

membrane phospholipids and PtdIns(3,4,5)P3 generated by phosphoinositide-3-kinase (PI3K) pathway are also implicated in NOX2 regulation. They contribute to Rac dissociation from GDI and its activation by GEFs after the Rac-RhoGDI complex translocation to the membrane [60, 61]. PtdIns(3,4,5)P3 is also central for neutrophil Rac-GEF regulation, acting in accordance with other signaling mediators unique to each GEF [172, 174]. Membrane phospholipids also help anchor p47phox and p40phox. The PX domain of p47phox binds to phosphoinositides, most strongly phosphatidylinositol 3,4-biphosphate [PtdIns(3,4)P2], as well as phosphatidic acid (PA) and phosphatidylserine (PS). The PX domain of p40phox is highly specific for phosphatidylinositol 3-phosphate [PtdIns(3)P] [147]. Another link that Rac proteins regulate the NADPH oxidase is through the downstream effector PAK. Among the PAK family, PAK1 and PAK2 are the most abundant members in leukocytes and are implicated in the regulation of NADPH oxidase activity through several direct and indirect mechanisms [175]. Besides RhoGDI, PAK1 could also phosphorylate p47phox in vitro, releasing it from the closed conformation [176]. In human neutrophils, PAK1 has been shown to colocalize with and phosphorylate p47phox and to directly bind to p22phox [177]. The Rac-PAK1-p47phox signaling provides a direct crosstalk between the NADPH oxidase components. Besides activation by Rac GTPases, PAK1 has been shown to be stimulated by lipids, such as sphingosine and sphingoid bases, and phosphorylate p47phox in vitro [178]. PAK2, on the other hand, has been reported to phosphorylate p67phox [179]. Moreover, PAK1 and PAK2 are important regulators in actin cytoskeleton remodeling, which in turn is involved in NADPH oxidase assembly [180].

5

Structural Aspects of Rho GTPase Regulation

The GDP-bound form of small GTPases is generally considered inactive, while the GTP-bound form can turn on downstream signaling by binding to effectors. Extensive structural studies of various Rho GTPases in either GDP- or GTP-bound form or in complex with regulators or effectors have provided a deep understanding to the molecular mechanisms of their activation and regulation. Although the “molecular switch” mechanism is widely accepted, it is becoming clear that the binary “ON” and “OFF” model is oversimplified.

5.1

The Molecular Switch Mechanism

The Ras family GTPases share conserved structural features and the switch mechanism. The core domain, termed

18

Rho Family GTPases and their Modulators

G-domain, carries the basic function of guanine nucleotide binding and hydrolysis, comprising of a six-stranded β sheet and five helices (Fig. 18.1b). Most Rho GTPases have an extra insertion region, which forms an additional α-helix together with a short, less stable second helix. A N/TKxD motif, an ExSAK motif, and a P-loop with Gx4GKS/T signature contribute to nucleotide binding. Two structural regions, termed Switch I and Switch II, undergo conformational change after GDP/GTP exchange and are involved in GTP hydrolysis [181]. The two switch regions are highly dynamic and show a large conformational space in the GDP-bound proteins. When GTP is loaded, two conserved residues from each switch region, T35 and G60 (numbering in Rac1), form hydrogen bonds with the γ-phosphate to pull the switch regions in a “closed” conformation to interact with effectors or GAPs. After the bound GTP is hydrolyzed, the switch regions are released back to the GDP-specific conformations. For GTPase-deficient atypical Rho GTPases (i.e., Rnd, RhoBTB, and RhoH), the key residues for GAP binding and GTP hydrolysis differ from the conserved amino acids used in the classical Rho GTPases and other Ras-related proteins. For fast-cycling atypical Rho GTPases (i.e., RhoD/RhoF, RhoU/RhoV), these residues are conserved. The mechanism behind their fast-cycling activities remains poorly understood.

5.2

Dynamics of the Switch Regions and Conformational Equilibrium

The switch regions are intrinsically highly dynamic and can exist in distinct sub-states even in the GTP-bound form. 31 P-NMR studies revealed that H-Ras bound to GTP, or its non-hydrolysable analogs, exhibits two lines for the resonances of the α- and γ-phosphate, suggesting a dynamic equilibrium between two interconverting conformations, i.e., state 1 and state 2 [182–184]. In support of this, crystal structures of H-Ras in complex of GTP analogs have been solved in both state 1 and state 2 conformations, with state 2 representing a “closed” active conformation featuring interactions of the γ-phosphate with both switch regions and state 1 an “open” inactive conformation which loses the γ-phosphate interaction with Switch I [185]. For Rho GTPases, the dynamics of the switch regions seem to be different from that of Ras. For example, the switch regions of Cdc42-GDP appears to have a conformational change on a ~ ms timescale in NMR, compared to only fast motion (ps to ns) in Ras-GDP [186]. Its switch regions rigidified upon binding to effectors such as PAK, ACK, and WASP [187–189]. Consistently, crystal structures indicated that Cdc42-GMPPCP predominantly exists in an inactive conformation similar to Cdc42-GDP, and effector binding induces or stabilizes the conformational change [190]. Rac2

297

and Rac3 also contain similar inactive conformations for GDP- and GTP analog-bound forms [191]. In contrast, Rac1-GMPPNP assumes active state 2 conformation [192]. While RhoC-GTPγS yields an active conformation, RhoC-GMPPNP assumes the same inactive conformation as RhoC-GDP [193]. The alternative spliced variant Rac1b, which has a fast nucleotide exchange property, exhibits open Switch I and highly dynamic Switch II in crystal structure [15]. A more direct evidence for the existence of a conformational equilibrium in Rho GTPase comes from our recent paper on RhoA, where 31P NMR spectra of RhoAGTPγS, but not RhoA-GMPPNP, show two states for the αand γ- phosphate resonances and a state 1 RhoA structure is resolved [194]. GDP/GTP exchange of Rho GTPases can also cause longrange allosteric effects outside the switch regions. The N-terminal region of GTPases (corresponding to residues 1–86 in Ras), referred to as effector lobe, harbors the P-loop and the switch regions, forming the active site for GTP hydrolysis and a binding surface for most effectors and regulators. The C-terminal half, referred to as the allosteric lobe, contains the N/TKxD and ExSAK motifs for nucleotide binding. The shared nucleotide binding of the two lobes enhances the communication between them. The allosteric lobe has more sequence variations among Rho GTPases and contributes to substrate specificity. The allostery is better characterized in Ras with substantial X-ray, NMR, and simulation studies [195, 196]. Rho GTPases are less studied, with most of the information coming from NMR studies on Cdc42 and Rac1. The insert region, the anti-parallel β-sheet (β2-β3), and the region around helix 3 and loop7 all show changes in dynamics when bound to effectors, although these areas are not always directly involved in effector interactions [196].

5.3

Structural Diversity of Rho Regulators and Effectors

The switch regions are engaged in almost all interactions of small GTPases with their regulators and effectors. The only exception for Rho GTPases is the interaction between ROCK1 and Rnd3 where ROCK1 phosphorylates Rnd3 and is not an effector protein [197]. The dynamic switch regions, as well as the allosteric sites including the insert region, β2-β 3, and α3/L7, form flexible and fluctuating interaction surfaces, allowing the binding of the GTPases to multiple partners with distinct structure folds. A large number of structure models are available for RhoA, Rac1, and Cdc42 in complex with various GEFs, GAPs, and RhoGDI1, providing opportunities to analyze the structural basis of how Rho GTPases are recognized and modulated by these regulators. The molecular mechanism underlying the action of GEFs, GAPs, and GDIs are well-

298

understood [26] and is discussed briefly in Sect. 3. Several large-scale studies have been carried out for in-depth understanding of the structure-function relationships for Dbl family GEFs [27], DOCK family GEFs [40], GAPs [49], and GDIs [58]. While the GAP domain of RhoGAPs and RhoGDIs lack selectivity in general, the catalytic domains of both Dbl and DOCK family GEFs show specificity or preferences for their substrates. Given the large number and complexity of their functions, how Rho regulators orchestrate spatiotemporalspecific Rho signaling in vivo will need further elucidations. There are ~30 X-ray and NMR structures of Rho GTPaseeffector complexes. The Rho binding domains (RBDs) of effectors have diverse structural features and can be classified into several structural classes: A) a pair of α-helices contacts the switch regions, such as ROCK1, PKN, formin-like proteins; B) CRIB-like domain that forms intermolecular β-sheet to extend the β2-β3 strands on Rac/Cdc42, such as ACK, WASP, Par6, PAK proteins; C) PH-domain, such as PLCβ2, PLCγ2, p190RhoGEF, LbcRhoGEF, PDZRhoGEF; D) others, such as p67phox and plexins. Representative structures of Rac1/Rac2 in complex with different classes of effectors are shown in Fig. 18.5. Most effectors, despite of their structural classes, engage Rho GTPases at the switch regions. Competition of PAK1 RBD domain with p67phox has been observed in the cell-free NADPH oxidase assay [198]. While structures of Rho-effector complexes have been reviewed in [199], we focus on the unique Rac1p67phox structure. The Rac-p67phox interaction is crucial for the activation of the NADPH oxidase enzyme. The binding site for Rac1 or Rac2 on p67phox was located to the N-terminal 200 residues, which contains four tetratricopeptide repeats (TPR) [158]. On Rac, residues involved in p67phox binding reside in the Switch I region, as mutants T35A, D38A, and Y40K lack p67phox binding activity and cannot stimulate oxidase activity. Further studies confirmed the role of the N-terminal region (residues 22–45) and found a second domain (residues 143–175) required for p67phox interaction [155]. These results are consistent with the crystal structure of Rac1 (Q61L,1–184)/p67phox (1–203) (PDB 1E96), which reveals residues S22, T25, N26, F28, G30, E31, A159, L160, and Q162 of Rac1 directly participating in p67phox binding [200]. Unusually, Switch II of Rac1 does not interact with p67phox. On p67phox, residues directly involved in Rac interaction comprise S37, D67, R102, N104, L106, and D108. The TPR domain of p67phox in complex with Rac-GTP does not show significant conformational change when compared to a structure of itself [201]. Cdc42 does not interact with

Y. Lin and Y. Zheng

p67phox despite of its similarity to Rac1. The only residues that are different between Cdc42 and Rac1 around the interaction site are A27Rac1 and G30Rac1, which are lysine and serine in Cdc42, respectively. The Rac1 A27K/G30S mutant cannot bind to p67phox, whereas a Cdc42 L27A/S30G mutant is able to bind [200].

5.4

Dynamics and Allostery Contributing to Substrate Specificity

Besides recognition of the key residues on the interaction surface, the dynamics of the switch regions and the resulted conformational space also contribute to the specificity of Rho GTPase binding to regulators/effectors. One example is substrate recognition by the DOCK family Rho GEFs revealed by the crystal structures of DOCK2/Rac1, DOCK9/Cdc42, and DOCK7/Cdc42 [40]. The catalytic domain, DHR2, consists of three lobes: lobe A mediates homodimerization while lobe B and C serve as binding sites for Rac/Cdc42. Lobe C contains an α10 insertion loop nucleotide sensor that interacts extensively with Switch I [202] and contributes partially to substrate specificity via recognition of residue 56 [203, 204]. Lobe B, interestingly, adopts a different arrangement relative to Lobe C in the Rac-specific DOCK2 vs. the Cdc42-specific DOCK9, corresponding to different Switch I conformations of Rac1 and Cdc42. The dual specific DOCK7 may alter the conformation of lobe B for Rac1 or Cdc42 recognition [205]. Another example is the Rac1p67phox interaction. A “peptide walking” strategy to identify the functional domains on Rac1 for NOX2 oxidase regulation implicated four domains: α3/L7 region (residues 103–107), insert region (residues 124–135), the C-terminal p67phox binding region (residues 163–169), and the polybasic C-terminal region (residues 183–188) [157]. While the polybasic C-terminal region is important for Rac1 membrane association, the α3/L7 region and insert region may contribute to the Rac1-p67phox interaction via an allosteric effect. Rho GTPases have overlapping interactive interfaces involved in binding to upstream regulators and downstream effectors, and thus, understanding the dynamic behavior is important to understand their biological function and regulation. In addition, certain activating mutations may change the dynamic behavior and conformational landscapes. Rac1 P29S mutation, frequently found in melanoma, has been shown to remarkably alter the conformational landscape and shift equilibrium toward a conformation with reduced affinity for Mg2+ cofactor [206]. The inherent dynamics of

18

Rho Family GTPases and their Modulators

299

Fig. 18.5 Representative 3-dimentional structures of Rac-effector complexes. (a) Complex of Rac1 with effectors of helical pairs: Rac1PKN (PDB 2RMK) and Rac1-CRYI-B (PDB 7AJK). (b) Complex of Rac3 with an effector with CRIB domain: Rac3-PAK4 (PDB 2OV2). (c) Complex of Rac1 with an effector containing PH domain: Rac1-PLCβ2

(PDB 2FJU). (d) Complex of Rac1 with effectors with other structures: Rac1-p67phox (PDB 1E96) and Rac1-Plexin A1 (PDB 3RYT). In all complexes, Rac GTPase is colored in grey with the switch regions highlighted in magenta. All structure figures were generated by PyMol (Schrödinger, LLC)

Rho GTPases and other Ras-related proteins and their conformational heterogeneity is of particular interest in the drug discovery effort targeting Rho GTPase signaling.

assess the activity of recombinant Rac proteins [207]. While those in vitro studies have allowed the discovery of important aspects of their regulation and function, the limitations are obvious. To dissect the precise function of specific Rho GTPases in complex physiological processes in the tissues where they are expressed, genetic animal models have been used to explore individual Rho GTPase function in vivo under normal or pathological conditions. Gene-targeted mouse lines for many Rho family members have been generated. Knockout mice of RhoB, RhoC, Rac2, Rac3, RhoG, RhoQ, RhoF, Rnd3, and RhoH are viable, whereas RhoA, Rac1, and Cdc42 knockout are embryonic lethal, and

6

Rho GTPase Function in Physiological Processes and Human Diseases

Much of the functional studies for Rho GTPases in the years 1990 to 2000 involved cultured mammalian cells expressing constitutively active or dominant negative mutants or cellfree assays such as NADPH oxidase enzyme activity to

300

Y. Lin and Y. Zheng

hence required conditional knockout. The phenotypes of the knockout mice for Rho GTPases have been summarized in several excellent reviews [17, 109, 208–211].

6.1

Genetic Analysis of Rac Subfamily GTPases

Rac proteins can stimulate lamellipodium and membraneruffle formation and induce membrane extension during phagocytosis. Rac1-/- Schwann cells, endothelial cells, and platelets show impaired lamellipodium formation [212– 214]. Rac1 is important for axon guidance, while its role in axon growth can be compensated by Rac3 [215]. In non-hemopoietic tissues, Rac1 is often the main player in NADPH oxidase regulation. Rac1-/- cardiomyocytes show reduced NADPH oxidase activity [216] and Rac1-/-osteoclasts show lower ROS levels [217]. Rac3-/mice show enhanced motor coordination and learning without detectable defects in brain structure or neuron organization [18]. Rac proteins have both unique and overlapping roles in hemopoietic cells. In macrophages, Rac1 is most abundant and its deficiency causes an altered morphology, reduced membrane ruffling and spreading [218], Rac2 deficiency causes reduced F-actin levels and lack of podosomes [219], whereas Rac1-/-, Rac2-/-, and Rac1-/-/Rac2-/macrophages all show normal migration and chemotaxis. Rac2 is important for phagocytosis and superoxide production in macrophages in response to some, but not all, stimuli [220]. In neutrophils, Rac1 and Rac2 appear to be expressed at equal amounts. Rac1-/- neutrophils display normal migratory activity but are unable to move toward the source of chemoattractants. By contrast, Rac2-/- neutrophils can orient toward the chemoattractant source but are unable to migrate efficiently [221]. Interestingly, RhoG is not required for neutrophil chemotaxis [222]. The activity of NADPH oxidase for superoxide production is drastically reduced in Rac2-/- neutrophils [223] but normal in Rac1-/- neutrophils [224]. While NADPH oxidase activity is reduced in RhoG-/neutrophils in response to some stimuli [222], this is likely due to a role of RhoG in signal transduction from receptor to the oxidase, probably acting upstream of Rac1/Rac2 by direct interaction with the DOCK2-binding protein ELMO [225]. Rac2 is also required for neutrophil primary granule release where Rac1 plays only a minor role [226]. In hemopoietic stem and progenitor cells (HSPCs), deficiency of Rac1, but not Rac2, strongly impairs engraftment [227]. Rac1 is essential for the entry of HSPCs into the cell cycle and progression upon extracellular stimulation while Rac2 is important for cytoskeletal responses, adhesion, spreading and Akt-dependent survival. Deficiency of both

Rac1 and Rac2 leads to massive mobilization of HSPCs from the bone marrow [228]. Rac1 and Rac2 shows overlapping functions in T-cells, B-cells, and platelets. Rac2-/- mice show no obvious defects in T-cell differentiation, but reduced proliferation and T cell receptor (TCR)-stimulated actin polymerization during T-cell activation [229], and reduced interferon gamma (INFγ) production in Th1 cells [230]. The defects in T-cells caused by Rac2 deficiency appear minor, suggesting its redundant role. Rac1 and Rac2 seem to have unique roles in production of common lymphoid progenitors but share a redundant role in later stages of T-cell development [231]. Rac2-/- mice show defects in the B-cell development, reduced levels of Ca2+ fluxes and cell proliferation [232], whereas Rac2-/- mice with combined conditional deletion of Rac1 in the B cell linage show almost completely blocked B-cell development [233]. RhoG-/- mice have normal T- and B-cell development with only a mildly increased response to antigens [20]. In platelets, Rac1 is the major player and there is no detectable protein level for Rac2; Rac1-/- and Rac1-/-/ Rac2-/- platelets show identical phenotype and Rac2-/platelets show a similar phenotype as wild type [212]. Rac1 is required for platelet aggregation and thrombus formation. RhoG-/- platelets display significantly reduced integrin activation and aggregation when stimulated by C-reactive protein (CRP) but yield a normal response to thrombin [234].

6.2

Rho GTPases in Human Diseases

With the wide variety of cellular roles Rho GTPases play, dysregulation of Rho signaling pathways has been implicated in many human diseases. In particular, Rho GTPases have been associated with cancer and appear to contribute to nearly all stages of cancer development and progression. Rho GTPase signaling and deregulation in cancer have been reviewed extensively [235–240]. Altered expression of Rho GTPases, or altered activities by dysregulations of their regulators, is often observed in cancers. Mutations of several Rho GTPases, notably RhoA and Rac1, have been recently discovered in several tumor types including diffuse gastric cancer and leukemia/lymphomas [241–243]. Both gain-offunction and loss-of-function mutations were identified in RhoA—this puzzling finding, together with recent genetic studies, suggest a more complicated paradigm of Rho GTPase functions in cancer as certain Rho family members can be either pro- or anti-tumorigenic in a tumor type-specific manner [244, 245]. For Rac1, gain-of-function mutations have been identified in cancer, including the hotspot of P29 and less frequent sites at C18, P34, A159, and A178 [235]. The Rac1 P29S mutation, occurring in 4%–9% of sun-exposed melanomas, is fast-cycling by itself and

18

Rho Family GTPases and their Modulators

promotes the proliferation and migration of melanoma [246, 247]. Consistently, gain-of-function mutations in Rac GEFs (P-Rex2, Vav family) and Rac downstream effectors (PAK1, PAK4, PAK5) were also revealed [248, 249]. Rac family may also have anti-tumorigenic functions. Rac GEF P-Rex2 and effector PAK3 have been found recurrently deleted in pancreatic cancer and desmoplastic melanoma, respectively [250, 251]. Rac GEF Tiam1 was shown to inhibit invasion of intestinal epithelial cells by antagonizing TAZ/YAP signaling [252]. Before the recent discoveries of recurrent mutations in cancer, disease-related mutations in Rho GTPases were relatively rare. The two hemopoietic-specific Rho GTPases, Rac2 and RhoH, were found to be mutated in human immunodeficiency diseases. Rac2 D57N mutation, which is dominant-negative, was found in human neutrophil immunodeficiency syndrome [253, 254]. Neutrophils from the patient with Rac2 D57N mutation show defective superoxide production, adhesion, and chemotaxis. A loss-of-function mutation of Rac2 (W56X) was later identified in patients with a form of common variable immunodeficiency [255]. Neutrophils from these patients show reduced chemotaxis and abnormal released granules. RHOH was first identified as a fusion partner and hypermutable gene in malignant lymphoma and high expression levels of RhoH were detected in lymphocytes [256, 257]. Lack of RhoH results in T cell deficiency and impaired T cell function both in human [258] and mice [259]. Mutation of residue Y38 in RhoH results in childhood Burkitt lymphoma. Mutations in the noncoding regions of RhoH have also been found in malignant B cells [260, 261]. In diseases of abnormal platelet function, several mutations in Rac GTPase signaling have been revealed. For example, mutations in the scaffolding protein Nbeal2 disrupt its interactions with the Rac GEF DOCK7 in gray platelet syndrome (GPS), linking this bleeding disorder with the molecular machinery mediating platelet granule secretion and hemostatic function [262]. While platelet disorder associated with Rac1 mutations in humans has yet to be described, recent studies report that somatic mutations in ELMO proteins, which regulate Rac activity through interactions with the GEF DOCK, are associated with congenital bleeding phenotypes [263, 264].

6.3

Therapeutic Targeting

Targeting Rho signaling has been an attractive approach for cancer therapeutic intervention [265–267]. Members of the Rac subfamily are mostly pro-tumorigenic, making them a desirable target. Like most small GTPases, Rac1 and its GEFs and GAPs are not easily druggable. Several strategies have been used to tackle this task: A) targeting Rac-GEF

301

interaction; B) targeting Rac-nucleotide binding; C) targeting Rac lipid modifications; D) targeting Rac downstream effectors, especially the PAK family kinases. Most of the small molecule probes of the above strategies are in the early pre-clinical stage and primarily used as pharmacological tools to dissect Rho GTPase functions. Given the challenges of inhibiting Rho GTPase activation directly, targeting downstream effectors remains a most promising approach. To this end, an inhibitor targeting the Rac-interactive surface of p67phox has been developed [268], and a few PAK inhibitors have progressed to phase I clinical trials, including PF-3758309 [269] and KPT-9274 [270]. PAK family members share structurally similar catalytic domains with a relatively flexible and open ATP-binding cleft [271]. Devising specific PAK inhibitors is challenging and needs further efforts. For immunodeficiency and platelet related abnormalities such as thrombocytopenia where Rho signaling pathways are defective, cell and gene therapy with patients’ HSPCs to complement or correct pathogenic mutations is a promising approach. For example, Wiskott-Aldrich Syndrome (WAS) is an inherited immunodeficiency caused by mutations in the gene encoding WASP, an effector downstream of Cdc42. Gene therapy studies using lentiviral vectors encoding functional WASP to genetically correct HSPCs from WAS patients are in clinical trials [272–274]. Gene editing platforms, particularly CRISPR/Cas9, allow correction of the affected gene in situ, offering a more promising approach. This strategy has been shown to allow precise correction of up to 60% of human HSPCs and displayed no major genotoxicity in mouse models [275].

7

Conclusion and Perspectives

The complexity of Rho GTPase-dependent signaling necessitates a diverse range of mechanisms to ensure proper cell functions. The large numbers of Rho GEFs and Rho GAPs compared to Rho GTPases indicate that specific cellular functions are likely mediated by specific regulators. The regulatory mechanisms and specificity conferred by the regulators are not well understood, as the majority of GEFs and GAPs remain poorly characterized. Recent work by the interactome approaches has provided a better picture of the intricate signaling assemblies centered around Rho GTPases. In addition, it is increasingly clear that the regulation of Rho GTPases goes beyond the classical GTP-GDP cycling. A range of posttranslational modifications on Rho GTPases themselves, as well as their regulators/effectors, are also involved in orchestrating the precise spatiotemporal activation of Rho GTPases in cell-type specific manner and in response to different stimuli. Such multi-layer regulatory mechanisms may ensure that each Rho GTPase is activated/

302

deactivated at the right subcellular location at the right time, enabling cells in multicellular organisms to respond to complex environments. Extensive crosstalk between Rho GTPases and related regulatory and effector proteins is often involved in physiological signaling processes. Further work is needed to expand our understanding of context-dependent roles of Rho GTPases and their regulatory events in specific tissue cell types. One approach to map the network of Rho GTPase regulation is the use of live imaging with Rho-FRET mouse [276] to define how different Rho GEFs and Rho GAPs bring about specific subsets of dynamic, Rho-dependent cell responses. Another key area is to define the Rho GTPase context-specific interactomes fueled by the advances in system biology, single cell RNA technologies, and proteomics. Rho GTPase signaling has been implicated in many human diseases including Rac the rare mutation mediated immunodeficiencies; yet therapeutic approaches targeting abnormal signaling caused by Rho GTPase mutations remains at early preclinical stages. Future studies of the structural dynamics and allostery of Rho GTPases are important for understanding the behaviors and compensatory resistance development of various Rho GTPase complexes in the context of genetic mutations. To this end, in addition to early leads of small molecules targeting individual Rho GTPase pathways, gene editing is emerging as an important tool for precise manipulation of mutant cells that may benefit Rho GTPase mutation associated genetic diseases, especially in hemopoietic cell types. Acknowledgements The work is partially supported by United States National Institutes of Health grants R01CA204895, R01AG063967, and R01 HL147536.

References 1. Wennerberg K, Rossman KL, Der CJ (2005) The Ras superfamily at a glance. J Cell Sci 118(Pt 5):843–846. https://doi.org/10.1242/ jcs.01660 2. Madaule P, Axel R (1985) A novel ras-related gene family. Cell 41(1):31–40. https://doi.org/10.1016/0092-8674(85)90058-3 3. Didsbury J, Weber RF, Bokoch GM, Evans T, Snyderman R (1989) rac, a novel ras-related family of proteins that are botulinum toxin substrates. J Biol Chem 264(28):16378–16382 4. Hodge RG, Ridley AJ (2016) Regulating Rho GTPases and their regulators. Nat Rev Mol Cell Biol 17(8):496–510. https://doi.org/ 10.1038/nrm.2016.67 5. McWilliam H, Li W, Uludag M, Squizzato S, Park YM, Buso N et al (2013) Analysis tool web services from the EMBL-EBI. Nucleic Acids Res 41(Web Server issue):W597–W600. https:// doi.org/10.1093/nar/gkt376 6. Waterhouse AM, Procter JB, Martin DM, Clamp M, Barton GJ (2009) Jalview version 2--a multiple sequence alignment editor and analysis workbench. Bioinformatics 25(9):1189–1191. https://doi. org/10.1093/bioinformatics/btp033

Y. Lin and Y. Zheng 7. Schaefer A, Reinhard NR, Hordijk PL (2014) Toward understanding RhoGTPase specificity: structure, function and local activation. Small GTPases 5(2):6. https://doi.org/10.4161/21541248.2014. 968004 8. Williams CL (2003) The polybasic region of Ras and Rho family small GTPases: a regulator of protein interactions and membrane association and a site of nuclear localization signal sequences. Cell Signal 15(12):1071–1080. https://doi.org/10.1016/s0898-6568(03) 00098-6 9. Foster R, Hu KQ, Lu Y, Nolan KM, Thissen J, Settleman J (1996) Identification of a novel human Rho protein with unusual properties: GTPase deficiency and in vivo farnesylation. Mol Cell Biol 16(6):2689–2699. https://doi.org/10.1128/MCB.16.6.2689 10. Nobes CD, Lauritzen I, Mattei MG, Paris S, Hall A, Chardin P (1998) A new member of the rho family, Rnd1, promotes disassembly of actin filament structures and loss of cell adhesion. J Cell Biol 141(1):187–197. https://doi.org/10.1083/jcb.141.1.187 11. Dallery-Prudhomme E, Roumier C, Denis C, Preudhomme C, Kerckaert JP, Galiegue-Zouitina S (1997) Genomic structure and assignment of the RhoH/TTF small GTPase gene (ARHH) to 4p13 by in situ hybridization. Genomics 43(1):89–94. https://doi.org/10. 1006/geno.1997.4788 12. Rivero F, Dislich H, Glockner G, Noegel AA (2001) The Dictyostelium discoideum family of rho-related proteins. Nucleic Acids Res 29(5):1068–1079. https://doi.org/10.1093/nar/29.5. 1068 13. Aspenstrom P, Ruusala A, Pacholsky D (2007) Taking Rho GTPases to the next level: the cellular functions of atypical Rho GTPases. Exp Cell Res 313(17):3673–3679. https://doi.org/10. 1016/j.yexcr.2007.07.022 14. Aspenstrom P (2020) Fast-cycling rho GTPases. Small GTPases 11(4):248–255. https://doi.org/10.1080/21541248.2017.1391365 15. Fiegen D, Haeusler LC, Blumenstein L, Herbrand U, Dvorsky R, Vetter IR et al (2004) Alternative splicing of Rac1 generates Rac1b, a self-activating GTPase. J Biol Chem 279(6):4743–4749. https://doi.org/10.1074/jbc.M310281200 16. Sugihara K, Nakatsuji N, Nakamura K, Nakao K, Hashimoto R, Otani H et al (1998) Rac1 is required for the formation of three germ layers during gastrulation. Oncogene 17(26):3427–3433. https://doi.org/10.1038/sj.onc.1202595 17. Heasman SJ, Ridley AJ (2008) Mammalian Rho GTPases: new insights into their functions from in vivo studies. Nat Rev Mol Cell Biol 9(9):690–701. https://doi.org/10.1038/nrm2476 18. Corbetta S, Gualdoni S, Albertinazzi C, Paris S, Croci L, Consalez GG et al (2005) Generation and characterization of Rac3 knockout mice. Mol Cell Biol 25(13):5763–5776. https://doi.org/10.1128/ MCB.25.13.5763-5776.2005 19. Roberts AW, Kim C, Zhen L, Lowe JB, Kapur R, Petryniak B et al (1999) Deficiency of the hematopoietic cell-specific Rho family GTPase Rac2 is characterized by abnormalities in neutrophil function and host defense. Immunity 10(2):183–196. https://doi.org/10. 1016/s1074-7613(00)80019-9 20. Vigorito E, Bell S, Hebeis BJ, Reynolds H, McAdam S, Emson PC et al (2004) Immunological function in mice lacking the Rac-related GTPase RhoG. Mol Cell Biol 24(2):719–729. https:// doi.org/10.1128/MCB.24.2.719-729.2004 21. Croft DR, Olson MF (2011) Transcriptional regulation of Rho GTPase signaling. Transcription 2(5):211–215. https://doi.org/10. 4161/trns.2.5.16904 22. Liu M, Bi F, Zhou X, Zheng Y (2012) Rho GTPase regulation by miRNAs and covalent modifications. Trends Cell Biol 22(7): 365–373. https://doi.org/10.1016/j.tcb.2012.04.004 23. Boureux A, Vignal E, Faure S, Fort P (2007) Evolution of the Rho family of ras-like GTPases in eukaryotes. Mol Biol Evol 24(1): 203–216. https://doi.org/10.1093/molbev/msl145

18

Rho Family GTPases and their Modulators

24. Hall A (1998) Rho GTPases and the actin cytoskeleton. Science 279(5350):509–514. https://doi.org/10.1126/science.279.5350.509 25. Etienne-Manneville S, Hall A (2002) Rho GTPases in cell biology. Nature 420(6916):629–635. https://doi.org/10.1038/nature01148 26. Cherfils J, Zeghouf M (2013) Regulation of small GTPases by GEFs, GAPs, and GDIs. Physiol Rev 93(1):269–309. https://doi. org/10.1152/physrev.00003.2012 27. Jaiswal M, Dvorsky R, Ahmadian MR (2013) Deciphering the molecular and functional basis of Dbl family proteins: a novel systematic approach toward classification of selective activation of the Rho family proteins. J Biol Chem 288(6):4486–4500. https:// doi.org/10.1074/jbc.M112.429746 28. Lemmon MA (2004) Pleckstrin homology domains: not just for phosphoinositides. Biochem Soc Trans 32(Pt 5):707–711. https:// doi.org/10.1042/BST0320707 29. Garcia-Mata R, Burridge K (2007) Catching a GEF by its tail. Trends Cell Biol 17(1):36–43. https://doi.org/10.1016/j.tcb.2006. 11.004 30. Rossman KL, Der CJ, Sondek J (2005) GEF means go: turning on Rho GTPases with guanine nucleotide-exchange factors. Nat Rev Mol Cell Biol 6(2):167–180. https://doi.org/10.1038/Nrm1587 31. Zheng Y (2001) Dbl family guanine nucleotide exchange factors. Trends Biochem Sci 26(12):724–732. https://doi.org/10.1016/ s0968-0004(01)01973-9 32. Boissier P, Huynh-Do U (2014) The guanine nucleotide exchange factor Tiam1: a Janus-faced molecule in cellular signaling. Cell Signal 26(3):483–491. https://doi.org/10.1016/j.cellsig.2013. 11.034 33. Michiels F, Stam JC, Hordijk PL, van der Kammen RA, Ruuls-Van Stalle L, Feltkamp CA et al (1997) Regulated membrane localization of Tiam1, mediated by the NH2-terminal pleckstrin homology domain, is required for Rac-dependent membrane ruffling and C-Jun NH2-terminal kinase activation. J Cell Biol 137(2): 387–398. https://doi.org/10.1083/jcb.137.2.387 34. Welch HC (2015) Regulation and function of P-Rex family Rac-GEFs. Small GTPases 6(2):49–70. https://doi.org/10.4161/ 21541248.2014.973770 35. Rodriguez-Fdez S, Bustelo XR (2019) The Vav GEF family: an evolutionary and functional perspective. Cell 8(5):465. https://doi. org/10.3390/cells8050465 36. Cote JF, Motoyama AB, Bush JA, Vuori K (2005) A novel and evolutionarily conserved PtdIns(3,4,5)P3-binding domain is necessary for DOCK180 signalling. Nat Cell Biol 7(8):797–807. https:// doi.org/10.1038/ncb1280 37. Premkumar L, Bobkov AA, Patel M, Jaroszewski L, Bankston LA, Stec B et al (2010) Structural basis of membrane targeting by the Dock180 family of Rho family guanine exchange factors (Rho-GEFs). J Biol Chem 285(17):13211–13222. https://doi.org/ 10.1074/jbc.M110.102517 38. Sakurai T, Kukimoto-Niino M, Kunimura K, Yamane N, Sakata D, Aihara R et al (2021) A conserved PI(4,5)P2-binding domain is critical for immune regulatory function of DOCK8. Life Sci Alliance 4(4):e202000873. https://doi.org/10.26508/lsa.202000873 39. Brugnera E, Haney L, Grimsley C, Lu M, Walk SF, ToselloTrampont AC et al (2002) Unconventional Rac-GEF activity is mediated through the Dock180-ELMO complex. Nat Cell Biol 4(8):574–582. https://doi.org/10.1038/ncb824 40. Kukimoto-Niino M, Ihara K, Murayama K, Shirouzu M (2021) Structural insights into the small GTPase specificity of the DOCK guanine nucleotide exchange factors. Curr Opin Struct Biol 71: 249–258. https://doi.org/10.1016/j.sbi.2021.08.001 41. Chen Y, Chen Y, Yin W, Han H, Miller H, Li J et al (2021) The regulation of DOCK family proteins on T and B cells. J Leukoc Biol 109(2):383–394. https://doi.org/10.1002/JLB.1MR0520221RR

303 42. Kunimura K, Uruno T, Fukui Y (2020) DOCK family proteins: key players in immune surveillance mechanisms. Int Immunol 32(1): 5–15. https://doi.org/10.1093/intimm/dxz067 43. Laurin M, Cote JF (2014) Insights into the biological functions of Dock family guanine nucleotide exchange factors. Genes Dev 28(6):533–547. https://doi.org/10.1101/gad.236349.113 44. Fukui Y, Hashimoto O, Sanui T, Oono T, Koga H, Abe M et al (2001) Haematopoietic cell-specific CDM family protein DOCK2 is essential for lymphocyte migration. Nature 412(6849):826–831. https://doi.org/10.1038/35090591 45. Moon SY, Zheng Y (2003) Rho GTPase-activating proteins in cell regulation. Trends Cell Biol 13(1):13–22. https://doi.org/10.1016/ s0962-8924(02)00004-1 46. Rittinger K, Walker PA, Eccleston JF, Smerdon SJ, Gamblin SJ (1997) Structure at 1.65 A of Rho A and its GTPase-activating protein in complex with a transition-state analogue. Nature 389(6652):758–762. https://doi.org/10.1038/39651 47. Scheffzek K, Ahmadian MR (2005) GTPase activating proteins: structural and functional insights 18 years after discovery. Cell Mol Life Sci 62(24):3014–3038. https://doi.org/10.1007/s00018-0055136-x 48. Garrett MD, Self AJ, van Oers C, Hall A (1989) Identification of distinct cytoplasmic targets for ras/R-ras and rho regulatory proteins. J Biol Chem 264(1):10–13 49. Amin E, Jaiswal M, Derewenda U, Reis K, Nouri K, Koessmeier KT et al (2016) Deciphering the molecular and functional basis of RHOGAP family proteins: a systematic approach toward selective inactivation of rho family proteins. J Biol Chem 291(39): 20353–20371. https://doi.org/10.1074/jbc.M116.736967 50. Ligeti E, Dagher MC, Hernandez SE, Koleske AJ, Settleman J (2004) Phospholipids can switch the GTPase substrate preference of a GTPase-activating protein. J Biol Chem 279(7):5055–5058. https://doi.org/10.1074/jbc.C300547200 51. Garcia-Mata R, Boulter E, Burridge K (2011) The 'invisible hand': regulation of RHO GTPases by RHOGDIs. Nat Rev Mol Cell Biol 12(8):493–504. https://doi.org/10.1038/nrm3153 52. Boulter E, Garcia-Mata R, Guilluy C, Dubash A, Rossi G, Brennwald PJ et al (2010) Regulation of Rho GTPase crosstalk, degradation and activity by RhoGDI1. Nat Cell Biol 12(5): 477–483. https://doi.org/10.1038/ncb2049 53. Brunet N, Morin A, Olofsson B (2002) RhoGDI-3 regulates RhoG and targets this protein to the Golgi complex through its unique N-terminal domain. Traffic 3(5):342–357. https://doi.org/10.1034/ j.1600-0854.2002.30504.x 54. Fukumoto Y, Kaibuchi K, Hori Y, Fujioka H, Araki S, Ueda T et al (1990) Molecular cloning and characterization of a novel type of regulatory protein (GDI) for the rho proteins, ras p21-like small GTP-binding proteins. Oncogene 5(9):1321–1328 55. Gorvel JP, Chang TC, Boretto J, Azuma T, Chavrier P (1998) Differential properties of D4/LyGDI versus RhoGDI: phosphorylation and rho GTPase selectivity. FEBS Lett 422(2):269–273. https://doi.org/10.1016/s0014-5793(98)00020-9 56. Lelias JM, Adra CN, Wulf GM, Guillemot JC, Khagad M, Caput D et al (1993) cDNA cloning of a human mRNA preferentially expressed in hematopoietic cells and with homology to a GDP-dissociation inhibitor for the rho GTP-binding proteins. Proc Natl Acad Sci U S A 90(4):1479–1483. https://doi.org/10. 1073/pnas.90.4.1479 57. Scherle P, Behrens T, Staudt LM (1993) Ly-GDI, a GDP-dissociation inhibitor of the RhoA GTP-binding protein, is expressed preferentially in lymphocytes. Proc Natl Acad Sci U S A 90(16):7568–7572. https://doi.org/10.1073/pnas.90.16.7568 58. Ahmad Mokhtar AMB, Ahmed SBM, Darling NJ, Harris M, Mott HR, Owen D (2021) A complete survey of RhoGDI targets reveals novel interactions with atypical small GTPases. Biochemistry 60(19):1533–1551. https://doi.org/10.1021/acs.biochem.1c00120

304 59. Chuang TH, Bohl BP, Bokoch GM (1993) Biologically active lipids are regulators of Rac.GDI complexation. J Biol Chem 268(35):26206–26211 60. Ugolev Y, Berdichevsky Y, Weinbaum C, Pick E (2008) Dissociation of Rac1(GDP).RhoGDI complexes by the cooperative action of anionic liposomes containing phosphatidylinositol 3,4,5trisphosphate, Rac guanine nucleotide exchange factor, and GTP. J Biol Chem 283(32):22257–22271. https://doi.org/10.1074/jbc. M800734200 61. Ugolev Y, Molshanski-Mor S, Weinbaum C, Pick E (2006) Liposomes comprising anionic but not neutral phospholipids cause dissociation of Rac (1 or 2) x RhoGDI complexes and support amphiphile-independent NADPH oxidase activation by such complexes. J Biol Chem 281(28):19204–19219. https://doi. org/10.1074/jbc.M600042200 62. DerMardirossian C, Schnelzer A, Bokoch GM (2004) Phosphorylation of RhoGDI by Pak1 mediates dissociation of Rac GTPase. Mol Cell 15(1):117–127. https://doi.org/10.1016/j.molcel.2004. 05.019 63. Xiao Y, Lin VY, Ke S, Lin GE, Lin FT, Lin WC (2014) 14-3-3tau promotes breast cancer invasion and metastasis by inhibiting RhoGDIalpha. Mol Cell Biol 34(14):2635–2649. https://doi.org/ 10.1128/MCB.00076-14 64. Pu J, Mao Y, Lei X, Yan Y, Lu X, Tian J et al (2013) FERM domain containing protein 7 interacts with the Rho GDP dissociation inhibitor and specifically activates Rac1 signaling. PLoS One 8(8):e73108. https://doi.org/10.1371/journal.pone.0073108 65. Lu Y, Liu X, Zhou J, Huang A, Zhou J, He C (2013) TROY interacts with Rho guanine nucleotide dissociation inhibitor alpha (RhoGDIalpha) to mediate Nogo-induced inhibition of neurite outgrowth. J Biol Chem 288(47):34276–34286. https://doi.org/ 10.1074/jbc.M113.519744 66. Cho HJ, Hwang YS, Yoon J, Lee M, Lee HG, Daar IO (2018) EphrinB1 promotes cancer cell migration and invasion through the interaction with RhoGDI1. Oncogene 37(7):861–872. https://doi. org/10.1038/onc.2017.386 67. Roberts PJ, Mitin N, Keller PJ, Chenette EJ, Madigan JP, Currin RO et al (2008) Rho family GTPase modification and dependence on CAAX motif-signaled posttranslational modification. J Biol Chem 283(37):25150–25163. https://doi.org/10.1074/jbc. M800882200 68. Winter-Vann AM, Casey PJ (2005) Post-prenylation-processing enzymes as new targets in oncogenesis. Nat Rev Cancer 5(5): 405–412. https://doi.org/10.1038/nrc1612 69. Heyworth PG, Knaus UG, Xu X, Uhlinger DJ, Conroy L, Bokoch GM et al (1993) Requirement for posttranslational processing of Rac GTP-binding proteins for activation of human neutrophil NADPH oxidase. Mol Biol Cell 4(3):261–269. https://doi.org/10. 1091/mbc.4.3.261 70. Kreck ML, Freeman JL, Abo A, Lambeth JD (1996) Membrane association of Rac is required for high activity of the respiratory burst oxidase. Biochemistry 35(49):15683–15692. https://doi.org/ 10.1021/bi962064l 71. Magalhaes MA, Glogauer M (2010) Pivotal advance: phospholipids determine net membrane surface charge resulting in differential localization of active Rac1 and Rac2. J Leukoc Biol 87(4):545–555. https://doi.org/10.1189/jlb.0609390 72. Filippi MD, Harris CE, Meller J, Gu Y, Zheng Y, Williams DA (2004) Localization of Rac2 via the C terminus and aspartic acid 150 specifies superoxide generation, actin polarity and chemotaxis in neutrophils. Nat Immunol 5(7):744–751. https://doi.org/10. 1038/ni1081 73. Yamauchi A, Marchal CC, Molitoris J, Pech N, Knaus U, Towe J et al (2005) Rac GTPase isoform-specific regulation of NADPH oxidase and chemotaxis in murine neutrophils in vivo. Role of the

Y. Lin and Y. Zheng C-terminal polybasic domain. J Biol Chem 280(2):953–964. https://doi.org/10.1074/jbc.M408820200 74. Phuyal S, Farhan H (2019) Multifaceted Rho GTPase Signaling at the Endomembranes. Front Cell Dev Biol 7:127. https://doi.org/10. 3389/fcell.2019.00127 75. Palamidessi A, Frittoli E, Garre M, Faretta M, Mione M, Testa I et al (2008) Endocytic trafficking of Rac is required for the spatial restriction of signaling in cell migration. Cell 134(1):135–147. https://doi.org/10.1016/j.cell.2008.05.034 76. Osmani N, Peglion F, Chavrier P, Etienne-Manneville S (2010) Cdc42 localization and cell polarity depend on membrane traffic. J Cell Biol 191(7):1261–1269. https://doi.org/10.1083/jcb. 201003091 77. Quassollo G, Wojnacki J, Salas DA, Gastaldi L, Marzolo MP, Conde C et al (2015) A RhoA Signaling pathway regulates dendritic Golgi outpost formation. Curr Biol 25(8):971–982. https:// doi.org/10.1016/j.cub.2015.01.075 78. Erickson JW, Zhang C, Kahn RA, Evans T, Cerione RA (1996) Mammalian Cdc42 is a brefeldin A-sensitive component of the Golgi apparatus. J Biol Chem 271(43):26850–26854. https://doi. org/10.1074/jbc.271.43.26850 79. Boivin D, Beliveau R (1995) Subcellular distribution and membrane association of Rho-related small GTP-binding proteins in kidney cortex. Am J Phys 269(2 Pt 2):F180–F189. https://doi.org/ 10.1152/ajprenal.1995.269.2.F180 80. Baldassare JJ, Jarpe MB, Alferes L, Raben DM (1997) Nuclear translocation of RhoA mediates the mitogen-induced activation of phospholipase D involved in nuclear envelope signal transduction. J Biol Chem 272(8):4911–4914. https://doi.org/10.1074/jbc.272.8. 4911 81. Lanning CC, Daddona JL, Ruiz-Velasco R, Shafer SH, Williams CL (2004) The Rac1 C-terminal polybasic region regulates the nuclear localization and protein degradation of Rac1. J Biol Chem 279(42):44197–44210. https://doi.org/10.1074/jbc. M404977200 82. Liu Y, Lv J, Liang X, Yin X, Zhang L, Chen D et al (2018) Fibrin stiffness mediates dormancy of tumor-repopulating cells via a Cdc42-driven Tet 2 epigenetic program. Cancer Res 78(14): 3926–3937. https://doi.org/10.1158/0008-5472.CAN-17-3719 83. Navarro-Lerida I, Sanchez-Alvarez M, Del Pozo MA (2021) Posttranslational modification and subcellular compartmentalization: emerging concepts on the regulation and physiopathological relevance of RhoGTPases. Cell 10(8):1990. https://doi.org/10.3390/ cells10081990 84. Wherlock M, Gampel A, Futter C, Mellor H (2004) Farnesyltransferase inhibitors disrupt EGF receptor traffic through modulation of the RhoB GTPase. J Cell Sci 117(Pt 15):3221–3231. https:// doi.org/10.1242/jcs.01193 85. Navarro-Lerida I, Sanchez-Perales S, Calvo M, Rentero C, Zheng Y, Enrich C et al (2012) A palmitoylation switch mechanism regulates Rac1 function and membrane organization. EMBO J 31(3):534–551. https://doi.org/10.1038/emboj.2011.446 86. Nishimura A, Linder ME (2013) Identification of a novel prenyl and palmitoyl modification at the CaaX motif of Cdc42 that regulates RhoGDI binding. Mol Cell Biol 33(7):1417–1429. https://doi.org/10.1128/MCB.01398-12 87. Greaves J, Chamberlain LH (2007) Palmitoylation-dependent protein sorting. J Cell Biol 176(3):249–254. https://doi.org/10.1083/ jcb.200610151 88. Lanning CC, Ruiz-Velasco R, Williams CL (2003) Novel mechanism of the co-regulation of nuclear transport of Smg GDS and Rac1. J Biol Chem 278(14):12495–12506. https://doi.org/10.1074/ jbc.M211286200 89. Olson MF (2018) Rho GTPases, their post-translational modifications, disease-associated mutations and pharmacological

18

Rho Family GTPases and their Modulators

inhibitors. Small GTPases 9(3):203–215. https://doi.org/10.1080/ 21541248.2016.1218407 90. Ellerbroek SM, Wennerberg K, Burridge K (2003) Serine phosphorylation negatively regulates RhoA in vivo. J Biol Chem 278(21):19023–19031. https://doi.org/10.1074/jbc.M213066200 91. Lang P, Gesbert F, Delespine-Carmagnat M, Stancou R, Pouchelet M, Bertoglio J (1996) Protein kinase A phosphorylation of RhoA mediates the morphological and functional effects of cyclic AMP in cytotoxic lymphocytes. EMBO J 15(3):510–519 92. Sauzeau V, Le Jeune H, Cario-Toumaniantz C, Smolenski A, Lohmann SM, Bertoglio J et al (2000) Cyclic GMP-dependent protein kinase signaling pathway inhibits RhoA-induced Ca2+ sensitization of contraction in vascular smooth muscle. J Biol Chem 275(28):21722–21729. https://doi.org/10.1074/jbc. M000753200 93. Forget MA, Desrosiers RR, Gingras D, Beliveau R (2002) Phosphorylation states of Cdc42 and RhoA regulate their interactions with Rho GDP dissociation inhibitor and their extraction from biological membranes. Biochem J 361(Pt 2):243–254. https://doi. org/10.1042/0264-6021:3610243 94. Tu S, Wu WJ, Wang J, Cerione RA (2003) Epidermal growth factor-dependent regulation of Cdc42 is mediated by the Src tyrosine kinase. J Biol Chem 278(49):49293–49300. https://doi.org/10. 1074/jbc.M307021200 95. Chang F, Lemmon C, Lietha D, Eck M, Romer L (2011) Tyrosine phosphorylation of Rac1: a role in regulation of cell spreading. PLoS One 6(12):e28587. https://doi.org/10.1371/journal.pone. 0028587 96. Tang J, Ip JP, Ye T, Ng YP, Yung WH, Wu Z et al (2014) Cdk5dependent Mst3 phosphorylation and activity regulate neuronal migration through RhoA inhibition. J Neurosci 34(22): 7425–7436. https://doi.org/10.1523/JNEUROSCI.5449-13.2014 97. Kwon T, Kwon DY, Chun J, Kim JH, Kang SS (2000) Akt protein kinase inhibits Rac1-GTP binding through phosphorylation at serine 71 of Rac1. J Biol Chem 275(1):423–428. https://doi.org/10. 1074/jbc.275.1.423 98. Abdrabou A, Brandwein D, Liu C, Wang Z (2019) Rac1 S71 mediates the interaction between Rac1 and 14-3-3 proteins. Cell 8(9):1006. https://doi.org/10.3390/cells8091006 99. Tong J, Li L, Ballermann B, Wang Z (2013) Phosphorylation of Rac1 T108 by extracellular signal-regulated kinase in response to epidermal growth factor: a novel mechanism to regulate Rac1 function. Mol Cell Biol 33(22):4538–4551. https://doi.org/10. 1128/MCB.00822-13 100. Uezu A, Okada H, Murakoshi H, del Vescovo CD, Yasuda R, Diviani D et al (2012) Modified SH2 domain to phototrap and identify phosphotyrosine proteins from subcellular sites within cells. Proc Natl Acad Sci U S A 109(43):E2929–E2938. https:// doi.org/10.1073/pnas.1207358109 101. Chen Y, Yang Z, Meng M, Zhao Y, Dong N, Yan H et al (2009) Cullin mediates degradation of RhoA through evolutionarily conserved BTB adaptors to control actin cytoskeleton structure and cell movement. Mol Cell 35(6):841–855. https://doi.org/10. 1016/j.molcel.2009.09.004 102. Oberoi TK, Dogan T, Hocking JC, Scholz RP, Mooz J, Anderson CL et al (2012) IAPs regulate the plasticity of cell migration by directly targeting Rac1 for degradation. EMBO J 31(1):14–28. https://doi.org/10.1038/emboj.2011.423 103. Torrino S, Visvikis O, Doye A, Boyer L, Stefani C, Munro P et al (2011) The E3 ubiquitin-ligase HACE1 catalyzes the ubiquitylation of active Rac1. Dev Cell 21(5):959–965. https:// doi.org/10.1016/j.devcel.2011.08.015 104. Wei J, Mialki RK, Dong S, Khoo A, Mallampalli RK, Zhao Y et al (2013) A new mechanism of RhoA ubiquitination and degradation: roles of SCF(FBXL19) E3 ligase and Erk2. Biochim Biophys Acta

305 1833(12):2757–2764. https://doi.org/10.1016/j.bbamcr.2013. 07.005 105. Zhao J, Mialki RK, Wei J, Coon TA, Zou C, Chen BB et al (2013) SCF E3 ligase F-box protein complex SCF(FBXL19) regulates cell migration by mediating Rac1 ubiquitination and degradation. FASEB J 27(7):2611–2619. https://doi.org/10.1096/fj.12-223099 106. Castillo-Lluva S, Tatham MH, Jones RC, Jaffray EG, Edmondson RD, Hay RT et al (2010) SUMOylation of the GTPase Rac1 is required for optimal cell migration. Nat Cell Biol 12(11): 1078–1085. https://doi.org/10.1038/ncb2112 107. Worby CA, Mattoo S, Kruger RP, Corbeil LB, Koller A, Mendez JC et al (2009) The fic domain: regulation of cell signaling by adenylylation. Mol Cell 34(1):93–103. https://doi.org/10.1016/j. molcel.2009.03.008 108. Bishop AL, Hall A (2000) Rho GTPases and their effector proteins. Biochem J 348(Pt 2):241–255 109. Bustelo XR, Sauzeau V, Berenjeno IM (2007) GTP-binding proteins of the Rho/Rac family: regulation, effectors and functions in vivo. BioEssays 29(4):356–370. https://doi.org/10.1002/bies. 20558 110. Paul F, Zauber H, von Berg L, Rocks O, Daumke O, Selbach M (2017) Quantitative GTPase affinity purification identifies Rho family protein interaction partners. Mol Cell Proteomics 16(1): 73–85. https://doi.org/10.1074/mcp.M116.061531 111. Bagci H, Sriskandarajah N, Robert A, Boulais J, Elkholi IE, Tran V et al (2020) Mapping the proximity interaction network of the Rho-family GTPases reveals signalling pathways and regulatory mechanisms. Nat Cell Biol 22(1):120–134. https://doi.org/10. 1038/s41556-019-0438-7 112. Jaffe AB, Hall A (2005) Rho GTPases: biochemistry and biology. Annu Rev Cell Dev Biol 21:247–269. https://doi.org/10.1146/ annurev.cellbio.21.020604.150721 113. Julian L, Olson MF (2014) Rho-associated coiled-coil containing kinases (ROCK): structure, regulation, and functions. Small GTPases 5:e29846. https://doi.org/10.4161/sgtp.29846 114. D’Avino PP (2017) Citron kinase - renaissance of a neglected mitotic kinase. J Cell Sci 130(10):1701–1708. https://doi.org/10. 1242/jcs.200253 115. Narumiya S, Tanji M, Ishizaki T (2009) Rho signaling, ROCK and mDia1, in transformation, metastasis and invasion. Cancer Metastasis Rev 28(1–2):65–76. https://doi.org/10.1007/s10555-0089170-7 116. Lammers M, Meyer S, Kuhlmann D, Wittinghofer A (2008) Specificity of interactions between mDia isoforms and Rho proteins. J Biol Chem 283(50):35236–35246. https://doi.org/10.1074/jbc. M805634200 117. Kumar R, Sanawar R, Li X, Li F (2017) Structure, biochemistry, and biology of PAK kinases. Gene 605:20–31. https://doi.org/10. 1016/j.gene.2016.12.014 118. Molli PR, Li DQ, Murray BW, Rayala SK, Kumar R (2009) PAK signaling in oncogenesis. Oncogene 28(28):2545–2555. https:// doi.org/10.1038/onc.2009.119 119. Watanabe T, Wang S, Kaibuchi K (2015) IQGAPs as key regulators of actin-cytoskeleton dynamics. Cell Struct Funct 40(2):69–77. https://doi.org/10.1247/csf.15003 120. Scita G, Confalonieri S, Lappalainen P, Suetsugu S (2008) IRSp53: crossing the road of membrane and actin dynamics in the formation of membrane protrusions. Trends Cell Biol 18(2):52–60. https:// doi.org/10.1016/j.tcb.2007.12.002 121. Henrique D, Schweisguth F (2003) Cell polarity: the ups and downs of the Par6/aPKC complex. Curr Opin Genet Dev 13(4): 341–350. https://doi.org/10.1016/s0959-437x(03)00077-7 122. Zhao Z, Manser E (2015) Myotonic dystrophy kinase-related Cdc42-binding kinases (MRCK), the ROCK-like effectors of Cdc42 and Rac1. Small GTPases 6(2):81–88. https://doi.org/10. 1080/21541248.2014.1000699

306 123. Unbekandt M, Olson MF (2014) The actin-myosin regulatory MRCK kinases: regulation, biological functions and associations with human cancer. J Mol Med (Berl) 92(3):217–225. https://doi. org/10.1007/s00109-014-1133-6 124. Hotta K, Tanaka K, Mino A, Kohno H, Takai Y (1996) Interaction of the Rho family small G proteins with kinectin, an anchoring protein of kinesin motor. Biochem Biophys Res Commun 225(1): 69–74. https://doi.org/10.1006/bbrc.1996.1132 125. Vignal E, Blangy A, Martin M, Gauthier-Rouviere C, Fort P (2001) Kinectin is a key effector of RhoG microtubule-dependent cellular activity. Mol Cell Biol 21(23):8022–8034. https://doi.org/10.1128/ MCB.21.23.8022-8034.2001 126. Mukai H (2003) The structure and function of PKN, a protein kinase having a catalytic domain homologous to that of PKC. J Biochem 133(1):17–27. https://doi.org/10.1093/jb/mvg019 127. Slater SJ, Seiz JL, Stagliano BA, Stubbs CD (2001) Interaction of protein kinase C isozymes with Rho GTPases. Biochemistry 40(14):4437–4445. https://doi.org/10.1021/bi001654n 128. Singh RK, Kumar S, Gautam PK, Tomar MS, Verma PK, Singh SP et al (2017) Protein kinase C-alpha and the regulation of diverse cell responses. Biomol Concepts 8(3–4):143–153. https://doi.org/ 10.1515/bmc-2017-0005 129. Hill CS, Wynne J, Treisman R (1995) The Rho family GTPases RhoA, Rac1, and CDC42Hs regulate transcriptional activation by SRF. Cell 81(7):1159–1170. https://doi.org/10.1016/s0092-8674 (05)80020-0 130. Wing MR, Snyder JT, Sondek J, Harden TK (2003) Direct activation of phospholipase C-epsilon by Rho. J Biol Chem 278(42): 41253–41258. https://doi.org/10.1074/jbc.M306904200 131. Houssa B, de Widt J, Kranenburg O, Moolenaar WH, van Blitterswijk WJ (1999) Diacylglycerol kinase theta binds to and is negatively regulated by active RhoA. J Biol Chem 274(11): 6820–6822. https://doi.org/10.1074/jbc.274.11.6820 132. Tibbles LA, Ing YL, Kiefer F, Chan J, Iscove N, Woodgett JR et al (1996) MLK-3 activates the SAPK/JNK and p38/RK pathways via SEK1 and MKK3/6. EMBO J 15(24):7026–7035 133. Fanger GR, Johnson NL, Johnson GL (1997) MEK kinases are regulated by EGF and selectively interact with Rac/Cdc42. EMBO J 16(16):4961–4972. https://doi.org/10.1093/emboj/16.16.4961 134. Chou MM, Blenis J (1996) The 70 kDa S6 kinase complexes with and is activated by the Rho family G proteins Cdc42 and Rac1. Cell 85(4):573–583. https://doi.org/10.1016/s0092-8674(00)81257-x 135. Zheng Y, Bagrodia S, Cerione RA (1994) Activation of phosphoinositide 3-kinase activity by Cdc42Hs binding to p85. J Biol Chem 269(29):18727–18730 136. Illenberger D, Schwald F, Pimmer D, Binder W, Maier G, Dietrich A et al (1998) Stimulation of phospholipase C-beta2 by the Rho GTPases Cdc42Hs and Rac1. EMBO J 17(21):6241–6249. https:// doi.org/10.1093/emboj/17.21.6241 137. Diekmann D, Abo A, Johnston C, Segal AW, Hall A (1994) Interaction of Rac with p67phox and regulation of phagocytic NADPH oxidase activity. Science 265(5171):531–533. https:// doi.org/10.1126/science.8036496 138. Eden S, Rohatgi R, Podtelejnikov AV, Mann M, Kirschner MW (2002) Mechanism of regulation of WAVE1-induced actin nucleation by Rac1 and Nck. Nature 418(6899):790–793. https://doi. org/10.1038/nature00859 139. Mukai H, Kitagawa M, Shibata H, Takanaga H, Mori K, Shimakawa M et al (1994) Activation of PKN, a novel 120-kDa protein kinase with leucine zipper-like sequences, by unsaturated fatty acids and by limited proteolysis. Biochem Biophys Res Commun 204(1):348–356. https://doi.org/10.1006/bbrc.1994.2466 140. Chong C, Tan L, Lim L, Manser E (2001) The mechanism of PAK activation. Autophosphorylation events in both regulatory and kinase domains control activity. J Biol Chem 276(20): 17347–17353. https://doi.org/10.1074/jbc.M009316200

Y. Lin and Y. Zheng 141. Joberty G, Perlungher RR, Sheffield PJ, Kinoshita M, Noda M, Haystead T et al (2001) Borg proteins control septin organization and are negatively regulated by Cdc42. Nat Cell Biol 3(10): 861–866. https://doi.org/10.1038/ncb1001-861 142. El-Benna J, Hurtado-Nedelec M, Marzaioli V, Marie JC, Gougerot-Pocidalo MA, Dang PM (2016) Priming of the neutrophil respiratory burst: role in host defense and inflammation. Immunol Rev 273(1):180–193. https://doi.org/10.1111/imr.12447 143. Panday A, Sahoo MK, Osorio D, Batra S (2015) NADPH oxidases: an overview from structure to innate immunity-associated pathologies. Cell Mol Immunol 12(1):5–23. https://doi.org/10. 1038/cmi.2014.89 144. Pick E (2014) Role of the Rho GTPase Rac in the activation of the phagocyte NADPH oxidase: outsourcing a key task. Small GTPases 5:e27952. https://doi.org/10.4161/sgtp.27952 145. Nauseef WM (2019) The phagocyte NOX2 NADPH oxidase in microbial killing and cell signaling. Curr Opin Immunol 60:130– 140. https://doi.org/10.1016/j.coi.2019.05.006 146. Nguyen GT, Green ER, Mecsas J (2017) Neutrophils to the ROScue: mechanisms of NADPH oxidase activation and bacterial resistance. Front Cell Infect Microbiol 7:373. https://doi.org/10. 3389/fcimb.2017.00373 147. Nunes P, Demaurex N, Dinauer MC (2013) Regulation of the NADPH oxidase and associated ion fluxes during phagocytosis. Traffic 14(11):1118–1131. https://doi.org/10.1111/tra.12115 148. Ziegler CS, Bouchab L, Tramier M, Durand D, Fieschi F, DupreCrochet S et al (2019) Quantitative live-cell imaging and 3D modeling reveal critical functional features in the cytosolic complex of phagocyte NADPH oxidase. J Biol Chem 294(11): 3824–3836. https://doi.org/10.1074/jbc.RA118.006864 149. Abo A, Pick E (1991) Purification and characterization of a third cytosolic component of the superoxide-generating NADPH oxidase of macrophages. J Biol Chem 266(35):23577–23585 150. Abo A, Pick E, Hall A, Totty N, Teahan CG, Segal AW (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1. Nature 353(6345):668–670. https://doi.org/10. 1038/353668a0 151. Knaus UG, Heyworth PG, Evans T, Curnutte JT, Bokoch GM (1991) Regulation of phagocyte oxygen radical production by the GTP-binding protein Rac 2. Science 254(5037):1512–1515. https://doi.org/10.1126/science.1660188 152. Knaus UG, Heyworth PG, Kinsella BT, Curnutte JT, Bokoch GM (1992) Purification and characterization of Rac 2. A cytosolic GTP-binding protein that regulates human neutrophil NADPH oxidase. J Biol Chem 267(33):23575–23582 153. Pick E, Gorzalczany Y, Engel S (1993) Role of the rac 1 p21-GDPdissociation inhibitor for rho heterodimer in the activation of the superoxide-forming NADPH oxidase of macrophages. Eur J Biochem 217(1):441–455. https://doi.org/10.1111/j.1432-1033. 1993.tb18264.x 154. Diebold BA, Bokoch GM (2001) Molecular basis for Rac2 regulation of phagocyte NADPH oxidase. Nat Immunol 2(3):211–215. https://doi.org/10.1038/85259 155. Diekmann D, Nobes CD, Burbelo PD, Abo A, Hall A (1995) Rac GTPase interacts with GAPs and target proteins through multiple effector sites. EMBO J 14(21):5297–5305 156. Toporik A, Gorzalczany Y, Hirshberg M, Pick E, Lotan O (1998) Mutational analysis of novel effector domains in Rac1 involved in the activation of nicotinamide adenine dinucleotide phosphate (reduced) oxidase. Biochemistry 37(20):7147–7156. https://doi. org/10.1021/bi9800404 157. Joseph G, Pick E (1995) “Peptide walking” is a novel method for mapping functional domains in proteins. Its application to the Rac1-dependent activation of NADPH oxidase. J Biol Chem 270(49):29079–29082. https://doi.org/10.1074/jbc.270.49.29079

18

Rho Family GTPases and their Modulators

158. Koga H, Terasawa H, Nunoi H, Takeshige K, Inagaki F, Sumimoto H (1999) Tetratricopeptide repeat (TPR) motifs of p67(phox) participate in interaction with the small GTPase Rac and activation of the phagocyte NADPH oxidase. J Biol Chem 274(35): 25051–25060. https://doi.org/10.1074/jbc.274.35.25051 159. Diebold BA, Fowler B, Lu J, Dinauer MC, Bokoch GM (2004) Antagonistic cross-talk between Rac and Cdc42 GTPases regulates generation of reactive oxygen species. J Biol Chem 279(27): 28136–28142. https://doi.org/10.1074/jbc.M313891200 160. Kao YY, Gianni D, Bohl B, Taylor RM, Bokoch GM (2008) Identification of a conserved Rac-binding site on NADPH oxidases supports a direct GTPase regulatory mechanism. J Biol Chem 283(19):12736–12746. https://doi.org/10.1074/jbc.M801010200 161. Heyworth PG, Bohl BP, Bokoch GM, Curnutte JT (1994) Rac translocates independently of the neutrophil NADPH oxidase components p47phox and p67phox. Evidence for its interaction with flavocytochrome b558. J Biol Chem 269(49):30749–30752 162. Gorzalczany Y, Sigal N, Itan M, Lotan O, Pick E (2000) Targeting of Rac1 to the phagocyte membrane is sufficient for the induction of NADPH oxidase assembly. J Biol Chem 275(51):40073–40081. https://doi.org/10.1074/jbc.M006013200 163. Alloul N, Gorzalczany Y, Itan M, Sigal N, Pick E (2001) Activation of the superoxide-generating NADPH oxidase by chimeric proteins consisting of segments of the cytosolic component p67 (phox) and the small GTPase Rac1. Biochemistry 40(48): 14557–14566. https://doi.org/10.1021/bi0117347 164. Berdichevsky Y, Mizrahi A, Ugolev Y, Molshanski-Mor S, Pick E (2007) Tripartite chimeras comprising functional domains derived from the cytosolic NADPH oxidase components p47phox, p67phox, and Rac1 elicit activator-independent superoxide production by phagocyte membranes: an essential role for anionic membrane phospholipids. J Biol Chem 282(30):22122–22139. https://doi.org/10.1074/jbc.M701497200 165. Gorzalczany Y, Alloul N, Sigal N, Weinbaum C, Pick E (2002) A prenylated p67phox-Rac1 chimera elicits NADPH-dependent superoxide production by phagocyte membranes in the absence of an activator and of p47phox: conversion of a pagan NADPH oxidase to monotheism. J Biol Chem 277(21):18605–18610. https://doi.org/10.1074/jbc.M202114200 166. Mizrahi A, Berdichevsky Y, Casey PJ, Pick E (2010) A prenylated p47phox-p67phox-Rac1 chimera is a quintessential NADPH oxidase activator: membrane association and functional capacity. J Biol Chem 285(33):25485–25499. https://doi.org/10.1074/jbc. M110.113779 167. Mizrahi A, Berdichevsky Y, Ugolev Y, Molshanski-Mor S, Nakash Y, Dahan I et al (2006) Assembly of the phagocyte NADPH oxidase complex: chimeric constructs derived from the cytosolic components as tools for exploring structure-function relationships. J Leukoc Biol 79(5):881–895. https://doi.org/10. 1189/jlb.1005553 168. Sarfstein R, Gorzalczany Y, Mizrahi A, Berdichevsky Y, Molshanski-Mor S, Weinbaum C et al (2004) Dual role of Rac in the assembly of NADPH oxidase, tethering to the membrane and activation of p67phox: a study based on mutagenesis of p67phoxRac1 chimeras. J Biol Chem 279(16):16007–16016. https://doi. org/10.1074/jbc.M312394200 169. Bechor E, Zahavi A, Berdichevsky Y, Pick E (2021) The molecular basis of Rac-GTP action-promoting binding of p67(phox) to Nox2 by disengaging the beta hairpin from downstream residues. J Leukoc Biol 110(2):219–237. https://doi.org/10.1002/JLB. 4HI1220-855RR 170. Bokoch GM, Diebold B, Kim JS, Gianni D (2009) Emerging evidence for the importance of phosphorylation in the regulation of NADPH oxidases. Antioxid Redox Signal 11(10):2429–2441. https://doi.org/10.1089/ARS.2009.2590

307 171. Bokoch GM, Zhao T (2006) Regulation of the phagocyte NADPH oxidase by Rac GTPase. Antioxid Redox Signal 8(9–10): 1533–1548. https://doi.org/10.1089/ars.2006.8.1533 172. Pantarelli C, Welch HCE (2018) Rac-GTPases and Rac-GEFs in neutrophil adhesion, migration and recruitment. Eur J Clin Investig 48(Suppl 2):e12939. https://doi.org/10.1111/eci.12939 173. Csepanyi-Komi R, Pasztor M, Bartos B, Ligeti E (2018) The neglected terminators: Rho family GAPs in neutrophils. Eur J Clin Investig 48(Suppl 2):e12993. https://doi.org/10.1111/eci. 12993 174. McCormick B, Chu JY, Vermeren S (2019) Cross-talk between Rho GTPases and PI3K in the neutrophil. Small GTPases 10(3): 187–195. https://doi.org/10.1080/21541248.2017.1304855 175. Taglieri DM, Ushio-Fukai M, Monasky MM (2014) P21-activated kinase in inflammatory and cardiovascular disease. Cell Signal 26(9):2060–2069. https://doi.org/10.1016/j.cellsig.2014.04.020 176. Knaus UG, Morris S, Dong HJ, Chernoff J, Bokoch GM (1995) Regulation of human leukocyte p21-activated kinases through G protein-coupled receptors. Science 269(5221):221–223. https:// doi.org/10.1126/science.7618083 177. Martyn KD, Kim MJ, Quinn MT, Dinauer MC, Knaus UG (2005) p21-activated kinase (Pak) regulates NADPH oxidase activation in human neutrophils. Blood 106(12):3962–3969. https://doi.org/10. 1182/blood-2005-03-0859 178. Daniels RH, Hall PS, Bokoch GM (1998) Membrane targeting of p21-activated kinase 1 (PAK1) induces neurite outgrowth from PC12 cells. EMBO J 17(3):754–764. https://doi.org/10.1093/ emboj/17.3.754 179. Ahmed S, Prigmore E, Govind S, Veryard C, Kozma R, Wientjes FB et al (1998) Cryptic Rac-binding and p21 (Cdc42Hs/Rac)activated kinase phosphorylation sites of NADPH oxidase component p67(phox). J Biol Chem 273(25):15693–15701. https://doi. org/10.1074/jbc.273.25.15693 180. van Bruggen R, Anthony E, Fernandez-Borja M, Roos D (2004) Continuous translocation of Rac2 and the NADPH oxidase component p67(phox) during phagocytosis. J Biol Chem 279(10): 9097–9102. https://doi.org/10.1074/jbc.M309284200 181. Vetter IR, Wittinghofer A (2001) The guanine nucleotide-binding switch in three dimensions. Science 294(5545):1299–1304. https:// doi.org/10.1126/science.1062023 182. Geyer M, Schweins T, Herrmann C, Prisner T, Wittinghofer A, Kalbitzer HR (1996) Conformational transitions in p21ras and in its complexes with the effector protein Raf-RBD and the GTPase activating protein GAP. Biochemistry 35(32):10308–10320. https://doi.org/10.1021/bi952858k 183. Spoerner M, Herrmann C, Vetter IR, Kalbitzer HR, Wittinghofer A (2001) Dynamic properties of the Ras switch I region and its importance for binding to effectors. Proc Natl Acad Sci U S A 98(9):4944–4949. https://doi.org/10.1073/pnas.081441398 184. Spoerner M, Nuehs A, Ganser P, Herrmann C, Wittinghofer A, Kalbitzer HR (2005) Conformational states of Ras complexed with the GTP analogue Gpp NHp or GppCH2p: implications for the interaction with effector proteins. Biochemistry 44(6):2225–2236. https://doi.org/10.1021/bi0488000 185. Shima F, Ijiri Y, Muraoka S, Liao J, Ye M, Araki M et al (2010) Structural basis for conformational dynamics of GTP-bound Ras protein. J Biol Chem 285(29):22696–22705. https://doi.org/10. 1074/jbc.M110.125161 186. Loh AP, Guo W, Nicholson LK, Oswald RE (1999) Backbone dynamics of inactive, active, and effector-bound Cdc42Hs from measurements of (15) N relaxation parameters at multiple field strengths. Biochemistry 38(39):12547–12557. https://doi.org/10. 1021/bi9913707 187. Abdul-Manan N, Aghazadeh B, Liu GA, Majumdar A, Ouerfelli O, Siminovitch KA et al (1999) Structure of Cdc42 in complex with

308 the GTPase-binding domain of the 'Wiskott-Aldrich syndrome' protein. Nature 399(6734):379–383. https://doi.org/10.1038/20726 188. Morreale A, Venkatesan M, Mott HR, Owen D, Nietlispach D, Lowe PN et al (2000) Structure of Cdc42 bound to the GTPase binding domain of PAK. Nat Struct Biol 7(5):384–388. https://doi. org/10.1038/75158 189. Mott HR, Owen D, Nietlispach D, Lowe PN, Manser E, Lim L et al (1999) Structure of the small G protein Cdc42 bound to the GTPase-binding domain of ACK. Nature 399(6734):384–388. https://doi.org/10.1038/20732 190. Phillips MJ, Calero G, Chan B, Ramachandran S, Cerione RA (2008) Effector proteins exert an important influence on the signaling-active state of the small GTPase Cdc42. J Biol Chem 283(20):14153–14164. https://doi.org/10.1074/jbc.M706271200 191. Bunney TD, Opaleye O, Roe SM, Vatter P, Baxendale RW, Walliser C et al (2009) Structural insights into formation of an active signaling complex between Rac and phospholipase C gamma 2. Mol Cell 34(2):223–233. https://doi.org/10.1016/j. molcel.2009.02.023 192. Hirshberg M, Stockley RW, Dodson G, Webb MR (1997) The crystal structure of human rac1, a member of the rho-family complexed with a GTP analogue. Nat Struct Biol 4(2):147–152. https://doi.org/10.1038/nsb0297-147 193. Dias SM, Cerione RA (2007) X-ray crystal structures reveal two activated states for RhoC. Biochemistry 46(22):6547–6558. https:// doi.org/10.1021/bi700035p 194. Lin Y, Lu S, Zhang J, Zheng Y (2021) Structure of an inactive conformation of GTP-bound RhoA GTPase. Structure 29(6): 553–63 e5. https://doi.org/10.1016/j.str.2020.12.015 195. Lu S, Jang H, Muratcioglu S, Gursoy A, Keskin O, Nussinov R et al (2016) Ras conformational ensembles, allostery, and signaling. Chem Rev 116(11):6607–6665. https://doi.org/10.1021/acs. chemrev.5b00542 196. Mott HR, Owen D (2018) Allostery and dynamics in small G proteins. Biochem Soc Trans 46(5):1333–1343. https://doi.org/ 10.1042/BST20170569 197. Komander D, Garg R, Wan PT, Ridley AJ, Barford D (2008) Mechanism of multi-site phosphorylation from a ROCK-I: RhoE complex structure. EMBO J 27(23):3175–3185. https://doi.org/10. 1038/emboj.2008.226 198. Sigal N, Gorzalczany Y, Pick E (2003) Two pathways of activation of the superoxide-generating NADPH oxidase of phagocytes in vitro--distinctive effects of inhibitors. Inflammation 27(3): 147–159. https://doi.org/10.1023/a:1023869828688 199. Mott HR, Owen D (2015) Structures of Ras superfamily effector complexes: what have we learnt in two decades? Crit Rev Biochem Mol Biol 50(2):85–133. https://doi.org/10.3109/10409238.2014. 999191 200. Lapouge K, Smith SJ, Walker PA, Gamblin SJ, Smerdon SJ, Rittinger K (2000) Structure of the TPR domain of p67phox in complex with Rac. GTP Mol Cell 6(4):899–907. https://doi.org/10. 1016/s1097-2765(05)00091-2 201. Grizot S, Fieschi F, Dagher MC, Pebay-Peyroula E (2001) The active N-terminal region of p67phox. Structure at 1.8 a resolution and biochemical characterizations of the A128V mutant implicated in chronic granulomatous disease. J Biol Chem 276(24): 21627–21631. https://doi.org/10.1074/jbc.M100893200 202. Yang J, Zhang Z, Roe SM, Marshall CJ, Barford D (2009) Activation of Rho GTPases by DOCK exchange factors is mediated by a nucleotide sensor. Science 325(5946):1398–1402. https://doi.org/ 10.1126/science.1174468 203. Kulkarni K, Yang J, Zhang Z, Barford D (2011) Multiple factors confer specific Cdc42 and Rac protein activation by dedicator of cytokinesis (DOCK) nucleotide exchange factors. J Biol Chem 286(28):25341–25351. https://doi.org/10.1074/jbc.M111.236455

Y. Lin and Y. Zheng 204. Wu X, Ramachandran S, Lin MC, Cerione RA, Erickson JW (2011) A minimal Rac activation domain in the unconventional guanine nucleotide exchange factor Dock180. Biochemistry 50(6): 1070–1080. https://doi.org/10.1021/bi100971y 205. Kukimoto-Niino M, Tsuda K, Ihara K, Mishima-Tsumagari C, Honda K, Ohsawa N et al (2019) Structural basis for the dual substrate specificity of DOCK7 guanine nucleotide exchange factor. Structure 27(5):741–8 e3. https://doi.org/10.1016/j.str.2019. 02.001 206. Toyama Y, Kontani K, Katada T, Shimada I (2019) Conformational landscape alternations promote oncogenic activities of Ras-related C3 botulinum toxin substrate 1 as revealed by NMR. Sci Adv 5(3):eaav8945. https://doi.org/10.1126/sciadv.aav8945 207. Pick E (2020) Cell-free NADPH oxidase activation assays: a triumph of reductionism. Methods Mol Biol 2087:325–411. https:// doi.org/10.1007/978-1-0716-0154-9_23 208. Thumkeo D, Watanabe S, Narumiya S (2013) Physiological roles of Rho and Rho effectors in mammals. Eur J Cell Biol 92(10–11): 303–315. https://doi.org/10.1016/j.ejcb.2013.09.002 209. Melendez J, Grogg M, Zheng Y (2011) Signaling role of Cdc42 in regulating mammalian physiology. J Biol Chem 286(4): 2375–2381. https://doi.org/10.1074/jbc.R110.200329 210. Wang L, Zheng Y (2007) Cell type-specific functions of Rho GTPases revealed by gene targeting in mice. Trends Cell Biol 17(2):58–64. https://doi.org/10.1016/j.tcb.2006.11.009 211. Zhou X, Zheng Y (2013) Cell type-specific signaling function of RhoA GTPase: lessons from mouse gene targeting. J Biol Chem 288(51):36179–36188. https://doi.org/10.1074/jbc.R113.515486 212. McCarty OJ, Larson MK, Auger JM, Kalia N, Atkinson BT, Pearce AC et al (2005) Rac1 is essential for platelet lamellipodia formation and aggregate stability under flow. J Biol Chem 280(47): 39474–39484. https://doi.org/10.1074/jbc.M504672200 213. Nodari A, Zambroni D, Quattrini A, Court FA, D'Urso A, Recchia A et al (2007) Beta1 integrin activates Rac1 in Schwann cells to generate radial lamellae during axonal sorting and myelination. J Cell Biol 177(6):1063–1075. https://doi.org/10.1083/jcb. 200610014 214. Tan W, Palmby TR, Gavard J, Amornphimoltham P, Zheng Y, Gutkind JS (2008) An essential role for Rac1 in endothelial cell function and vascular development. FASEB J 22(6):1829–1838. https://doi.org/10.1096/fj.07-096438 215. Chen L, Liao G, Waclaw RR, Burns KA, Linquist D, Campbell K et al (2007) Rac1 controls the formation of midline commissures and the competency of tangential migration in ventral telencephalic neurons. J Neurosci 27(14):3884–3893. https://doi.org/10.1523/ JNEUROSCI.3509-06.2007 216. Satoh M, Ogita H, Takeshita K, Mukai Y, Kwiatkowski DJ, Liao JK (2006) Requirement of Rac1 in the development of cardiac hypertrophy. Proc Natl Acad Sci U S A 103(19):7432–7437. https://doi.org/10.1073/pnas.0510444103 217. Wang Y, Lebowitz D, Sun C, Thang H, Grynpas MD, Glogauer M (2008) Identifying the relative contributions of Rac1 and Rac2 to osteoclastogenesis. J Bone Miner Res 23(2):260–270. https://doi. org/10.1359/jbmr.071013 218. Wells CM, Walmsley M, Ooi S, Tybulewicz V, Ridley AJ (2004) Rac1-deficient macrophages exhibit defects in cell spreading and membrane ruffling but not migration. J Cell Sci 117(Pt 7): 1259–1268. https://doi.org/10.1242/jcs.00997 219. Wheeler AP, Wells CM, Smith SD, Vega FM, Henderson RB, Tybulewicz VL et al (2006) Rac1 and Rac2 regulate macrophage morphology but are not essential for migration. J Cell Sci 119 (Pt 13):2749–2757. https://doi.org/10.1242/jcs.03024 220. Yamauchi A, Kim C, Li S, Marchal CC, Towe J, Atkinson SJ et al (2004) Rac2-deficient murine macrophages have selective defects in superoxide production and phagocytosis of opsonized particles.

18

Rho Family GTPases and their Modulators

J Immunol 173(10):5971–5979. https://doi.org/10.4049/jimmunol. 173.10.5971 221. Sun CX, Downey GP, Zhu F, Koh AL, Thang H, Glogauer M (2004) Rac1 is the small GTPase responsible for regulating the neutrophil chemotaxis compass. Blood 104(12):3758–3765. https://doi.org/10.1182/blood-2004-03-0781 222. Condliffe AM, Webb LM, Ferguson GJ, Davidson K, Turner M, Vigorito E et al (2006) RhoG regulates the neutrophil NADPH oxidase. J Immunol 176(9):5314–5320. https://doi.org/10.4049/ jimmunol.176.9.5314 223. Kim C, Dinauer MC (2001) Rac2 is an essential regulator of neutrophil nicotinamide adenine dinucleotide phosphate oxidase activation in response to specific signaling pathways. J Immunol 166(2):1223–1232. https://doi.org/10.4049/jimmunol.166.2.1223 224. Glogauer M, Marchal CC, Zhu F, Worku A, Clausen BE, Foerster I et al (2003) Rac1 deletion in mouse neutrophils has selective effects on neutrophil functions. J Immunol 170(11):5652–5657. https://doi.org/10.4049/jimmunol.170.11.5652 225. Katoh H, Negishi M (2003) RhoG activates Rac1 by direct interaction with the Dock180-binding protein Elmo. Nature 424(6947): 461–464. https://doi.org/10.1038/nature01817 226. Abdel-Latif D, Steward M, Macdonald DL, Francis GA, Dinauer MC, Lacy P (2004) Rac2 is critical for neutrophil primary granule exocytosis. Blood 104(3):832–839. https://doi.org/10.1182/blood2003-07-2624 227. Gu Y, Filippi MD, Cancelas JA, Siefring JE, Williams EP, Jasti AC et al (2003) Hematopoietic cell regulation by Rac1 and Rac2 guanosine triphosphatases. Science 302(5644):445–449. https:// doi.org/10.1126/science.1088485 228. Cancelas JA, Lee AW, Prabhakar R, Stringer KF, Zheng Y, Williams DA (2005) Rac GTPases differentially integrate signals regulating hematopoietic stem cell localization. Nat Med 11(8): 886–891. https://doi.org/10.1038/nm1274 229. Yu H, Leitenberg D, Li B, Flavell RA (2001) Deficiency of small GTPase Rac2 affects T cell activation. J Exp Med 194(7):915–926. https://doi.org/10.1084/jem.194.7.915 230. Li B, Yu H, Zheng W, Voll R, Na S, Roberts AW et al (2000) Role of the guanosine triphosphatase Rac2 in T helper 1 cell differentiation. Science 288(5474):2219–2222. https://doi.org/10.1126/sci ence.288.5474.2219 231. Guo F, Cancelas JA, Hildeman D, Williams DA, Zheng Y (2008) Rac GTPase isoforms Rac1 and Rac2 play a redundant and crucial role in T-cell development. Blood 112(5):1767–1775. https://doi. org/10.1182/blood-2008-01-132068 232. Croker BA, Tarlinton DM, Cluse LA, Tuxen AJ, Light A, Yang FC et al (2002) The Rac2 guanosine triphosphatase regulates B lymphocyte antigen receptor responses and chemotaxis and is required for establishment of B-1a and marginal zone B lymphocytes. J Immunol 168(7):3376–3386. https://doi.org/10.4049/jimmunol. 168.7.3376 233. Walmsley MJ, Ooi SK, Reynolds LF, Smith SH, Ruf S, Mathiot A et al (2003) Critical roles for Rac1 and Rac2 GTPases in B cell development and signaling. Science 302(5644):459–462. https:// doi.org/10.1126/science.1089709 234. Goggs R, Harper MT, Pope RJ, Savage JS, Williams CM, Mundell SJ et al (2013) RhoG protein regulates platelet granule secretion and thrombus formation in mice. J Biol Chem 288(47): 34217–34229. https://doi.org/10.1074/jbc.M113.504100 235. Bustelo XR (2018) RHO GTPases in cancer: known facts, open questions, and therapeutic challenges. Biochem Soc Trans 46(3): 741–760. https://doi.org/10.1042/BST20170531 236. Crosas-Molist E, Samain R, Kohlhammer L, Orgaz J, George S, Maiques O et al (2021) RhoGTPase signalling in cancer progression and dissemination. Physiol Rev 102(1):455–510. https://doi. org/10.1152/physrev.00045.2020

309 237. Haga RB, Ridley AJ (2016) Rho GTPases: regulation and roles in cancer cell biology. Small GTPases 7(4):207–221. https://doi.org/ 10.1080/21541248.2016.1232583 238. Li H, Peyrollier K, Kilic G, Brakebusch C (2014) Rho GTPases and cancer. Biofactors 40(2):226–235. https://doi.org/10.1002/ biof.1155 239. Porter AP, Papaioannou A, Malliri A (2016) Deregulation of Rho GTPases in cancer. Small GTPases 7:1–16. https://doi.org/10. 1080/21541248.2016.1173767 240. Vega FM, Ridley AJ (2008) Rho GTPases in cancer cell biology. FEBS Lett 582(14):2093–2101. https://doi.org/10.1016/j.febslet. 2008.04.039 241. Aspenstrom P (2018) Activated Rho GTPases in cancer-the beginning of a new paradigm. Int J Mol Sci 19(12):3949. https://doi.org/ 10.3390/ijms19123949 242. Jung H, Yoon SR, Lim J, Cho HJ, Lee HG (2020) Dysregulation of Rho GTPases in human cancers. Cancers (Basel) 12(5):1179. https://doi.org/10.3390/cancers12051179 243. Hodge RG, Schaefer A, Howard SV, Der CJ (2020) RAS and RHO family GTPase mutations in cancer: twin sons of different mothers? Crit Rev Biochem Mol Biol 55(4):386–407. https://doi.org/10. 1080/10409238.2020.1810622 244. Zandvakili I, Lin Y, Morris JC, Zheng Y (2017) Rho GTPases: anti- or pro-neoplastic targets? Oncogene 36(23):3213–3222. https://doi.org/10.1038/onc.2016.473 245. Svensmark JH, Brakebusch C (2019) Rho GTPases in cancer: friend or foe? Oncogene 38(50):7447–7456. https://doi.org/10. 1038/s41388-019-0963-7 246. Davis MJ, Ha BH, Holman EC, Halaban R, Schlessinger J, Boggon TJ (2013) RAC1P29S is a spontaneously activating cancerassociated GTPase. Proc Natl Acad Sci U S A 110(3):912–917. https://doi.org/10.1073/pnas.1220895110 247. Krauthammer M, Kong Y, Ha BH, Evans P, Bacchiocchi A, McCusker JP et al (2012) Exome sequencing identifies recurrent somatic RAC1 mutations in melanoma. Nat Genet 44(9): 1006–1014. https://doi.org/10.1038/ng.2359 248. Kazanietz MG, Caloca MJ (2017) The Rac GTPase in cancer: from old concepts to new paradigms. Cancer Res 77(20):5445–5451. https://doi.org/10.1158/0008-5472.CAN-17-1456 249. Radu M, Semenova G, Kosoff R, Chernoff J (2014) PAK signalling during the development and progression of cancer. Nat Rev Cancer 14(1):13–25 250. Shain AH, Garrido M, Botton T, Talevich E, Yeh I, Sanborn JZ et al (2015) Exome sequencing of desmoplastic melanoma identifies recurrent NFKBIE promoter mutations and diverse activating mutations in the MAPK pathway. Nat Genet 47(10): 1194–1199. https://doi.org/10.1038/ng.3382 251. Waddell N, Pajic M, Patch AM, Chang DK, Kassahn KS, Bailey P et al (2015) Whole genomes redefine the mutational landscape of pancreatic cancer. Nature 518(7540):495–501. https://doi.org/10. 1038/nature14169 252. Diamantopoulou Z, White G, Fadlullah MZH, Dreger M, Pickering K, Maltas J et al (2017) TIAM1 antagonizes TAZ/YAP both in the destruction complex in the cytoplasm and in the nucleus to inhibit invasion of intestinal epithelial cells. Cancer Cell 31(5): 621–34 e6. https://doi.org/10.1016/j.ccell.2017.03.007 253. Ambruso DR, Knall C, Abell AN, Panepinto J, Kurkchubasche A, Thurman G et al (2000) Human neutrophil immunodeficiency syndrome is associated with an inhibitory Rac2 mutation. Proc Natl Acad Sci U S A 97(9):4654–4659. https://doi.org/10.1073/ pnas.080074897 254. Williams DA, Tao W, Yang F, Kim C, Gu Y, Mansfield P et al (2000) Dominant negative mutation of the hematopoietic-specific Rho GTPase, Rac2, is associated with a human phagocyte immunodeficiency. Blood 96(5):1646–1654

310 255. Alkhairy OK, Rezaei N, Graham RR, Abolhassani H, Borte S, Hultenby K et al (2015) RAC2 loss-of-function mutation in 2 siblings with characteristics of common variable immunodeficiency. J Allergy Clin Immunol 135(5):1380–4 e1-5. https://doi. org/10.1016/j.jaci.2014.10.039 256. Dallery E, Galiegue-Zouitina S, Collyn-d'Hooghe M, Quief S, Denis C, Hildebrand MP et al (1995) TTF, a gene encoding a novel small G protein, fuses to the lymphoma-associated LAZ3 gene by t (3; 4) chromosomal translocation. Oncogene 10(11): 2171–2178 257. Pasqualucci L, Neumeister P, Goossens T, Nanjangud G, Chaganti RS, Kuppers R et al (2001) Hypermutation of multiple protooncogenes in B-cell diffuse large-cell lymphomas. Nature 412(6844):341–346. https://doi.org/10.1038/35085588 258. Crequer A, Troeger A, Patin E, Ma CS, Picard C, Pedergnana V et al (2012) Human RHOH deficiency causes T cell defects and susceptibility to EV-HPV infections. J Clin Invest 122(9): 3239–3247. https://doi.org/10.1172/JCI62949 259. Gu Y, Chae HD, Siefring JE, Jasti AC, Hildeman DA, Williams DA (2006) RhoH GTPase recruits and activates Zap 70 required for T cell receptor signaling and thymocyte development. Nat Immunol 7(11):1182–1190. https://doi.org/10.1038/ni1396 260. Gaidano G, Pasqualucci L, Capello D, Berra E, Deambrogi C, Rossi D et al (2003) Aberrant somatic hypermutation in multiple subtypes of AIDS-associated non-Hodgkin lymphoma. Blood 102(5):1833–1841. https://doi.org/10.1182/blood-2002-11-3606 261. Montesinos-Rongen M, Van Roost D, Schaller C, Wiestler OD, Deckert M (2004) Primary diffuse large B-cell lymphomas of the central nervous system are targeted by aberrant somatic hypermutation. Blood 103(5):1869–1875. https://doi.org/10.1182/ blood-2003-05-1465 262. Mayer L, Jasztal M, Pardo M, Aguera de Haro S, Collins J, Bariana TK et al (2018) Nbeal2 interacts with Dock7, Sec16a, and Vac14. Blood 131(9):1000–1011. https://doi.org/10.1182/blood-201708-800359 263. Cetinkaya A, Xiong JR, Vargel I, Kosemehmetoglu K, Canter HI, Gerdan OF et al (2016) Loss-of-function mutations in ELMO2 cause intraosseous vascular malformation by impeding RAC1 signaling. Am J Hum Genet 99(2):299–317. https://doi.org/10.1016/j. ajhg.2016.06.008 264. Mehawej C, Hoischen A, Farah RA, Marey I, David M, Stora S et al (2018) Homozygous mutation in ELMO2 may cause Ramon syndrome. Clin Genet 93(3):703–706. https://doi.org/10.1111/cge. 13166 265. Lin Y, Zheng Y (2015) Approaches of targeting Rho GTPases in cancer drug discovery. Expert Opin Drug Discovery 10(9): 991–1010. https://doi.org/10.1517/17460441.2015.1058775

Y. Lin and Y. Zheng 266. Clayton NS, Ridley AJ (2020) Targeting Rho GTPase signaling networks in cancer. Front Cell Dev Biol 8:222. https://doi.org/10. 3389/fcell.2020.00222 267. Maldonado MDM, Dharmawardhane S (2018) Targeting Rac and Cdc42 GTPases in cancer. Cancer Res 78(12):3101–3111. https:// doi.org/10.1158/0008-5472.CAN-18-0619 268. Bosco EE, Kumar S, Marchioni F, Biesiada J, Kordos M, Szczur K et al (2012) Rational design of small molecule inhibitors targeting the Rac GTPase-p67(phox) signaling axis in inflammation. Chem Biol 19(2):228–242. https://doi.org/10.1016/j.chembiol.2011. 12.017 269. Murray BW, Guo C, Piraino J, Westwick JK, Zhang C, Lamerdin J et al (2010) Small-molecule p21-activated kinase inhibitor PF-3758309 is a potent inhibitor of oncogenic signaling and tumor growth. Proc Natl Acad Sci U S A 107(20):9446–9451. https://doi.org/10.1073/pnas.0911863107 270. Rane C, Senapedis W, Baloglu E, Landesman Y, Crochiere M, Das-Gupta S et al (2017) A novel orally bioavailable compound KPT-9274 inhibits PAK4, and blocks triple negative breast cancer tumor growth. Sci Rep 7:42555. https://doi.org/10.1038/srep42555 271. Semenova G, Chernoff J (2017) Targeting PAK1. Biochem Soc Trans 45(1):79–88. https://doi.org/10.1042/BST20160134 272. Aiuti A, Biasco L, Scaramuzza S, Ferrua F, Cicalese MP, Baricordi C et al (2013) Lentiviral hematopoietic stem cell gene therapy in patients with Wiskott-Aldrich syndrome. Science 341(6148): 1233151. https://doi.org/10.1126/science.1233151 273. Sereni L, Castiello MC, Di Silvestre D, Della Valle P, Brombin C, Ferrua F et al (2019) Lentiviral gene therapy corrects platelet phenotype and function in patients with Wiskott-Aldrich syndrome. J Allergy Clin Immunol 144(3):825–838. https://doi.org/ 10.1016/j.jaci.2019.03.012 274. Ferrua F, Cicalese MP, Galimberti S, Giannelli S, Dionisio F, Barzaghi F et al (2019) Lentiviral haemopoietic stem/progenitor cell gene therapy for treatment of Wiskott-Aldrich syndrome: interim results of a non-randomised, open-label, phase 1/2 clinical study. Lancet Haematol 6(5):e239–ee53. https://doi.org/10.1016/ S2352-3026(19)30021-3 275. Rai R, Romito M, Rivers E, Turchiano G, Blattner G, Vetharoy W et al (2020) Targeted gene correction of human hematopoietic stem cells for the treatment of Wiskott-Aldrich syndrome. Nat Commun 11(1):4034. https://doi.org/10.1038/s41467-020-17626-2 276. Donnelly SK, Bravo-Cordero JJ, Hodgson L (2014) Rho GTPase isoforms in cell motility: Don't FRET, we have FRET. Cell Adhes Migr 8(6):526–534. https://doi.org/10.4161/cam.29712

Part IV Tools, Inhibitors, and Neighbors

Tools to Identify Noxes and their Regulators

19

Katrin Schröder

Abstract

The family of NADPH oxidases consists of at least 7 members: Nox1, Nox2, Nox3 Nox4, Nox5, Duox1 and Duox2. The sole function of all NADPH oxidases discovered so far is the formation of reactive oxygen species (ROS). For active ROS formation, NADPH oxidases rely on assembly of up to 5 subunits together with the individual Nox. This however, is a quite important feature, as overproduction of ROS can be harmful. On the other hand, ROS are essential second messengers that facilitate signal transduction and cellular homeostasis. Given the importance and simultaneously the danger of NADPH oxidases, they represent attractive targets for pharmacological treatments that modulate their activity. Despite this, before establishing any modulators it is important to understand what is known about unaltered NADPH oxidase complex formation and assembly. In this chapter, we will highlight the discovery of the NADPH oxidases and their regulators. Keywords

NADPH oxidases · Tools of identification · Cytosolic subunits · Non-protein regulators · Reactive oxygen species

Abbreviations CGD ER EST FPLC

Chronic granulomatous disease Endoplasmic reticulum Expressed sequence tags Fast protein liquid chromatography

K. Schröder (✉) Institute of Cardiovascular Physiology, Vascular Research Center, Faculty of Medicine, Goethe-University, Frankfurt, Germany e-mail: [email protected]

HA-tag His-tag IMAGE NADPH NCF MPSS mRNA Myc-tag PKC PMA RNA Ros RT-PCR siRNA SDS/ PAGE SH3 Sod Stat TLR TGF TNFα TPR

1

Hemagglutinin tag Polyhistidine-tag Integrated molecular analysis of genomes and their expression Nicotinamide adenine dinucleotide phosphate Neutrophil cytosolic factor Massively parallel signature sequencing Messenger ribonucleic acid Peptide tag derived from the c-Myc protein Protein kinase C Phorbol-12-myristat-13-acetat Ribonucleic acid Reactive oxygen species Reverse transcription polymerase chain reaction Short interfering RNA Sodium dodecyl sulphate–polyacrylamide gel electrophoresis Src homology 3 Superoxide dismutase Signal transducers and activators of transcription Toll like receptor Transforming growth factor Tumor necrosis factor alpha Tetratricopeptide repeat

Introduction

The family of NADPH oxidases consists of 7 members: Nox1, Nox2, Nox3 Nox4, Nox5, Duox1 and Duox2. The sole function of all NADPH oxidases discovered so far is the formation of reactive oxygen species (ROS). Overproduction of ROS can be harmful, but on the other hand, ROS are established second messengers that facilitate signal transduction of several hormones, growth factors and cytokines as well as cellular homeostasis. Accordingly, it is obvious that

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_19

313

314

K. Schröder

ROS formation by NADPH oxidases has to occur in a controlled manner and needs to be highly regulated. In fact, the activity of most NADPH oxidases is controlled by cytosolic subunits and by processes such as expression and degradation. Interestingly, the members of the NADPH oxidase family are expressed in a tissue and cell specific manner and most often cannot substitute for each other. Although it is not clear what controls cell specific expression patterns, it appears likely that cells exhibit different options, and subsequently individual strategies, to control the activity of NADPH oxidases. This chapter highlights the discovery of the NADPH oxidases and their regulators.

1.1

History and First Discoveries

The first NADPH oxidase discovered is Nox2 in the late 1970ies [1]. Before identification of the family members of NADPH oxidases, it was known that phagocytes express a membrane bound heme-containing protein that is involved in superoxide anion formation as a major part of first line host defense by phagocytes [2]. The midpoint potential of the protein in question is -245 mV, which led to the name cytochrome b-245. Phagocytes isolated from the blood of patients suffering from the inherited chronic granulomatous disease (CGD) fail to produce superoxide anions [3, 4]. Accordingly, CGD is a disease whose affected individuals suffer from a massively impaired host defense and weakened protection from infections with a wide variety of microorganisms [5]. A strategy of extended genomic cloning coupled with searches of transcribed regions was used to define disease loci in the genome of patients [6]. Eventually, it was found on the X-chromosome, the absence of which is the basis of the now so called X-linked CGD (X-CGD). Later the product of the X-CGD gene was isolated and identified as the β-chain of cytochrome b-245 [7, 8]. Although the two studies appeared simultaneously, they were entirely independent from each other. While Dinauer et al. [7] generated an antibody and performed Western Blots, Teahan et al. [8] based their study on sucrose gradient centrifugation, fast protein liquid chromatography (FPLC) and subsequent silver stained SDS/PAGEs. As time went by, the identified protein was named gp91phox and eventually Nox2 [9]. The small α-chain of cytochrome b-245 was identified on material purified by stepwise affinity chromatography on wheat germ agglutinin and heparin conjugated to Sepharose followed by SDS/PAGE. This approach revealed a 91 kDa (Nox2) and a 22 kDa protein band. The later turned out to be p22phox, which is expressed in many cells other than phagocytes [10–12]. Identification of the cytosolic subunits p47phox (NCF1) and p67phox (NCF2) was made possible by the discovery of the cell-free systems, allowing for Nox2 activation [13, 14].

Two parallel studies identified the cytosolic components with the aid of cell-free assays by mixing chromatographic fractions of cytosol with membranes and activating those with arachidonic acid and NADPH [15, 16]. Already in that study [15], p47phox was shown to be phosphorylated upon Phorbol-12-myristat-13-acetat (PMA) stimulation of the phagocytes. Protein kinase C, phosphorylates p47phox, which is essential for activation of the Nox2-complex [17]. Eventually, also a third protein was identified, p40phox, which later turned out to be a specific component of the phagocytic NADPH oxidase, where it rather supports activity than being an essential component of the complex [18, 19]. By that time, it was unclear, if p47phox is phosphorylated by NCF1 of if it actually is NCF1. It was also unclear whether NCF2 is p67phox. This issue was solved when Lomax et al. cloned p47phox [20] and Leto et al. p67phox [21] and found them to be the proposed cytosolic subunits of Nox2. Assembling of all subunits at the membrane occurs only when the membrane bound components Nox2 and p22phox both are present [22]. In fact, Nox2 and p22phox form a complex and the resulting heterodimer is known as flavo-cytochrome b558 [23]. Cloning and mutation analyses identified two tandem src homology 3 (SH3) domains in p47phox that bind to proline-rich sequence in p22phox, dependent on relief of autoinhibition [24, 25]. The C-terminal SH3 domain of p67phox binds to a proline-rich C-terminal sequence in p47phox and p67phox directly interacts with cytochrome b558 [26]. Besides p67phox, Rac GTPases (Rac2 in neutrophils [27] and Rac1 in cells not containing Rac2, such macophages [28, 29]) are absolutely required for activity of Nox2. Rac2 is proposed to position the activation domain of p67phox on cytochrome b558 [30] and directly binds Nox2 [31]. Both findings support the pivotal role of Rac2 in neutrophilic Nox2-complexformation and activity. With this, in the mid 1990ies a complex model of assembly of several Nox2 centered NADPH oxidase components was proposed, which has not been proved wrong until today.

1.2

Finding the Homologues

The availability of sequence data, together with improved possibilities to analyse them, spurred the discovery of homologues of known proteins. Based on sequence similarities, genome loci were predicted to code for homologues of Nox2. One human expressed sequence tag (EST) sequence was identified on chromosome 22 and cloning and functional studies revealed the newly discovered sequence to code for Nox1 [32]. The authors found Nox1 to play a role in growth, transformation and tumorigenicity of NIH3T3 cells (fibroblasts). Fibroblasts tumorigenecity was confirmed in vivo for inflammation induced sarcoma

19

Tools to Identify Noxes and their Regulators

formation [33]. This first description of Nox1 suggested it to rely on other cytosolic subunits than p47phox and p67phox, as ROS formation by overexpressed Nox1 remains unaffected by additional overexpression of the two in NIH3T3 cells. In 2003 Bánfi et al. found Nox1 is highly expressed in the colon, but the known cytosolic subunits of Nox2, p47phox and p67phox, were missing. They searched for homologues of p47phox and p67phox and identified two novel mRNAs predominantly expressed in colon epithelium, which they named NoxO1 (Nox organizer 1) and NoxA1 (Nox activator 1) [34]. NoxO1, which substitutes for p47phox, lacks the auto-inhibitory domain, including the first SH 3 domain and the protein kinase C-targeted phosphorylation sites. Accordingly, the two novel homologues do not need further activation and led to stimulus-independent high-level superoxide generation [35]. Nevertheless, both NoxO1-independent and -dependent Nox1 centered NADPH oxidase complexes depend on Rac1. A shorter splice variant of Nox1 was described, that was thought to be a H+-channel [36]. Although others reported the same finding for Nox1 [37] and Nox2 [38], it turned out that at least the proposed alternative splicing and therefore expression of a short form of Nox1 was an artifact of template switching during cDNA synthesis [39]. Nox1 was discovered a second time in 2000 by Kikuchi et al. who recognized that it was discovered already. However, using the sequence of Nox1 for further analysis Kikuchi et al. found a new homologue [37], which later turned out to be Nox3 [40]. Interestingly, the first report of Nox3 identified Nox3 mRNA by RT-PCR in fetal kidney and in the hepatoblastoma-derived cell line HepG2, but not in liver. A little later, Nox3 was discovered again and was found in fetal kidney, liver, lung and spleen but basically absent in adult tissue [41]. Although cell lines often do not reflect the situation in vivo, Nox3 may play a role in tumor necrosis factor alpha (TNFα)-signaling in hepatocytes and the liver’s insulin-response of mice [42] and indeed, Nox3 mRNA is found in human hepatocytes of human liver biopsies [43]. Despite this interesting fact, no function of Nox3 was identified in the liver. Pfaffenholz et al. found Nox3 to be important for otoconia formation in the (adult) vestibular system and therefore contributes to balance and gravity sensing [44]. At the same time, Banfi et al. found that Nox3 is active upon PMA treatment, if overexpressed in Hek293 cells alone or in combination with p47phox and p67phox or p47phox and NoxA1, while co expression of NoxO1 and NoxA1 resulted in constitutive ROS formation [45]. Ueno et al. reported a stimulus and subunit-independent activity of Nox3 [46]. Although based on overexpression this result is true for various cells and may indeed represent a unique and distinguishing feature of Nox3. However, according to Banfi et al., Nox3 in the inner ear is accompanied by p47phox and NoxA1. Rac appears to be an optional component of Nox3-

315

centered NADPH oxidase complexes. Overexpression studies with mutated Rac showed that Rac plays a positive role in Nox3 activation in the presence of p47(phox) and either p67 (phox) or NoxA1, whereas Rac fails to upregulate Nox3 activity when p47(phox) is replaced with NoxO1 [47]. Nox4 was discovered twice and in a similar way: In 2000 Geiszt et al. and in 2001 Shiose et al. conducted blast nucleotide searches within expressed sequence tags (ESTs) with nucleotide sequences corresponding to conserved C-terminal regions of gp91phox [48, 49]. With this approach, both groups identified different ESTs related to Nox4. Geiszt et al. identified a mouse cDNA EST clone (GenBank™ accession no. AI 226641) and two human EST clones with GenBank™ accession nos. AW237557 and AI241222. Shiose et al. identified a cDNA clone with the GenBank™ accession no. AI742260. All ESTs were uploaded and predicted by the Integrated Molecular Analysis of Genomes and their Expression (IMAGE) Consortium [50]. This is an example of how multiple discoveries of the same cDNA can cause confusion. No more than 1000–2000 different prevalent and intermediate frequency mRNAs comprise 50–65% of the total mRNA mass in a cell, provoking redundant identification of mRNAs from cDNA libraries. Accordingly, the IMAGE consortium recognized and reduced the danger of multiple discoveries of the same gene already by a subtractive hybridization approach, designed specifically to eliminate or significantly reduce the representation of large pools of arrayed and sequenced clones from normalized libraries yet to be surveyed [50]. Eventually both working groups sorted the identified sequence into the group of NADPH oxidases. Northern blots and in situ hybridizations revealed the chromosomal localization of the identified sequence as well as tissue distribution of its mRNA. Shiose et al. raised an antibody and performed immunehistochemical stainings and Western blots. A little later, Nox4 was cloned a third time, again based on sequence similarities with Nox2 [41]. Different from Nox1, Nox2 and Nox3, Nox4 is not located at the plasma membrane. Nox4 contains a signal peptide at its very N-terminal part, which determines endoplasmic reticulum (ER)-localization, where it produced H2O2 instead of superoxide anions [51]. Discovery and first cloning of Nox5 was based on a Blast search using Nox3 as a query. Two unfinished genomic clones were found and cloned. Northern blot experiments revealed a 2.2 kb band in all fetal tissues tested, while present at low amounts in adult spleen and testis, along with larger transcripts at 2.6 and 6 kb [41]. Unfortunately, this interesting finding was not further analyzed by the authors. Like the other NADPH oxidase homologues, Nox5 was simultaneously discovered by a second group [52]. Banfi et al. performed BLAST nucleotide searches in GenBank™ with the C terminus of human Nox2 and performed Northern blots and in situ hybridization. They found Nox5 to be expressed in

316

spleen and testis as well and in lymph nodes. Like Cheng et al. they discovered Nox5 to be present at two different lengths, but Banfi et al. went on and investigated that phenomenon. The two transcript variants found in spleen and testis were found to result from different transcription initiation sites and named Nox5α and β by the authors. Further sequencing revealed two additional splice variants referred to as NOX5γ and NOX5δ. The unique feature of Nox5, when compared to other Noxs identified by both groups is the presence of EF hands. Banfi et al. performed experiments to show the functional importance of those by providing evidence that Nox5 can be activated by elevated intracellular Ca2+-level [52, 53]. Based on two different assays measuring ROS, namely H2O2 and superoxide, (luminol/horseradish peroxidase and ferricytochrome c/superoxide dismutase (SOD)) the authors also identified superoxide anions as the product of Nox5 and with the aid of patch clamp technique the authors identified a Ca2+-activated H+-current in Nox5 overexpressing cells [52]. While writing this section I cannot omit to mention that this [52] is a study showing all the beauty of science in a careful, extensive and appropriate description of a newly discovered protein. Besides the “around-65 kDa-members” of the NADPH oxidase family two large molecular weight homologs of Nox2 were identified in 1999 and 2000; Duox1 and Duox2 with both of which mainly expressed in thyroid and a variety of other tissues [54, 55]. The size of 175–180 kDa homologs results from an N-terminal peroxidase domain. Right from their discovery, the Duoxes were clearly associated with their physiological role. Especially Duox2 is involved in thyroid hormonogenesis. Biallelic nonsense mutations of the enzyme result in severe congenital hypothyroidism [56]. Additionally, subunits of Duox1 and 2 have been identified, from an atlas of human gene expression from massively parallel signature sequencing (MPSS), based on tissue specific mRNA expression profiling [57] by Grasberger et al. in 2006 [58]. Techniques based on different tags of cloned and overexpressed proteins have been used to verify predicted intracellular localizations of known proteins. For example, localization of human influenza hemagglutinin (HA)-tagged Duox2 was analyzed with and without co-overexpression of DuoxA2 for its intracellular localization, identifying DuoxA2 as a maturation factor of Duox2, that keeps its target protein in the ER [58]. With all the mentioned discoveries based on classical techniques such as sequence comparisons, cloning and mutagenesis, Northern and Western blots, the family of NADPH oxidases was complete in the early 2000s.

K. Schröder

1.3

How to Identify the Interactors and their Function?

Constitutive high level ROS formation by NADPH oxidases can harm a cell or even an organism. Accordingly, disassembly of the multi component enzyme-complex is as important as assembly in order to ensure controlled and adequate ROS formation at all times. This section will give attention to basic aspects of protein-protein interactions involved in assembly and disassembly of multi protein complexes. Protein-protein interactions as such, can be stable or transient, and strong or weak. Transient interactions are temporary in nature, strong or weak, fast or slow and often require conditions, such as phosphorylation, other posttranslational modifications, conformational changes or specific localization within a cell. Within the family members of NADPH oxidases, the nature of interaction of the Nox and the p22phox subunit varies. While the Nox2-p22phox dimer can be dissociated on PAGE only by SDS, very mild detergents are needed, if a Nox4 pulldown assay should pull down p22phox [59]. Despite this variation in strength, a stable interaction of the Nox1–4 subunits with p22phox is required to stabilize the Nox proteins. Once discovered, the molecular basis of this phenomenon may be useful to develop pharmacological interventions, to control expression of the individual Nox1–4 subunits. Apart from the membrane bound subunits of the NADPH oxidase complex, also the cytosolic subunits discovered so far, assemble in a transient way, indicating the existence of mechanisms of disruption of the complex and termination of ROS formation. Uncovering additional interactors of NAPPH oxidase subunits is required for a better understanding of regulation of NADPH oxidase complex assembly and disassembly. Further, posttranslational modifications of the complex subunits, such as oxidation could provide the basis for disruption of the NADPH oxidase complex and need to be analyzed. It is in the nature of protein interaction research that strong and permanent interactions of proteins can be found relatively easy (e.g. by co-immunoprecipitations and pull-down assays), while transient and weak interactions require more sophisticated approaches (e.g. cross-linking or label transfer analyses) (Fig. 19.1). It is unnecessary to mention that verifying or excluding interactions proposed by an educated guess are easier than trying to find unexpected interaction partners in an untargeted way. The following section will highlight some targeted approaches, followed by a smaller section for relatively new untargeted approaches together with an outlook.

19

Tools to Identify Noxes and their Regulators

317

Pulldown experiments

Sequence Analysis

for weak and transient interaction

for strong interaction

Y

Cloning and expression

bead coupled antibody

nearby protein

p67phox

Nox1

biotin ligase

p67phox

biotin tag

biotin

Fluorescence Microscopy Western Blot and Mass spectrometry based tools

Nox1

Intensity

Y Y

p22 phox membrane

m/z Artificial intelligence

Duox1 p22phox

Nox1

Nox2

Nox3

Nox4

Duox2

Nox5

Fig. 19.1 Summary scheme of methods and tools to identify NADPH oxidases and their subunits. Sequence analyses were used to identify homologues of already discovered proteins. Association of the subunits was often verified by fluorescence microscopy or pull down experiments. Transient interactions can be analyzed by techniques such as BioID, where a biotin-ligase fused to a protein of interest

attaches a biotin to nearby proteins. Proteins pulled down or biotinylated can be identified e.g. by Western blots or mass spectrometry. Eventually, artificial intelligence can be used to model intramembrane proteins such as the Nox and Duox subunits of the NADPH oxidase complexes. In a next step, artificial intelligence may be able to predict potential interaction partners, which then can be verified experimentally

1.4

1.4.1 Rac The small GTPase Rac was early characterized as an essential subunit of the Nox2 NADPH oxidase complex. Later, it was discovered to be part of the active Nox1 and Nox3 NADPH oxidase complexes as well. Those findings made Rac an apparent candidate applicable for targeted approaches to identify or verify interactions of newly discovered subunits with active Nox1–3 NADPH oxidase complexes. Interaction of NoxA1 and p67phox with Rac was discovered by classic mutagenesis study and co-precipitation: A tag such as HA is fused to the protein of interest and the target-protein is modified by a His- or Myc-tag. Both tagged proteins are overexpressed in one cell. Interaction of the two can be

Targeted Approaches

On the heels of identification of cytosolic subunits of the NADPH oxidases, researchers were looking for kinases and other posttranslational modifiers that may activate or terminate the assembly of the NADPH oxidase subunits. For example, the most prominent kinase to phosphorylate p47phox, protein kinase C (PKC), was identified based on a sequence domain in p47phox suitable to bind PKC and verified with the aid of PKC inhibitors [60]. The corresponding phosphatase however, remains unspecified, with one potential candidate being protein phosphatase 2A (PP2A) [61].

318

analyzed with or without treatment of the cell with a stimulus. In order to do so, proteins are cross-linked and the protein of interest is isolated via its tag. Eventually, interaction of the two proteins is confirmed or excluded based on Western Blot analyses [30, 35].

1.4.2 p22phox The major maturation factor that stabilizes NADPH oxidases Nox1–4 and contributes to their localization is p22phox [62]. The physical association of p22phox with the integral membrane Nox protein varies from strong (Nox2) to weak (Nox4). Nevertheless, interaction of the Nox proteins and p22phox was shown several times [63]. Downregulation of p22phox by a CRISPR/Cas approach results in a loss of Nox1 and Nox4 [64] and point mutations in the proline-rich region, truncated forms of p22phox and p22phox shRNA inhibit ROS formation by Nox1, Nox2, Nox3 and Nox4 [65]. Truncated forms of p22phox missing the proline-rich region inhibited ROS formation by Nox1, Nox2, and Nox3 but not by Nox4 or Nox5, while shRNA mediated knockdown of p22phox also inhibited Nox4 mediated ROS formation. Nox5 remained unaffected by any manipulations of p22phox. Similarly, p22phoxP156Q, a mutation that disrupts SH3 binding by the proline rich region, potently inhibited reactive oxygen production from Nox1 and Nox2 but not from Nox4 and Nox5. Accordingly, p22phoxP156Q prevents membrane anchoring of the organizer subunits of the subunitdependent NADPH oxidase Nox1–3 [65]. In Hek293 cells overexpressing Nox3 together with its cytosolic subunits, application of a p22phox siRNA reduced Nox3 mediated ROS formation [66]. As a potential mechanism, the authors suggest that glycosylation and maturation of Nox3 requires p22phox. The findings mentioned are based on overexpression of fusion proteins, haem-spectum analysis of membrane sections, ROS measurements and immunofluorescence. These studies stress the importance of p22phox for stability and function of Nox1, Nox2, Nox3, and Nox4, and emphasize the key role of the proline rich region for regulating Nox proteins whose activity is dependent on p47phox or NoxO1. New analyses using untargeted approaches will unmask potential regulators of p22phox, including interacting proteins and oxidation. 1.4.3 Nox4 Nox4s activity is solely based on its expression level and correlates with its appropriate mRNA level [67]. An ATP binding site identified in Nox4, has been shown to directly bind ATP, which negatively regulates Nox4 activity [68]. However, besides p22phox mediated stabilization of the Nox4 protein [64, 69], no regulatory subunit of Nox4 was identified until today. Recently a binary luciferase

K. Schröder

reporter assay (NanoBiT®) was suggested as a targeted approach to verify the association of Nox4 and p22phox [70]. This technique is based on the interaction of a Large BiT (LgBiT; 18 kDa) subunit and a small complimentary peptide Small BiT (SmBiT; 11 amino acid peptide). Both, LgBiT and SmBiT are fused to individual proteins of interest and expressed in cells. When the two fusion-proteins come into close proximity LgBiT and SmBiT form an active enzyme, that generate a luminescent signal in the presence of a substrate [71]. This reporter allowed quantitative determination of accurate, reduced, or failed complex assembly of Nox4 and p22phox, which was confirmed by analyzing Nox4 mediated H2O2 release, protein expression, and dimer localization [72]. The Griendling laboratory discovered an interaction of Nox4/p22phox with the polymerase delta-interacting protein 2 (Poldip2). In fact, they suggest that Poldip2 associates with p22phox to activate Nox4 [73]. Poldip2 therefore, is not a direct interactor of Nox4, but may enhance Nox4s association with p22phox, which in turn promotes Nox4s stability and subsequent ROS formation by the enzyme. As pointed out above, Nox4 protein expression and activity are directly correlated to its mRNA expression, at least in an overexpression system [67]. In native cells, Nox4 mRNA expression is spurred by treatment with transforming growth factor beta (TGFβ) [74]. Interestingly, there is evidence for an acute (meaning, within a shorter time than would be needed for the synthesis of more protein) TGFβ-induced activation of H2O2 formation by Nox4 (unpublished observation from several laboratories). This however, interferes with all known characteristics of Nox4 and cannot be explained so far. Accordingly, researchers put a lot of effort into finding interaction partners of Nox4 and identified a number of more or less likely candidates including phosphorylated Signal Transducers and Activators of Transcription (Stat3) [75], toll like receptor 4 (TLR4) [76], mitochondrial cytochrome b5 reductase 3 (CYB5R3) [77] and even nuclear proteins such as prelaminin A [78]. However, all those just interact with Nox4 to control its degradation, without any contribution to an acute activation of the enzyme.

1.4.4 Nox5 in Ca2+ Dependent ROS Formation Although a conformational change, resulting in association of the N-terminal EF-hands and the C-terminal catalytic domain of Nox5 has been shown to be responsible for the Ca2+-induced superoxide formation, a later study further analyzed the mode of activation of Nox5 and identified calmodulin as a regulator of Nox5. Within the sequence of the C-terminal domain in Nox5, Tirone and Cox identified a calmodulin-binding domain (CaM-BD). Using GST pulldown and purified protein in a targeted approach, they found this domain to interact selectively with calmodulin

19

Tools to Identify Noxes and their Regulators

(CaM) and not with the Nox5-EF hands. CaM binding however, does not lead to additional activation, but increases the Ca2+-sensitivity of Nox5 [79].

1.5

Untargeted Approaches and Outlook

1.5.1

Untargeted Analysis of Protein-Protein Interaction Transient or weak close proximity of two proteins can be monitored e.g. by techniques such as BioID. In order to perform a BioID assay, a protein of interest is fused to a biotin ligase. The fusion protein will then be overexpressed in a cell of interest and upon adding biotin, all proteins nearby the BioID-fusion-protein will be marked by a biotin. After a biotin pull down, e.g. with streptavidin, the biotin-marked proteins will be identified, e.g. by a mass spectrometry [80]. Alternatively, simple fluorescent microscopy can visualize the area of action of the fusion protein. The interested reader is referred to [81] to learn about more application options. Even a co-staining for ROS modification and ROS themselves, in a manner similar to proximity ligation, appears to be possible [82]. With the aid of techniques such as BioID, it is possible to identify transient and weak protein-protein interactions in an untargeted approach (Fig. 19.1). A possible disadvantage could be that some proteins of interest upon fusion with the biotin-ligase may change essential structures and subsequently lose their usual intracellular localization or binding features. 1.5.2 Non-protein Regulators Non-protein regulators of NADPH oxidases are of high interest, especially in terms of pharmacologic interventions. According to the assumption that ROS are harmful, the search for Nox inhibitors was predominant, until it became clear that the members of the NADPH oxidase family exert quite diverse functions and that general inhibition of NADPH oxidases should not be the only goal. Since then, the search for specific inhibitors for single members of the family is going on, despite the fact that the benefit of inhibiting NADPH oxidase is still a matter of debate. Nevertheless, several natural compound inhibitors of NADPH oxidases [83] and a few small molecule inhibitors [84] have been identified [See chapter „Isoform-Selective NOX Inhibitors: Advances and Future Perspectives“, by C.M. Dustin, E. Cifuentes-Pagano, and P.J. Pagano in this book]. On the other hand, it might be helpful, to boost the activity of specific NADPH oxidases, based on the assumption that they have beneficial effects. Nox1 activation could promote gut wound healing [85]. Nox4-activation could boost angiogenesis and reduce heart hypertrophy, as shown in a model of pressure overload induced by thoracic aortic constriction

319

[86]. From a statistical point of view, analyzes of huge small molecule libraries should lead to the discovery of NADPH oxidase activators as often as to the discovery of inhibitors of NADPH oxidase. Accordingly, there must be some knowledge about small molecule activators of NADPH oxidases, which is just not published. Although pure speculation, it is likely that the “negative image” of NADPH oxidases not only spurs the attention for inhibitors but also induces publication-bias for harmful effects of small molecule activators or inducers of NADPH oxidases. One example is cisplatin, an ototoxic cytostatic that acts as a potent inducer of ROS in the inner ear and in hair cells [87]. Cisplatin was suggested to increase ROS formation by Nox3/ NoxA1/NoxO1 and Nox5, if overexpressed in Hek293 cells [45]. Importantly, superoxide formation by xanthine/xanthine oxidase was increased as well, indicating that the effect of cisplatin is not restricted to NADPH oxidases but rather reflects an inhibition downstream of the ROS producing enzyme and in fact, cisplatin turned out to inhibit superoxide dismutase [88, 89].

1.5.3 Artificial Intelligence Artificial intelligence (AI) is a promising future technology. It is fundamental to and expected to become essential for all aspects of research [90, 91]. One specific aspect of AI is machine learning; the ability of a computer to generate expert knowledge and experience based on existing and available data sets and potentially computer generated algorithms [92]. The computer then can sort things into categories based on autonomous decisions including the choice of the appropriate underlying pattern. This skill turns the passive tool computer into an active tool, which is able to make decisions within the process of data analysis. Immediate integration of the results from an analysis with available literature and results from past (human-verified) analysis reinforces the quality of each new data analysis. So called dry laboratory experiments will be the motor of medical research that e.g. will boost personalized medicine by enabling researchers to find the reason for non-responders to therapeutics and therapy resistances [93]. AI holds the potential to massively advance protein and nucleic acid research as well as patient big data usage and analysis. One example for how AI and especially machine learning could be used in biomedical and pharmacological research is the platform AlphaFold [94]. With the aid of this platform, it is possible to predict 3D structures of unknown proteins, based on algorithms and datasets from other proteins, the 3D structures of which are known already (Fig. 19.1). If the next step was to extend the functions of this or an alternative platform by machine learning, it will be possible to model 3D interactions of two or more proteins and eventually to artificially analyze how small molecules or post-translational modifications may alter 3D structures and subsequent

320

assembly of multi protein complexes [95]. Although most of those findings made by AI will have to be verified conventionally, AI will change the way of how molecular discoveries will be made [96]. I am convinced that AI will expand the horizon of possibilities of how we identify components and interactors of NADPH oxidases.

References 1. Segal AW, Jones OT (1978) Novel cytochrome b system in phagocytic vacuoles of human granulocytes. Nature 276:515–517. https:// doi.org/10.1038/276515a0 2. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. J Clin Invest 52:741–744. https://doi.org/10. 1172/JCI107236 3. Quie PG, White JG, Holmes B et al (1967) In vitro bactericidal capacity of human polymorphonuclear leukocytes: diminished activity in chronic granulomatous disease of childhood. J Clin Invest 46:668–679. https://doi.org/10.1172/JCI105568 4. Nathan DG, Baehner RL, Weaver DK (1969) Failure of nitro blue tetrazolium reduction in the phagocytic vacuoles of leukocytes in chronic granulomatous disease. J Clin Invest 48:1895–1904. https:// doi.org/10.1172/JCI106156 5. Segal AW, Cross AR, Garcia RC et al (1983) Absence of cytochrome b-245 in chronic granulomatous disease. A multicenter European evaluation of its incidence and relevance. N Engl J Med 308:245–251. https://doi.org/10.1056/NEJM198302033080503 6. Royer-Pokora B, Kunkel LM, Monaco AP et al (1986) Cloning the gene for an inherited human disorder--chronic granulomatous disease--on the basis of its chromosomal location. Nature 322:32–38. https://doi.org/10.1038/322032a0 7. Dinauer MC, Orkin SH, Brown R et al (1987) The glycoprotein encoded by the X-linked chronic granulomatous disease locus is a component of the neutrophil cytochrome b complex. Nature 327: 717–720. https://doi.org/10.1038/327717a0 8. Teahan C, Rowe P, Parker P et al (1987) The X-linked chronic granulomatous disease gene codes for the beta-chain of cytochrome b-245. Nature 327:720–721. https://doi.org/10.1038/327720a0 9. Lambeth J, Cheng G, Arnold RS et al (2000) Novel homologs of gp91phox. Trends Biochem Sci 25:459–461. https://doi.org/10. 1016/S0968-0004(00)01658-3 10. Dinauer MC, Pierce EA, Bruns GA et al (1990) Human neutrophil cytochrome b light chain (p22-phox). Gene structure, chromosomal location, and mutations in cytochrome-negative autosomal recessive chronic granulomatous disease. J Clin Invest 86:1729–1737. https:// doi.org/10.1172/JCI114898 11. Parkos CA, Dinauer MC, Walker LE et al (1988) Primary structure and unique expression of the 22-kilodalton light chain of human neutrophil cytochrome b. Proc Natl Acad Sci 85:3319–3323. https:// doi.org/10.1073/pnas.85.10.3319 12. Parkos CA, Allen RA, Cochrane CG et al (1987) Purified cytochrome b from human granulocyte plasma membrane is comprised of two polypeptides with relative molecular weights of 91,000 and 22,000. J Clin Invest 80:732–742. https://doi.org/10.1172/ JCI113128 13. Bromberg Y, Pick E (1984) Unsaturated fatty acids stimulate NADPH-dependent superoxide production by cell-free system derived from macrophages. Cell Immunol 88:213–221. https://doi. org/10.1016/0008-8749(84)90066-2 14. Bromberg Y, Pick E (1985) Activation of NADPH-dependent superoxide production in a cell-free system by sodium dodecyl sulfate. J Biol Chem 260:13539–13545. https://doi.org/10.1016/S0021-9258 (17)38756-2

K. Schröder 15. Nunoi H, Rotrosen D, Gallin JI et al (1988) Two forms of autosomal chronic granulomatous disease lack distinct neutrophil cytosol factors. Science 242:1298–1301. https://doi.org/10.1126/science. 2848319 16. Volpp BD, Nauseef WM, Clark RA (1988) Two cytosolic neutrophil oxidase components absent in autosomal chronic granulomatous disease. Science 242:1295–1297. https://doi.org/10.1126/ science.2848318 17. Segal AW, Heyworth PG, Cockcroft S et al (1985) Stimulated neutrophils from patients with autosomal recessive chronic granulomatous disease fail to phosphorylate a Mr-44,000 protein. Nature 316:547–549. https://doi.org/10.1038/316547a0 18. Someya A, Nagaoka I, Yamashita T (1993) Purification of the 260 kDa cytosolic complex involved in the superoxide production of Guinea pig neutrophils. FEBS Lett 330:215–218. https://doi.org/ 10.1016/0014-5793(93)80276-Z 19. Wientjes FB, Hsuan JJ, Totty NF et al (1993) p40phox, a third cytosolic component of the activation complex of the NADPH oxidase to contain src homology 3 domains. Biochem J 296(Pt 3): 557–561. https://doi.org/10.1042/bj2960557 20. Lomax KJ, Leto TL, Nunoi H et al (1989) Recombinant 47-kilodalton cytosol factor restores NADPH oxidase in chronic granulomatous disease. Science 245:409–412. https://doi.org/10. 1126/science.2547247 21. Leto TL, Lomax KJ, Volpp BD et al (1990) Cloning of a 67-kD neutrophil oxidase factor with similarity to a noncatalytic region of p60c-src. Science 248:727–730. https://doi.org/10.1126/science. 1692159 22. Heyworth PG, Curnutte JT, Nauseef WM et al (1991) Neutrophil nicotinamide adenine dinucleotide phosphate oxidase assembly. Translocation of p47-phox and p67-phox requires interaction between p47-phox and cytochrome b558. J Clin Invest 87:352– 356. https://doi.org/10.1172/JCI114993 23. Parkos CA, Allen RA, Cochrane CG et al (1988) The quaternary structure of the plasma membrane b-type cytochrome of human granulocytes. Biochimica et Biophysica Acta (BBA) - Bioenergetics 932:71–83. https://doi.org/10.1016/0005-2728(88)90140-5 24. Leto TL, Adams AG, de Mendez I (1994) Assembly of the phagocyte NADPH oxidase: binding of Src homology 3 domains to proline-rich targets. Proc Natl Acad Sci 91:10650–10654. https:// doi.org/10.1073/pnas.91.22.10650 25. Sareila O, Jaakkola N, Olofsson P et al (2013) Identification of a region in p47phox/NCF1 crucial for phagocytic NADPH oxidase (NOX2) activation. J Leukoc Biol 93:427–435. https://doi.org/10. 1189/jlb.1211588 26. Dang PM, Cross AR, Babior BM (2001) Assembly of the neutrophil respiratory burst oxidase: a direct interaction between p67PHOX and cytochrome b558. Proc Natl Acad Sci 98:3001–3005. https:// doi.org/10.1073/pnas.061029698 27. Knaus UG, Heyworth PG, Evans T et al (1991) Regulation of phagocyte oxygen radical production by the GTP-binding protein Rac 2. Science 254:1512–1515. https://doi.org/10.1126/science. 1660188 28. Zhao X, Carnevale KA, Cathcart MK (2003) Human monocytes use Rac1, not Rac2, in the NADPH oxidase complex. J Biol Chem 278: 40788–40792. https://doi.org/10.1074/jbc.M302208200 29. Abo A, Pick E, Hall A et al (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1. Nature 353:668–670. https://doi.org/10.1038/353668a0 30. Koga H, Terasawa H, Nunoi H et al (1999) Tetratricopeptide repeat (TPR) motifs of p67 participate in interaction with the small GTPase Rac and activation of the phagocyte NADPH oxidase. J Biol Chem 274:25051–25060. https://doi.org/10.1074/jbc.274.35.25051 31. Kao Y-Y, Gianni D, Bohl B et al (2008) Identification of a conserved Rac-binding site on NADPH oxidases supports a direct GTPase regulatory mechanism. J Biol Chem 283:12736–12746. https://doi.org/10.1074/jbc.M801010200

19

Tools to Identify Noxes and their Regulators

32. Suh YA, Arnold RS, Lassegue B et al (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401:79–82. https:// doi.org/10.1038/43459 33. Helfinger V, Freiherr von Gall F, Henke N et al (2021) Genetic deletion of Nox4 enhances cancerogen-induced formation of solid tumors. Proc Natl Acad Sci 118:e2020152118. https://doi.org/10. 1073/pnas.2020152118 34. Bánfi B, Clark RA, Steger K et al (2003) Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 278:3510–3513. https://doi.org/10.1074/jbc.C200613200 35. Takeya R, Ueno N, Kami K et al (2003) Novel human homologues of p47 and p67 participate in activation of superoxide-producing NADPH oxidases. J Biol Chem 278:25234–25246. https://doi.org/ 10.1074/jbc.M212856200 36. Bánfi B, Maturana A, Jaconi S et al (2000) A mammalian H+ channel generated through alternative splicing of the NADPH oxidase homolog NOH-1. Science 287:138–142. https://doi.org/10. 1126/science.287.5450.138 37. Kikuchi H, Hikage M, Miyashita H et al (2000) NADPH oxidase subunit, gp91phox homologue, preferentially expressed in human colon epithelial cells. Gene 254:237–243. https://doi.org/10.1016/ S0378-1119(00)00258-4 38. Henderson LM, Meech RW (1999) Evidence that the product of the human X-linked CGD gene, gp91-phox, is a voltage-gated H(+) pathway. J Gen Physiol 114:771–786. https://doi.org/10.1085/jgp. 114.6.771 39. Geiszt M, Lekstrom K, Leto TL (2004) Analysis of mRNA transcripts from the NAD(P)H oxidase 1 (Nox1) gene. Evidence against production of the NADPH oxidase homolog-1 short (NOH-1S) transcript variant. J Biol Chem 279:51661–51668. https://doi.org/10.1074/jbc.M409325200 40. Homo sapiens NOX3 (2022). https://www.biocyc.org/gene?orgid= HUMAN&id=HS01151. Accessed 03 Jan 2022 41. Cheng G, Cao Z, Xu X et al (2001) Homologs of gp91 phox: cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269:131– 140. https://doi.org/10.1016/s0378-1119(01)00449-8 42. Gao D, Nong S, Huang X et al (2010) The effects of palmitate on hepatic insulin resistance are mediated by NADPH oxidase 3-derived reactive oxygen species through JNK and p38MAPK pathways. J Biol Chem 285:29965–29973. https://doi.org/10.1074/ jbc.M110.128694 43. Tissue expression of NOX3 - Staining in liver - The Human Protein Atlas (2022). https://www.proteinatlas.org/ENSG00000074771NOX3/tissue/liver. Accessed 04 Jan 2022 44. Paffenholz R, Bergstrom RA, Pasutto F et al (2004) Vestibular defects in head-tilt mice result from mutations in Nox3, encoding an NADPH oxidase. Genes Dev 18:486–491. https://doi.org/10. 1101/gad.1172504 45. Bánfi B, Malgrange B, Knisz J et al (2004) NOX3, a superoxidegenerating NADPH oxidase of the inner ear. J Biol Chem 279: 46065–46072. https://doi.org/10.1074/jbc.M403046200 46. Ueno N, Takeya R, Miyano K et al (2005) The NADPH oxidase Nox3 constitutively produces superoxide in a p22phox-dependent manner: its regulation by oxidase organizers and activators. J Biol Chem 280:23328–23339. https://doi.org/10.1074/jbc.M414548200 47. Miyano K, Sumimoto H (2007) Role of the small GTPase Rac in p22phox-dependent NADPH oxidases. Biochimie 89:1133–1144. https://doi.org/10.1016/j.biochi.2007.05.003 48. Geiszt M, Kopp JB, Várnai P et al (2000) Identification of renox, an NAD(P)H oxidase in kidney. Proc Natl Acad Sci 97:8010–8014. https://doi.org/10.1073/pnas.130135897 49. Shiose A, Kuroda J, Tsuruya K et al (2001) A novel superoxideproducing NAD(P)H oxidase in kidney. J Biol Chem 276:1417– 1423. https://doi.org/10.1074/jbc.M007597200

321 50. Bonaldo MF, Lennon G, Soares MB (1996) Normalization and subtraction: two approaches to facilitate gene discovery. Genome Res 6:791–806. https://doi.org/10.1101/gr.6.9.791 51. Helmcke I, Heumüller S, Tikkanen R et al (2009) Identification of structural elements in Nox1 and Nox4 controlling localization and activity. Antioxid Redox Signal 11:1279–1287. https://doi.org/10. 1089/ars.2008.2383 52. Bánfi B, Molnár G, Maturana A et al (2001) A ca(2+)-activated NADPH oxidase in testis, spleen, and lymph nodes. J Biol Chem 276:37594–37601. https://doi.org/10.1074/jbc.M103034200 53. Bánfi B, Tirone F, Durussel I et al (2004) Mechanism of Ca2+ activation of the NADPH oxidase 5 (NOX5). J Biol Chem 279: 18583–18591. https://doi.org/10.1074/jbc.M310268200 54. de Deken X, Wang D, Many MC et al (2000) Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275:23227–23233. https://doi.org/10.1074/ jbc.M000916200 55. Dupuy C, Ohayon R, Valent A et al (1999) Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. Cloning of the porcine and human cdnas. J Biol Chem 274:37265–37269. https:// doi.org/10.1074/jbc.274.52.37265 56. Moreno JC, Bikker H, Kempers MJE et al (2002) Inactivating mutations in the gene for thyroid oxidase 2 (THOX2) and congenital hypothyroidism. N Engl J Med 347:95–102. https://doi.org/10. 1056/NEJMoa012752 57. Jongeneel CV, Delorenzi M, Iseli C et al (2005) An atlas of human gene expression from Massively Parallel Signature Sequencing (MPSS). Genome Res 15:1007–1014. https://doi.org/10.1101/gr. 4041005 58. Grasberger H, Refetoff S (2006) Identification of the maturation factor for dual oxidase. Evolution of an eukaryotic operon equivalent. J Biol Chem 281:18269–18272. https://doi.org/10.1074/jbc. C600095200 59. Prior K-K, Wittig I, Leisegang MS et al (2016) The endoplasmic reticulum chaperone calnexin is a NADPH oxidase NOX4 interacting protein. J Biol Chem 291:7045–7059. https://doi.org/ 10.1074/jbc.M115.710772 60. Nauseef WM, Volpp BD, McCormick S et al (1991) Assembly of the neutrophil respiratory burst oxidase. Protein kinase C promotes cytoskeletal and membrane association of cytosolic oxidase components. J Biol Chem 266:5911–5917 61. Eteläinen T, Kulmala V, Svarcbahs R et al (2021) Prolyl oligopeptidase inhibition reduces oxidative stress via reducing NADPH oxidase activity by activating protein phosphatase 2A. Free Radic Biol Med 169:14–23. https://doi.org/10.1016/j. freeradbiomed.2021.04.001 62. Zhu Y, Marchal CC, Casbon A-J et al (2006) Deletion mutagenesis of p22phox subunit of flavocytochrome b558: identification of regions critical for gp91phox maturation and NADPH oxidase activity. J Biol Chem 281:30336–30346. https://doi.org/10.1074/jbc. M607191200 63. Ambasta RK, Kumar P, Griendling KK et al (2004) Direct interaction of the novel Nox proteins with p22phox is required for the formation of a functionally active NADPH oxidase. J Biol Chem 279:45935–45941. https://doi.org/10.1074/jbc.M406486200 64. Prior K-K, Leisegang MS, Josipovic I et al (2016) CRISPR/Cas9mediated knockout of p22phox leads to loss of Nox1 and Nox4, but not Nox5 activity. Redox Biol 9:287–295. https://doi.org/10.1016/j. redox.2016.08.013 65. Kawahara T, Ritsick D, Cheng G et al (2005) Point mutations in the proline-rich region of p22phox are dominant inhibitors of Nox1- and Nox2-dependent reactive oxygen generation. J Biol Chem 280: 31859–31869. https://doi.org/10.1074/jbc.M501882200 66. Nakano Y, Banfi B, Jesaitis AJ et al (2007) Critical roles for p22phox in the structural maturation and subcellular targeting of Nox3. Biochem J 403:97–108. https://doi.org/10.1042/BJ20060819

322 67. Serrander L, Cartier L, Bedard K et al (2007) NOX4 activity is determined by mRNA levels and reveals a unique pattern of ROS generation. Biochem J 406:105–114. https://doi.org/10.1042/ BJ20061903 68. Shanmugasundaram K, Nayak BK, Friedrichs WE et al (2017) NOX4 functions as a mitochondrial energetic sensor coupling cancer metabolic reprogramming to drug resistance. Nat Commun 8: 997. https://doi.org/10.1038/s41467-017-01106-1 69. Zana M, Péterfi Z, Kovács HA et al (2018) Interaction between p22phox and Nox4 in the endoplasmic reticulum suggests a unique mechanism of NADPH oxidase complex formation. Free Radic Biol Med 116:41–49. https://doi.org/10.1016/j.freeradbiomed.2017. 12.031 70. O’Neill S, Knaus UG (2019) Protein-protein interaction assay to analyze NOX4/p22phox heterodimerization. Methods Mol Biol 1982:447–458. https://doi.org/10.1007/978-1-4939-9424-3_26 71. NanoBiT® PPI Starter Systems (2022). https://www.promega.de/ products/protein-interactions/live-cell-protein-interactions/nanobitppi-starter-systems/?catNum=N2014. Accessed 18 Jan 2022 72. O'Neill S, Mathis M, Kovačič L et al (2018) Quantitative interaction analysis permits molecular insights into functional NOX4 NADPH oxidase heterodimer assembly. J Biol Chem 293:8750–8760. https:// doi.org/10.1074/jbc.RA117.001045 73. Lyle AN, Deshpande NN, Taniyama Y et al (2009) Poldip2, a novel regulator of Nox4 and cytoskeletal integrity in vascular smooth muscle cells. Circ Res 105:249–259. https://doi.org/10.1161/ CIRCRESAHA.109.193722 74. Cucoranu I, Clempus R, Dikalova A et al (2005) NAD(P)H oxidase 4 mediates transforming growth factor-beta1-induced differentiation of cardiac fibroblasts into myofibroblasts. Circ Res 97:900–907. https://doi.org/10.1161/01.RES.0000187457.24338.3D 75. Zhang Q-Y, Wang Z-J, Miao L et al (2019) Neuroprotective effect of SCM-198 through stabilizing endothelial cell function. Oxidative Med Cell Longev 2019:7850154. https://doi.org/10.1155/2019/ 7850154 76. Chen X, Xu S, Zhao C et al (2019) Role of TLR4/NADPH oxidase 4 pathway in promoting cell death through autophagy and ferroptosis during heart failure. Biochem Biophys Res Commun 516:37–43. https://doi.org/10.1016/j.bbrc.2019.06.015 77. Yuan S, Hahn SA, Miller MP et al (2021) Cooperation between CYB5R3 and NOX4 via coenzyme Q mitigates endothelial inflammation. Redox Biol 47:102166. https://doi.org/10.1016/j.redox. 2021.102166 78. Casciaro F, Beretti F, Zavatti M et al (2018) Nuclear Nox4 interaction with prelamin a is associated with nuclear redox control of stem cell aging. Aging (Albany NY) 10:2911–2934. https://doi.org/10. 18632/aging.101599 79. Tirone F, Cox JA (2007) NADPH oxidase 5 (NOX5) interacts with and is regulated by calmodulin. FEBS Lett 581:1202–1208. https:// doi.org/10.1016/j.febslet.2007.02.047 80. Roux KJ, Kim D, Burke B et al (2018) BioID: a screen for proteinprotein interactions. Curr Protoc Protein Sci 91:19.23.1–19.23.15. https://doi.org/10.1002/cpps.51

K. Schröder 81. Sears RM, May DG, Roux KJ (2019) BioID as a tool for proteinproximity labeling in living cells. Methods Mol Biol 2012:299–313. https://doi.org/10.1007/978-1-4939-9546-2_15 82. Tsutsumi R, Harizanova J, Stockert R et al (2017) Assay to visualize specific protein oxidation reveals spatio-temporal regulation of SHP2. Nat Commun 8:466. https://doi.org/10.1038/s41467-01700503-w 83. Maraldi T (2013) Natural compounds as modulators of NADPH oxidases. Oxidative Med Cell Longev 2013:271602. https://doi. org/10.1155/2013/271602 84. Reis J, Massari M, Marchese S et al (2020) A closer look into NADPH oxidase inhibitors: validation and insight into their mechanism of action. Redox Biol 32:101466. https://doi.org/10.1016/j. redox.2020.101466 85. Khoshnevisan R, Anderson M, Babcock S et al (2020) NOX1 regulates collective and planktonic cell migration: insights from patients with pediatric-onset IBD and NOX1 deficiency. Inflamm Bowel Dis 26:1166–1176. https://doi.org/10.1093/ibd/izaa017 86. Zhang M, Brewer AC, Schröder K et al (2010) NADPH oxidase-4 mediates protection against chronic load-induced stress in mouse hearts by enhancing angiogenesis. Proc Natl Acad Sci U S A 107: 18121–18126. https://doi.org/10.1073/pnas.1009700107 87. Kopke RD, Liu W, Gabaizadeh R et al (1997) Use of organotypic cultures of Corti's organ to study the protective effects of antioxidant molecules on cisplatin-induced damage of auditory hair cells. Am J Otol 18:559–571 88. Banci L, Bertini I, Blaževitš O et al (2012) Interaction of cisplatin with human superoxide dismutase. J Am Chem Soc 134:7009–7014. https://doi.org/10.1021/ja211591n 89. Mapuskar KA, Steinbach EJ, Zaher A et al (2021) Mitochondrial superoxide dismutase in cisplatin-induced kidney injury. Antioxidants (Basel) 10(9):1329. https://doi.org/10.3390/ antiox10091329 90. Greener JG, Kandathil SM, Moffat L et al (2022) A guide to machine learning for biologists. Nat Rev Mol Cell Biol 23:40–55. https://doi.org/10.1038/s41580-021-00407-0 91. Yousefi PD, Suderman M, Langdon R et al (2022) DNA methylation-based predictors of health: applications and statistical considerations. Nat Rev Genet 23:369–383. https://doi.org/10.1038/ s41576-022-00465-w 92. Alpayd{n E (2020) Introduction to machine learning. In: Adaptive computation and machine learning ser, 4th edn. MIT Press, Cambridge 93. Nagarajan N, Yapp EKY, Le NQK et al (2019) Application of computational biology and artificial intelligence Technologies in Cancer Precision Drug Discovery. Biomed Res Int 2019:8427042. https://doi.org/10.1155/2019/8427042 94. Database APS (2022) AlphaFold Protein structure database. https:// alphafold.ebi.ac.uk/. Accessed 18 Feb 2022 95. Jones DT, Thornton JM (2022) The impact of AlphaFold2 one year on. Nat Methods 19:15–20. https://doi.org/10.1038/s41592-02101365-3 96. Ma Q, Xu D (2022) Deep learning shapes single-cell data analysis. Nat Rev Mol Cell Biol 23:303–304. https://doi.org/10.1038/s41580022-00466-x

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

20

Jacek Zielonka and Matea Juric

Abstract

NADPH oxidases (Noxes) are a family of enzymes that catalyze the oxidation of NADPH by molecular oxygen to produce superoxide radical anion (O2•–) and hydrogen peroxide (H2O2). Formation of the aforementioned reactive oxygen species plays a critical role in immune defense and redox signaling while their overproduction is implicated in numerous pathological conditions associated with increased reactive oxygen species formation and oxidative damage to biomolecules. In this chapter, we review the methods used to detect Nox-derived O2•– and H2O2 in cell-free and cellular systems. We discuss the small molecule spectroscopic probes used, the chemical mechanisms responsible for the formation of the detectable products, and potential limitations of the probes. Finally, we recommend methods to measure O2•– and H2O2. Keywords

NADPH oxidase · Superoxide · Hydrogen peroxide · Probes · Fluorescence · Chemiluminescence · Spin trapping

In a cell-free system, where Nox is the sole consumer of NADPH and oxygen (O2), and producer of NADP+ and O2•–/ H2O2, monitoring any of those species may be sufficient to measure Nox activity [2, 3]. For practical reasons, consumption of NADPH is favored over NADP+ formation, but both O2 consumption and reactive oxygen species (ROS) formation have been utilized in cell-free Nox assays. However, intact cells make the monitoring of Nox activity more challenging as there are multiple cellular metabolic pathways involving substrates and products related to Nox activity. In cells with low mitochondrial oxygen consumption rates (e.g., neutrophils), O2 consumption may be used to monitor Nox activity after subtraction of the mitochondrial contribution [3–5]. However, the most common way to monitor Nox activity in intact cells is to measure ROS (O2•– and/or H2O2) production [6–8]. This is because Nox proteins are the major cellular sources of O2•– and H2O2 in many cases, and function as their emitter to the extracellular milieu. The chemical and mechanistic principles of O2•– and H2O2 detection and quantification in the context of monitoring Nox activity are described in this chapter.

2 1

Introduction

NADPH oxidases (Noxes) are a family of enzymes catalyzing the oxidation of NADPH with a concomitant reduction of molecular oxygen to superoxide radical anion (O2•–), which may dismutate into hydrogen peroxide (H2O2) within the enzyme compartment or in an aqueous environment (Fig. 20.1) [1].

O2•– is the primary product of the Nox-catalyzed reduction of O2, and, as such, its measurement provides one of the most direct ways to monitor Nox activity [1, 2]. As mentioned in the Introduction, however, some Nox isoforms (Nox4, Duox1, Duox2) may covert O2•– into H2O2 before its release from the reduction center into the bulk solution.

2.1 J. Zielonka (✉) · M. Juric Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, USA e-mail: [email protected]; [email protected]

Detection of Nox-Derived Superoxide

Chemical Properties of Superoxide

O2•– is a product of the one-electron reduction of O2, and in protic solvents it exists in equilibrium with its protonated form, hydroperoxyl radical (HO2•, reaction 1) [9, 10]. The

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_20

323

324

J. Zielonka and M. Juric

most convenient methods to detect and quantify O2•–. The assay is based on a single-step electron transfer from O2•– to cyt c(Fe3+) to form ferrocytochrome c (cyt c(Fe2+)) (reaction 3) [12]. O2 • – þ cyt c Fe3þ → O2 þ cyt c Fe2þ

Fig. 20.1 The enzymatic activity of Nox enzymes

pKa value of HO2• is 4.8, which implicates that at pH 7.4, O2•– is the major form (>99%) of superoxide. Nevertheless, the protonation reaction is important, as HO2• formation is responsible for the spontaneous dismutation of O2•– (reaction 2) and for the many other oxidation reactions mediated by O2•–. HO2 • þ H2 O ⇆ O2 • – þ H3 Oþ

ðreaction 1Þ

HO2 • þ O2 • – → HO2 – þ O2

ðreaction 2Þ

The dismutation of O2•– is a pH-dependent process and is relatively fast at pH 7.4 (reaction 2, k = 2 × 105 M-1 s-1) [10]. In biological systems, the conversion of O2•– into H2O2 (represented in reaction 2 by its deprotonated form, HO2-) is further catalyzed by superoxide dismutases (SODs). In addition to undergoing the conversion to H2O2, O2•– may react with metal ions/metalloproteins (e.g., aconitases) or other radical species, including nitric oxide (•NO) and amino acid radicals (free or in proteins) [9]. The reaction of O2•– with • NO results in the formation of peroxynitrite (ONOO-), while radical-radical reactions with amino acid radicals typically result in the formation of amino acid hydroperoxides [9]. Due to the above-mentioned reactivity of O2•–, the lifetime of O2•– in biological systems is very short, resulting in a low steady-state level. Thermodynamically, O2•– may act as both an oxidant (Eo(O2•–,H+/HO2-) = 1.03 V) and as reductant (Eo(O2/ O2•–) = -0.33 V) [11]. However, as previously stated, oxidation reactions are typically mediated by protonation of O2•– to HO2•, which exhibits stronger oxidizing properties (Eo(HO2•,H+/H2O2) = 1.46 V). Due to its low steady-state concentration and unfavorable spectroscopic properties (absorption band in the UV range, with maximum at 245 nm, ε = 2350 M-1 cm-1) [10], the detection and quantification of O2•– are typically accomplished using molecular probes, which upon reaction with superoxide form an easily detectable product.

2.2

Small Molecule Probes for Superoxide

2.2.1 Reduction of Ferricytochrome c Superoxide dismutase-inhibitable reduction of ferricytochrome c (cyt c(Fe3+)) is one of the oldest and

ðreaction 3Þ

Due to a significant difference in the one-electron reduction potentials between oxygen and cyt c(Fe3+) (E(cyt c(Fe3+)/cyt c(Fe2+) = +0.22 V), the reaction is fast (k (cyt c(Fe3+) + O2•–)= 5 × 105 M-1 s-1) and practically irreversible. Formation of cyt c(Fe2+) is accompanied by a change in UV-visible absorption spectrum, with a buildup of a narrow absorption band with a maximum at 550 nm. The difference in molar absorption coefficients between cyt c(Fe2+) and cyt c(Fe3+) of 2.1 × 104 M-1 cm-1 may be used for quantitative analyses of the reduction process. The cyt c(Fe3+)-based reduction is widely used in cell-free assays [2, 13] for different Nox isoforms and has also been applied to monitor Nox-derived O2•– production from intact cells, blood vessels, and in tissue homogenates [14, 15]. The major advantages of cytochrome c-based O2•– measurements are the simplicity of the reaction (one-step process), fast reaction kinetics, 1:1 stoichiometry enabling easy quantification of O2•–, and widespread access to spectrophotometers and plate readers with absorption measurements in biochemical laboratories. It is worth noting that the now routinely used kinetic cyt c(Fe3+) reduction assay in 96-well plates was introduced in 1990 and remains the assay of choice for quantitative analyses of extracellular O2•– generation by activated neutrophils [15]. The major limitations of the assay are the lack of cell permeability and the lack of selectivity toward O2•–. Cyt cbased assays are most often used in cell-free measurements and in experiments with neutrophils or neutrophil-like cells, where O2•– is emitted into the extracellular space. Cyt c(Fe3+) may be reduced directly by redox enzymes (e.g., diaphorases) and non-enzymatic reductants, and attempts have been made to minimize the interaction of cyt c with flavoproteins by way of cyt c acetylation or succinylation. In most cases, the contribution of O2•– to the reduction of cyt c(Fe3+) can be determined by monitoring the extent of inhibition of the reduction process in the presence of SOD.

2.2.2 Reduction of Tetrazolium Salts Another class of probes utilizing the reductive power of O2•– are tetrazolium salts [16, 17]. Historically, the most widely used tetrazolium salt for superoxide detection was nitroblue tetrazolium (NBT), while iodo-nitro-tetrazolium (INT) and more recently water soluble tetrazolium (WST-1) also have been used. In the presence of O2•–, the tetrazolium cation (T+) is reduced to formazan (TH, Fig. 20.2), which can be detected by spectrophotometry. In the case of the INT and WST-1

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

325

Fig. 20.2 Chemical structures of selected terazolium cations (T+) and the reduction products (TH)

cations, the conversion to TH involves two electrons, while in case of NBT, the complete reduction of both tetrazolium moieties requires four electrons. The reduction of T+ to TH by O2•– is, therefore, a multi-step process involving the formation of an intermediate tetrazolinyl radical (T•, reaction 4). O2

•–

þ

þ T → O2 þ T

•

ðreaction 4Þ

Reduction of NBT by O2•– has been characterized kinetically by pulse radiolysis resulting in a rate constant of 5.9 × 104 M-1 s-1 [18]. The rate constant for the reduction of WST-1 by O2•– was estimated as ~3 × 104 M-1 s-1, based on stopped flow measurements [19]. Alternatively, redoxactive enzymes may directly reduce T+ via O2•–-independent one-electron or two-electron reduction producing T• or TH, respectively [20, 21]. Once formed, T• must undergo another one-electron reduction to form the strongly absorbing TH. The best-characterized pathway for such a process is the radical-radical disproportionation reaction, in which one molecule of T• undergoes oxidation to T+ and the other undergoes reduction to TH (reaction 5). 2T • þ Hþ → Tþ þ TH

ðreaction 5Þ

The possibility of T• reoxidation by O2, with the formation of O2•– (reaction 6, a reverse reaction to reaction 4), is a subject

of debate [16], as the ability of SOD to inhibit TH formation is used as both a proof and quantitative measure of O2•– detection by tetrazolium salts. T • þ O2 ⇆ Tþ þ O2 • –

ðreaction 6Þ

If this equilibrium were to occur at the same time scale as the TH formation, the addition of SOD would shift the equilibrium to the right resulting in T• reoxidation and inhibition of TH formation, even if the T+ reduction process did not include O2•– [22]. The major advantage of using tetrazolium salts for O2•– detection include the straightforward detection of the strongly light absorbing TH product, and the relatively high rate constant of their direct reaction with O2•–. A plate readerbased densitometric assay for intracellular O2•– based on NBT reduction in stimulated peritoneal macrophages was reported in 1981 [23]. The availability of highly watersoluble derivatives (WST-1, WST-8) simplifies the experimental protocol for cell-free and extracellular assays [24, 25], and the WST-1 probe was successfully used to monitor the activity of various Nox enzymes in cell-free assays [3, 26]. The electrochemical properties of WST-1 were reported, with the determined one-electron reduction potential Eo(T+/T•) = -0.20 V vs the Ag/AgCl reference electrode [19], demonstrating the feasibility of the WST-1 reduction by O2•– and the shift of the equilibrium described by reaction

326

6 to the left side by three orders of magnitude under typical assay conditions (25 °C, 0.25 mM O2, 0.5 mM WST-1). The major drawback of using tetrazolium salts is their lack of selectivity. They can be reduced to THs by an array of reductants, including reduced enzymatic redox centers [16]. Both the flavin and heme centers of the NADPH oxidase 2 (Nox2) protein were reported to directly reduce tetrazolium salts [27]. Although in cell-free and extracellular assays, the addition of SOD may help establish the contribution of O2•– to the TH formation, the potential occurrence of reaction 6 could result in misinterpretation of the inhibitory effect of SOD [22, 28]. Therefore, further research on the redox chemistry of tetrazolium salts is warranted to establish the conditions and/or chemical structures to avoid the confounding consequences of reaction 6. The application of tetrazolium salts for intracellular O2•– detection is significantly limited due to O2•– -independent reduction pathways. Tetrazolium salts (e.g., MTT) are used as cell viability stains [16, 17], and the availability of efficient and cell-permeable SOD mimetics is limited. Furthermore, increased reduction in cells with higher Nox expression/ activity may be due to a superoxide-independent reduction, as is displayed by increased reduction of INT by liposomes containing cyt b558 in the presence of cytosolic components [29], lack of correlation with cyt c(Fe3+) reduction by activated macrophages [23], as well as in Nox4-expressing human embryonic kidney cells, which showed increased NBT staining without increased superoxide formation [4].

2.2.3 Lucigenin Chemiluminescence One of the most widely utilized assays for NADPH oxidase activity, especially in intact cells and tissue homogenates, is lucigenin (Luc2+)-derived chemiluminescence [30, 31]. The assay is based on the formation of the unstable dioxetane (Luc-O2, Fig. 20.3) derivative of Luc2+, which spontaneously

J. Zielonka and M. Juric

breaks down to two N-methylacridone (MAcrO) molecules, one of which is in the electronically excited state (MAcrO*). The transition from the excited state to the ground state is accompanied by the emission of light, detected as chemiluminescence. To convert Luc2+ to its Luc-O2 derivative, it must first be converted to its one-electron reduced species, Luc•+ (Fig. 20.3), which upon reaction with O2•– produces Luc-O2. Although O2•– is able to reduce Luc2+ to produce Luc•+, this reaction is neither thermodynamically nor kinetically favored. In fact, the reverse reaction in which O2 is reduced by Luc•+ to generate O2•– is expected to predominate [32, 33]. Therefore, in the presence of redox-active enzymes, the ability of Luc2+ to produce O2•– via redox cycling should always be considered [34, 35]. It has been proposed that at low concentrations, the redox cycling activity of Luc2+ may be minimized, and that in many systems the Luc2+-derived luminescence correlates well with other assays for O2•– [31]. However, due to the low rate constant of the reaction of Luc2+ with O2•–, lowering the probe concentration would also affect its ability to intercept O2•– and thus decrease the efficiency of its detection, most notably in the presence of competing O2•– scavengers such as SODs in tissue homogenates. The major advantages of Luc2+-based detection of O2•– are the sensitivity of chemiluminescence detection with a low background, and the relative specificity of the product (excited acridone) to O2•–. Although the dioxetane derivative can also be formed in the presence of H2O2, the reaction requires strong alkaline conditions due to the requirement of H2O2 deprotonation [36]. The major disadvantage of using Luc2+ as a probe for O2•– detection is its ability to redox cycle, resulting in the probemediated production of the species to be detected and overestimation and/or misidentification of the reducing species [37]. Therefore, the data obtained with Luc2+ must be

Fig. 20.3 Scheme of lucigenin reaction with O2•– to produce chemiluminescence and potential generation of O2•– by lucigenin radical

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

validated using probes and assays which lack the ability to produce O2•– while retaining specificity to this molecule. The NADPH-dependence of the chemiluminescence signal should not be used as an indicator of Nox activity, as NADPH-dependent, Luc2+-derived luminescence may be due to cytochrome P450 activity [38, 39].

2.2.4 Oxidation of Luminol and Analogs Another class of chemiluminescent probes is based on the oxidation of luminol and its derivatives in the presence of O2•–. The probes showing a similar mechanism of chemiluminescence generation include luminol, isoluminol, and the L-012 probe [40–43]. Luminol and L-012 have been used in numerous studies elucidating the mechanism of action utilized by different Nox isoforms [44–47]. As with lucigenin, to produce light, the probe is first converted to its radical intermediate, which reacts with O2•– to yield a hydroperoxide derivative, which upon further intramolecular transformations leads to a phthalate (or its analog) in the excited state. Transition from the excited state to a ground state is accompanied by the emission of blue light (Fig. 20.4). The spectral properties and intensity of the emitted light can be modified by the addition or conjugation of the luminol moiety with a fluorophore able to emit at longer wavelengths [48–50]. In contrast to the mechanism of O2•– detection by lucigenin, which involves the one-electron reduction of the probe, luminol-based detection of O2•– involves one-electron oxidation of the probe. As shown for the L-012 probe, oxidation of the probe by O2•– is rather inefficient, and peroxidatic activity (e.g., peroxidases + H2O2) has been proposed as a requirement for the initial oxidation step [51]. Once the luminol-derived radical is formed, it may react with O2•– to produce chemiluminescence. However, the radical may also react with oxygen to produce O2•– [34, 51]. It has been demonstrated that one-electron oxidation of luminol and L-012 by peroxidase in the presence of H2O2 is sufficient to produce chemiluminescence, and the signal is dependent on

327

the presence of O2 and inhibitable by SOD [51–53]. In fact, peroxidase-catalyzed oxidation of luminol has been widely used as a chemiluminescent assay for H2O2 [54]. Although the advantages of luminol-based probes are similar to those of lucigenin which include high sensitivity and the relative specificity of the chemiluminescent signal and the product formed to O2•–, the requirement of peroxidatic oxidation of the probe to its radical makes the probe more useful for a general oxidative environment (e.g., oxidative burst) rather than as a selective probe for O2•–. The possibility of tuning the spectral properties of the luminescence signal via chemical luminescence energy transfer makes luminol-type probes attractive candidates for in vivo imaging of ROS. The major disadvantages of the probe involve its potential to produce O2•– upon one-electron oxidation in the presence of O2, and the possibility of forming probe-derived quinone, which upon reaction with H2O2 would produce the same signal as in the case of O2•–.

2.2.5 Oxidation of Coelenterazine and Analogs Coelenterazine and its analogs (CLA, MCLA, Fig. 20.5) are another class of chemiluminescent probes for O2•– detection [55, 56]. This class of probes has been used to monitor Nox2derived O2•– [57–59]. The reaction proceeds via one-electron oxidation of the probe to its radical, which upon reaction with O2•– and further transformations produces chemiluminescence (Fig. 20.5). Similar to luminol-type probes, the presence of peroxidase and H2O2 has been reported to be sufficient to produce the luminescence signal [60]. In contrast to luminol-type probes, however, no reaction of the radical formed toward O2 has been reported. Due to their relatively low pKa (pKa = 7.6) [61], the coelenterazine-type probes are easily oxidizable, and HO2• present in equilibrium with O2•– is able to convert the probes to their radicals. It has been postulated that the first step of the oxidation process of CLA involves one-electron oxidation of the deprotonated probe by

Fig. 20.4 Scheme of oxidation of luminol and analogs to produce chemiluminescence

328

J. Zielonka and M. Juric

Fig. 20.5 Scheme of oxidation of coelenterazine and analogs to produce chemiluminescence

HO2• [62]. Thus, the presence of peroxidase or strong one-electron oxidants is not required to produce chemiluminescence. It has been shown that singlet oxygen is also able to induce similar chemistry as O2•–, resulting in the chemiluminescence signal [63]. As with other chemiluminescent probes, conjugates with various fluorophores were developed to modify the spectral properties of the luminescence signal [64–66]. The major drawback of using CLA (and related probes) is its susceptibility to autoxidation via the reaction of its anionic form with O2 [61] resulting in high background chemiluminescence, which may limit the use of the probe to systems with relatively high fluxes of O2•–. The presence of additional one-electron oxidants in the system should also be considered when more quantitative analyses are required.

2.2.6 Oxidation of Ethidium-Based Probes Among the array of fluorescent probes for O2•– detection, hydroethidine (HE) and its derivatives (water-soluble and cell membrane-impermeable hydropropidine [HPr+] [67] and mitochondria-targeted analog [MitoSOX Red] [68, 69], Fig. 20.6) are the most widely used in cells and are the probes of choice to detect intracellular O2•– [70, 71]. In the presence of O2•–, HE undergoes conversion to red fluorescent 2-hydroxyethidium (2-OH-E+) [72, 73]. This reaction involves one-electron oxidation of HE to its radical cation, HE•+, by the hydroperoxyl radical, followed by the reaction of the resulting HE•+ with O2•– [74, 75]. HE derivatives involve HPr+ used to monitor Nox2-derived extracellular O2•– [4, 5, 67]. Monitoring the conversion of HE and HPr+ to their respective hydroxylated oxidation products has been the method used to detect Nox2-derived O2•– both in vitro and in vivo [76–80]. The major advantage of this class of probes is the formation of a highly specific product, 2-hydroxyethidium (2-OH-E+), observed only in the presence of O2•–. Because the probe may be oxidized by the HO2• radical [75], the presence of peroxidase or other oxidants is not required,

although it may affect reaction stoichiometry and increase the yield of 2-OH-E+ [81]. The major disadvantage of HE-based assays is the potential formation of other red fluorescent product(s), including ethidium, leading to potential misinterpretation of the fluorescence signal [70]. Therefore, the assay requires identification of the product(s) formed and selective quantification of the superoxide-specific hydroxylated product. This is typically achieved with the use of chromatographic techniques, including (ultra) high-performance liquid chromatography (HPLC) with fluorescence, electrochemical, or mass spectrometric detection [71, 73, 82–84]. The use of cell membraneimpermeable HPr+ minimizes intracellular, O2•–-independent oxidation, and the assay was successfully applied in fluorescence plate reader-based measurements in cell-based assays of Nox2 activity [4, 5, 85]. Quantitative analyses of the hydroxylated products should take into account the potential conversion of HE into its radical, HE•+, via O2•–-independent pathways. The presence of peroxidase has been shown to increase the yield of 2-OH-E+ at a constant flux of O2•– [81]. Should peroxidatic oxidation occur, it can be identified by the chromatographic detection of HE-derived dimeric products (E+-E+ and analogs), characteristic for one-electron oxidation of the probe (Fig. 20.6) [69, 86].

2.2.7 EPR Spin Trapping Electron paramagnetic resonance (EPR) is a technique that allows the direct detection of free radicals, due to the presence of an unpaired electron. Although direct detection of O2•– in biological systems is practically impossible due to a very low steady-state O2•– level, the product of the reaction of O2•– with cyclic nitrones is significantly more persistent, and has been detected using the EPR technique in many biochemical and biological systems (Fig. 20.7) [87– 89]. The conversion of short-lived free radicals (R•) into more persistent radical adducts is called spin trapping, and cyclic nitrones are a class of spin traps particularly useful to detect O2•– [90].

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

329

Fig. 20.6 Scheme of oxidation of hydroethidine and analogs, and the products formed

The spin trapping technique found many applications, particularly in cell-free assays, for characterization of O2•– production in various enzymatic systems, including neutrophil NADPH oxidase, nitric oxide synthases, xanthine oxidase, etc. [91–93]. In addition to cell-free assays, EPR spin

trapping has been applied to monitor Nox2-derived O2•– emitted from intact cells [4, 94–96]. The advantages of the EPR spin trapping of O2•– using cyclic nitrones include the simplicity of the reaction (one-step process) and the formation of the product with a characteristic

Fig. 20.7 Scheme of spin trapping of O2•– by cyclic nitrones and potential reduction products of the spin adduct

330

EPR spectrum that is highly specific for O2•–. In fact, detection of the superoxide adduct is sufficient to confirm the presence of O2•–. The major drawback of the EPR spin trapping technique is the requirement of specialized instrumentation (EPR spectrometer) and appropriate expertise. The low rate constant of the reaction of cyclic nitrones with O2•– is another limitation, which results in the requirement of high concentrations of the spin trap for efficient trapping of O2•–. Furthermore, being that the superoxide adduct is a hydroperoxide, it may undergo conversion to a hydroxyl radical adduct either spontaneously (as in the case of DMPO) or via the action of biological reductants, both non-enzymatic (e.g., ascorbate, glutathione) and enzymatic (e.g., catalyzed by glutathione peroxidases) (Fig. 20.7). This increases the potential of misidentifying the trapped species. Finally, the nitroxide moiety of the adduct may undergo transformation to EPR-silent product(s), e.g., via the reduction to hydroxylamine or oxidation to oxoammonium cation. Cyclic nitroxides are also known to catalytically remove O2•– [97, 98], and this process may affect the quantitative analysis of O2•– when the spin adduct accumulates to the level at which it can compete with the spin trap for O2•–.

2.2.8 Other Probes for O2•– Over the last two decades, several new classes of probes for O2•– detection have been reported. Hydrocyanines, which are two-electron reduction products of fluorescent cyanine dyes, have been proposed for selective detection of O2•– and have been applied to monitor Nox activity [99– 102]. Hydroxylamines, which produce stable nitroxides upon one-electron oxidation are another class of probes proposed for O2•– detection using EPR spectroscopy [103, 104]. Because both hydrocyanines and hydroxylamines involve an oxidation reaction without producing an O2•–specific product, such reactions are expected to be accomplished by other biological oxidants, including ONOO-derived radicals and via peroxidase-catalyzed oxidation reactions. A different approach to detect O2•– is based on its nucleophilic character, and several classes of probes have been reported, including arylsulfonyl [105–107], triflate [108, 109], and diphenylphosphinyl [110–112] derivatives of fluorophores. Some of these probes have already been used to detect O2•– in model biological systems, including to monitor the activity of NADPH oxidases [109, 110, 113]. These classes of redox probes, however, still require full characterization of their chemical reactivity (e.g., reaction mechanism, stoichiometry, kinetics) before they can be used with a high degree of confidence for selective detection and quantitative analyses of O2•–.

J. Zielonka and M. Juric

3

Detection of Hydrogen Peroxide

Hydrogen peroxide is the product of the two-electron reduction of O2 and may be formed directly by many oxidases (e.g., monoamine oxidases) or via spontaneous or SOD-catalyzed dismutation of O2•– [114]. Other potential sources of H2O2 include recombination of hydroxyl radicals (•OH) or hydrolysis of some organic and inorganic hydroperoxides. H2O2 is a major product released from several members of the Nox family of enzymes including Nox4, Duox1, and Duox2 [1].

3.1

Chemical Properties of H2O2

H2O2 is a weak acid (pKa = 11.7), and at physiological pH in an aqueous solution, it is mostly present in its neutral form. In this form, H2O2 acts as an electrophile and is responsible for many oxidative processes in cells, including the oxidation of thiols. However, in its deprotonated form, HO2- (present at pH 7.4 as a 0.005% fraction of the total H2O2 pool), it acts as a nucleophile and is responsible for the oxidation of various classes of redox probes for H2O2. The concentration of H2O2 in water can be determined by spectrophotometry, as it absorbs light in the UV region. Typically, the absorbance at 240 nm is measured and converted into concentration using the extinction coefficient value of 39.4 M-1 cm-1 [115] or 43.6 M-1 cm-1 [116]. In many biological redox reactions, H2O2 acts as a strong two-electron oxidant (Eo(H2O2, 2H+/2H2O) = +1.76 V) [11] and undergoes reduction to water. With the exception of a few enzymatic targets (e.g., peroxiredoxins, catalase, peroxidases), such oxidation reactions are rather slow, and may account for H2O2-induced damage to biomolecules only under extended exposure conditions and/or supraphysiological levels of H2O2 [117]. However, H2O2-induced oxidation reactions can be catalyzed by some of the enzymatic targets mentioned above (e.g., peroxidases). For example, H2O2 can be converted by myeloperoxidase to hypochlorous acid (HOCl), resulting in fast chlorination and oxidation of biomolecules. Peroxidases may also catalyze one-electron oxidation reactions by H2O2, as observed in the case of the oxidation of phenolic substrates (e.g., tyrosine) to phenoxyl radicals and the oxidation of nitrite anion to nitrogen dioxide radical. One-electron oxidation reactions can also be catalyzed by the transition metal ions via the generation of a strong one-electron oxidizing hydroxyl radical (Eo(HO•, H+/H2O) = +2.72 V) in a Fenton reaction (reaction 7) [118]. H2 O2 þ Fe2þ → HO • þ HO– þ Fe3þ

ðreaction 7Þ

20

3.2

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

Small Molecule Probes for H2O2

Due to the low reactivity of H2O2, the most widely used assays rely on the peroxidase-catalyzed oxidation reaction using appropriate probes which produce easily detectable products upon oxidation. Peroxidase-catalyzed oxidation of the fluorescent probes into non-fluorescent products is less commonly used. In all peroxidase-based assays, it is important to avoid additional peroxidase substrates, which may not only compete with the probe, but may also introduce additional complexity to the system. For example, one-electron oxidation of NAD(P)H would result in the formation of the corresponding radical, NAD(P)•, which is known to react rapidly with O2 to produce O2•–.[119, 120] Over the last two decades, new classes of H2O2 probes have been developed, which react directly with H2O2 without the requirement of enzymatic catalysis. These classes of probes may be preferred when the detection of H2O2 is carried out in a solution containing NADPH, as in cell-free Nox assays. For most H2O2 assays, the use of catalase as a catalytic H2O2 scavenger is recommended to confirm the intermediacy of H2O2 in the oxidation of the probe.

3.2.1 Oxidation of Reduced Fluorophores The commonly used probe, dihydrodichlorofluorescein (DCFH), has been designed to detect H2O2 in the presence of peroxidase (typically horseradish peroxidase [HRP]), with the formation of a strongly fluorescent product, DCF (Fig. 20.8a) [121, 122]. The oxidation reaction proceeds via two consecutive one-electron oxidation steps, with the involvement of the phenoxyl radical intermediate (DCF•–). A cell-permeable diacetate derivative of DCFH, DCFH-DA, has been widely used to detect intracellular H2O2. Because the non-catalyzed reaction between DCFH and H2O2 is very slow, a catalyst (e.g., transition metal ions) is required, which

331

makes the oxidation system complex and the interpretation of the experimental data difficult. Analogous oxidation reactions of dihydrorhodamine (DHR, Fig. 20.8b) to the fluorescent rhodamine-123, with the intermediate formation of an anilinyl-type of radical (Rh•), were used to monitor H2O2 and other oxidants in cells [123]. The advantage of these assays is sensitive and real-time detection of H2O2. The requirement of peroxidase in the oxidation system may serve as an additional layer of selectivity for H2O2-dependent oxidation. One of the major limitations of both DCFH and DHR probes is the reactivity of the intermediate probe-derived radicals toward O2 leading to the production of O2•– (Fig. 20.8, dashed arrows) [124–126]. Upon dismutation of O2•–, H2O2 would be formed and drive probe oxidation at the expense of O2 in addition to H2O2. While this may result in “self”-amplification of the signal, and be beneficial when quantitative analysis is not required, this chemistry may result in H2O2 production, even when the initial oxidizing species was different from H2O2. Thus, in such cases, H2O2 formed would propagate oxidation of the probe, and the formation of the product would be catalase-sensitive, resulting in potential misidentification of the oxidant. Additionally, potential confounding effects include photosensitivity of the probe, and the photosensitizing effects of the fluorescent products.[127, 128].

3.2.2 Oxidation of Luminol and Analogs In the presence of peroxidase, luminol and its analogs (e.g., L-012) are oxidized by H2O2 to produce chemiluminescence (as discussed above) [54]. This assay has been widely applied for sensitive monitoring of H2O2 and in the screening campaigns for the identification and validation of NADPH oxidase inhibitors [129–131]. Oxidation of luminol by HRP/H2O2 is a complex, multi-step process, involving O2•–

Fig. 20.8 Scheme of peroxidatic oxidation of DCFH (a) and DHR (b) and potential generation of O2•– by intermediate radicals

332

J. Zielonka and M. Juric

Fig. 20.9 Scheme of peroxidatic oxidation of Amplex Red to fluorescent resorufin

and the reduction of O2, as demonstrated by the inhibition of the signal in the presence of SOD or upon deoxygenation of the solution [51, 52, 132]. The major advantages of the luminol-based assay include high sensitivity of H2O2 detection, and compatibility with high-throughput screening platforms. Due to a complex reaction chemistry involving the consumption of O2 and formation of O2•–, this assay is not quantitative and may produce a chemiluminescence signal even when the initial oxidizing species was not H2O2. In addition, the screening of Nox inhibitors using luminol (or L-012)-based detection may yield peroxidase inhibitors, in addition to bona fide Nox inhibitors [129].

3.2.3 Oxidation of Amplex Red By far, the assay most commonly used to detect H2O2 in extracellular space or in cell-free systems is based on the peroxidase-catalyzed oxidation of Amplex Red (10-acetyl3,7-dihydroxyphenoxazine) to red fluorescent resorufin (λexc: 570 nm, λemi: 583 nm) [86, 133, 134]. The oxidation reaction involves intermediary formation of the phenoxyl radical derivative followed by another one-electron oxidation and deacetylation steps (Fig. 20.9) [135, 136]. In contrast to the H2O2 probes described above (DCFH, DHR, luminol), no reaction of the intermediate radical with O2 has been reported, and quantitative conversion of Amplex Red into resorufin has been observed. The advantages of the Amplex Red assay include high sensitivity, compatibility with real-time monitoring of H2O2 production, and the quantitative nature of H2O2 detection. Due to these properties, the Amplex Red-based assay is exceedingly used in high-throughput screening campaigns searching for new Nox inhibitors, and for establishing their inhibitory potency [3, 5, 26, 137, 138]. The major weaknesses of the assay include the probe’s photosensitivity and photosensitization by the oxidation product, resorufin [139–141]. Therefore, samples require protection from light, and real-time monitoring of probe oxidation should be complemented by end-point measurements of samples incubated without constant exposure to the excitation light. Potential involvement of redox cycling by resorufin via diaphorase activity should be

considered when quantitative analysis is required [142– 144]. As with other peroxidase-dependent assays, peroxidase substrates and inhibitors will interfere with the assay, as has been shown for NADH, glutathione, dietary antioxidants, and anthracyclines [145–147].

3.2.4 Oxidation of Other Phenolic Probes Peroxidase-catalyzed oxidation of simple phenolic (PhOH) probes by H2O2 has also been applied to monitor H2O2. The relevant probes used include tyrosine (Tyr) [148], p-hydroxyphenylacetic acid (PHPA) [149, 150], and homovanillic acid (HVA) [151, 152], which undergo one-electron oxidation in the presence of H2O2 and peroxidase to form the corresponding phenoxyl radicals (PhO•, Fig. 20.10). Phenoxyl radicals undergo subsequent dimerization to produce fluorescent diphenolic products, (PhOH)2, which can be detected directly by spectrofluorometry or by HPLC analysis using instruments equipped with a fluorescence detector. Peroxidase-catalyzed oxidation of other phenolic compounds, and monitoring their consumption (e.g., scopoletin [153, 154]) or the formation of the oxidation product(s) (e.g., phenol red oxidation product(s) [155–157]) has been also reported and used for H2O2 measurement. The advantages of the assay are exemplified by its relative simplicity and the lack of reduction of O2 to O2•– by the reaction intermediates. Unsurprisingly, the peroxidasecatalyzed oxidation of phenolic probes has been a popular assay in studies of H2O2 generation by different members of the Nox family of enzymes [158–161]. The weaknesses of the assay include subpar sensitivity and the spectral properties of the fluorescent product, which may overlap with endogenous cellular fluorophores. In such cases, HPLC-based separation and quantification of the diphenolic product is recommended. 3.2.5 Oxidation of Boronates Over the last two decades, new classes of redox probes have been developed that react with H2O2 without the requirement of a catalyst. Of those, the boronate-based probes garnered intense interest, and numerous fluorophores and reporters based on other detection modalities have been derivatized to introduce a redox-sensing boronate moiety, with the goal

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

333

Fig. 20.10 Scheme of peroxidatic oxidation of phenolic probes to diphenolic products

of detecting and imaging H2O2 in biochemical and biological systems [162–164]. Peroxidase is not required for these probes; therefore, the possibility of a change in probe response due to assay interference decreases, and there is an opportunity for rigorous characterization of NADPH oxidase inhibitors [3–5, 85, 165–168]. The reaction of boronates (boronic acids and esters) with H2O2 leads to the replacement of the boronate moiety by a hydroxyl (-OH) group to produce a reporting molecule (Fig. 20.11). Since many fluorophores are phenolic-type compounds (e.g., umbelliferone, fluorescein, resorufin), replacement of the phenolic hydroxyl group by a boronate moiety provides a convenient way to design H2O2 fluorogenic probes (Fig. 20.11a). Numerous probes have been designed, where the boronate-based sensing moiety is linked to a fluorophore via a self-immolative benzyl linker (Fig. 20.11b and c). In such case, both phenol- and anilinetype fluorophores can be derivatized for H2O2 detection, and form an intermediate phenolic product upon reaction with H2O2. The primary phenolic product is not stable and undergoes quinone methide (QM) elimination to produce a more stable phenolic product (Fig. 20.11b), or via elimination of QM and carbon dioxide (CO2), it may produce an anilinetype reporter (Fig. 20.11c). The reaction between boronate probes and H2O2 involves its deprotonated form (HOO-), which is present in only ~0.005% of the total amount of H2O2 at pH 7.4. This results in a relatively low rate constant for the reaction under physiological conditions (k ~ 100 M-1 s-1 at pH 7.4) [164, 169, 170]. The described oxidative chemistry of boronates is not, however, limited to H2O2, and several other biological oxidants have been shown to react with boronate compounds [163, 164]. It has been observed that boronates can be oxidized by acidic amino acid hydroperoxides (AA-OOH, k ~ 101 M-1 s-1) [171], peroxymonocarbonate (HCO4-, k ~ 102 M-1 s-1) [172], hypochlorous acid (HOCl, k ~ 104 M-1 s-1) [169], and peroxynitrite (ONOO-, k ~ 106 M-1 s-1) [169, 170]. Such

reactions provide the opportunity to use boronate probes to monitor those species in cellular and cell-free systems. The formation of HCO4- from the reaction between H2O2 and CO2 may improve the performance of boronate probes for H2O2 detection, where CO2 would act as a catalyst [83], and differentiate H2O2 from ONOO-, for which the presence of CO2 results in decreased probe oxidation [169, 173]. The involvement of ONOO- in oxidation of boronate probes can be also identified by profiling the products formed and detecting ONOO--specific minor products [5, 83, 173– 176]. The major advantages of boronate-based probes for H2O2 are that they involve a direct reaction without the requirement of any catalyst, have 1:1 stoichiometry, and allow the same chemical principles to be applied to an array of chemical reporters for different detection modalities (spectrophotometry, fluorescence, chemiluminescence, bioluminescence, positron emission tomography, etc.) and different spectral regions [162–164]. One of the limitations of boronate-based probes is that their relatively slow reaction kinetics is an impediment to quantitative real-time H2O2 monitoring and sensitive detection of intracellular H2O2. Raising the probe concentration may at least partially overcome this limitation in the case of cell-free assays or when monitoring extracellular H2O2. Involvement of the self-immolative linker in the probe design (Fig. 20.11b and c) introduces kinetic complexity as the rate of formation of the detectable product (reporter) may be limited by the rate of elimination of the linker. Another limitation for H2O2 detection is the ability of boronates to react with several other biologically relevant two-electron oxidants. Therefore, appropriate control experiments, including the use of inhibitors of nitric oxide synthases and myeloperoxidases to prevent the formation of ONOO- and HOCl as well as the use of catalase to confirm the involvement of H2O2 in probe oxidation, should be included in the experimental protocol whenever feasible.

334

J. Zielonka and M. Juric

Fig. 20.11 Scheme of oxidation of boronate probes to phenolic- (a, b) and anilinyl- (c) type products

3.2.6 Other Probes for H2O2 In addition to boronates, other classes of small molecule probes have been developed in recent years. These typically are based on the oxidative chemistry of H2O2 initiated by the nucleophilic attack of its deprotonated form, HOO-, on an electrophilic center of the probe. Such probes include, e.g., α-keto acids [177], α-keto amides [178], α,β-diketones [179, 180], and benzenesulfonates [113, 181–183] as the H2O2-sensing moieties. The subsequent chemistry after the nucleophilic attack depends on the actual probe design, and may involve the formation of carboxylic acids, and phenoland aniline-type products as the reporting molecules. Therefore, similar to boronate-based probes, a wide range of reporters can be designed by taking advantage of the oxidative chemistry between H2O2 and these classes of probes. Oxalate derivatives are a special case of α,β-diketones that, in combination with fluorescent dyes, have been proposed for chemiluminescent detection of H2O2. Upon reaction with H2O2, the high-energy intermediate interacts with the fluorophore to chemically excite it, resulting in the emission of chemiluminescence [184–187]. Although the aforementioned classes of probes have already been successfully applied for the detection of H2O2 in biological systems, at least some of those probes are also expected to be oxidized by other biologically relevant oxidants (e.g., HOCl, ONOO-) [187]. More data on the reaction mechanism, kinetics, and selectivity are needed to

accurately interpret the experimental results, in order to define their optimal applications and understand their limitations. Also, knowledge of a probe’s chemistry provides a basis for potential future improvement in their detection performance. In addition to the small molecule probes for H2O2 detection described above, the advent of genetically encoded fluorescent probes for H2O2 provides complementary tools for real-time monitoring and imaging of H2O2 production and dynamics in cells, which are applicable in studies of Nox-dependent redox signaling. This topic is the subject of numerous review articles [188–190] and is outside of the scope of this chapter.

4

Simultaneous Detection of O2•– And H2O2

Simultaneous monitoring of O2•– and H2O2 in the extracellular medium has been proposed to identify the species released from different Nox isoforms [4]. This approach is based on the incubation of cells with a combination of the probes for O2•– (hydroethidine) and for H2O2 (coumarin boronic acid), with subsequent rapid HPLC-based separation and quantification of both probes and oxidation products formed. Due to the fast analysis, the same samples (cell media) can be repeatedly analyzed over time to provide real time-type

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

oxidant generation profiles [4]. Such an approach also could be used for analyses of intracellular oxidants, but this would require sample destruction for analysis, and thus would be an end-point measurement. Future combinations of various probes for O2•– and H2O2 may allow real-time, fluorescence-based analyses without the requirement of HPLCbased separation.

5

Recommendations and Outlook

Application of redox probes with established reaction chemistry, known reaction kinetics, and known selectivity/specificity is a sine qua non for the rigorous characterization of O2•– and H2O2 production by NADPH oxidases in cell-free and cellular systems. Reliable high-throughput screeningcompatible probes are also needed to identify novel chemical scaffolds for selective inhibitors of different Nox isoforms. In fact, a recent report demonstrates that most commonly used Nox inhibitors failed to show inhibitory activity in cell-free models of Nox isoforms, suggesting that interference with the detection method is a common problem in the characterization of potential Nox inhibitors [3, 168]. Being that most probes for H2O2 require peroxidase, the interference with the HRP-catalyzed probe oxidation or inhibition of myeloperoxidase present in isolated membranes may lead to false identification of new inhibitors. At present, HE (and its water soluble, cell membraneimpermeant derivative HPr+)-based assays for O2•– and boronate-based probes for H2O2 are recommended based on the current understanding of the chemistry, stoichiometry, and reaction kinetics for both classes of probes. It is imperative that HE-based measurements are complemented by HPLC-based identification of the O2•–-specific hydroxylated product. In systems utilizing boronate-based probes for H2O2 detection, whereby the formation of HOCl or ONOO- is possible, implementation of myeloperoxidase and nitric oxide synthase inhibitors as well as HPLC-based profiling of the products formed may be necessary to confirm the identity of the oxidant. Whenever possible, redox assays should involve the use of O2•– and H2O2 catalytic scavengers (SOD and catalase, respectively) to confirm their involvement in the conversion of the probe into its detectable product. It should be emphasized, however, that in the case of probes capable of producing O2•– and H2O2 (e.g., DCFH, DHR, lucigenin, L-012) such controls may be insufficient and may result in misidentification of the species responsible for the initial probe oxidation. Over the last two decades, immense progress has been made in chemical biology and a deluge of reports have been published on new redox-sensitive probes, including those for O2•– and H2O2. Many of these probes are expected to help

335

establish rigorous, plate reader-based assays for real-time and high-throughput monitoring of Nox-derived ROS. However, before their widespread use, the characterization of their reaction mechanism, kinetics, and stoichiometry and of the reactivity of any reaction intermediates and products are needed. With this knowledge, the meaningful application of such probes both in vitro and in vivo may be possible, furthering progress in understanding the role of the NADPH oxidase family of enzymes in physiology and disease. Acknowledgments The authors would like to thank Lydia Washechek for her superb editorial work on the manuscript.

References 1. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313. https://doi.org/10.1152/physrev.00044.2005 2. Pick E (2020) Cell-free NADPH oxidase activation assays: a triumph of reductionism. Methods Mol Biol 2087:325–411. https:// doi.org/10.1007/978-1-0716-0154-9_23 3. Augsburger F et al (2019) Pharmacological characterization of the seven human NOX isoforms and their inhibitors. Redox Biol 26: 101272. https://doi.org/10.1016/j.redox.2019.101272 4. Zielonka J et al (2014) High-throughput assays for superoxide and hydrogen peroxide: design of a screening workflow to identify inhibitors of NADPH oxidases. J Biol Chem 289:16176–16189. https://doi.org/10.1074/jbc.M114.548693 5. Zielonka J et al (2016) Mitigation of NADPH oxidase 2 activity as a strategy to inhibit Peroxynitrite formation. J Biol Chem 291: 7029–7044. https://doi.org/10.1074/jbc.M115.702787 6. Kalyanaraman B, Hardy M, Podsiadly R, Cheng G, Zielonka J (2017) Recent developments in detection of superoxide radical anion and hydrogen peroxide: opportunities, challenges, and implications in redox signaling. Arch Biochem Biophys 617:38– 47. https://doi.org/10.1016/j.abb.2016.08.021 7. Kalyanaraman B, Hardy M, Zielonka J (2016) A critical review of methodologies to detect reactive oxygen and nitrogen species stimulated by NADPH oxidase enzymes: implications in pesticide toxicity. Curr Pharmacol Rep 2:193–201. https://doi.org/10.1007/ s40495-016-0063-0 8. Nauseef WM (2014) Detection of superoxide anion and hydrogen peroxide production by cellular NADPH oxidases. Biochim Biophys Acta 1840:757–767. https://doi.org/10.1016/j.bbagen. 2013.04.040 9. Winterbourn CC (2020) Biological chemistry of superoxide radicals. ChemTexts 6:7. https://doi.org/10.1007/s40828-0190101-8 10. Bielski BH, Cabelli DE (1991) Highlights of current research involving superoxide and perhydroxyl radicals in aqueous solutions. Int J Radiat Biol 59:291–319. https://doi.org/10.1080/ 09553009114550301 11. Armstrong DA et al (2015) Standard electrode potentials involving radicals in aqueous solution: inorganic radicals (IUPAC technical report). Pure Appl Chem 87:1139–1150. https://doi.org/10.1515/ pac-2014-0502 12. Koppenol WH, van Buuren KJ, Butler J, Braams R (1976) The kinetics of the reduction of cytochrome c by the superoxide anion radical. Biochim Biophys Acta 449:157–168. https://doi.org/10. 1016/0005-2728(76)90130-4

336 13. Shiose A et al (2001) A novel superoxide-producing NAD(P)H oxidase in kidney. J Biol Chem 276:1417–1423. https://doi.org/10. 1074/jbc.M007597200 14. Guzik TJ et al (2006) Coronary artery superoxide production and Nox isoform expression in human coronary artery disease. Arterioscler Thromb Vasc Biol 26:333–339. https://doi.org/10. 1161/01.ATV.0000196651.64776.51 15. Mayo LA, Curnutte JT (1990) Kinetic microplate assay for superoxide production by neutrophils and other phagocytic cells. Methods Enzymol 186:567–575. https://doi.org/10.1016/00766879(90)86151-k 16. Berridge MV, Herst PM, Tan AS (2005) Tetrazolium dyes as tools in cell biology: new insights into their cellular reduction. In: Biotechnology annual review, vol 11. Elsevier, pp 127–152. https:// doi.org/10.1016/S1387-2656(05)11004-7 17. Stockert JC, Horobin RW, Colombo LL, Blázquez-Castro A (2018) Tetrazolium salts and formazan products in cell biology: viability assessment, fluorescence imaging, and labeling perspectives. Acta Histochem 120:159–167. https://doi.org/10. 1016/j.acthis.2018.02.005 18. Bielski BHJ, Shiue GG, Bajuk S (1980) Reduction of nitro blue tetrazolium by CO2- and O2- radicals. J Phys Chem 84:830–833. https://doi.org/10.1021/j100445a006 19. Oritani T, Fukuhara N, Okajima T, Kitamura F, Ohsaka T (2004) Electrochemical and spectroscopic studies on electron-transfer reaction between novel water-soluble tetrazolium salts and a superoxide ion. Inorg Chim Acta 357:436–442. https://doi.org/10.1016/ j.ica.2003.05.007 20. Thayer WS (1990) Superoxide-dependent and superoxideindependent pathways for reduction of nitroblue tetrazolium in isolated rat cardiac myocytes. Arch Biochem Biophys 276:139– 145. https://doi.org/10.1016/0003-9861(90)90020-y 21. Schor NA, Stedman RB, Epstein N, Schally G (1982) Rat splenic D-T diaphorase and NAD(P)H-nitroblue tetrazolium reductase. Their use to assess the action of polycyclic hydrocarbons in the lymphatic system. Virchows Arch B Cell Pathol Incl Mol Pathol 41:83–93. https://doi.org/10.1007/BF02890273 22. Auclair C, Torres M, Hakim J (1978) Superoxide anion involvement in NBT reduction catalyzed by NADPH-cytochrome P-450 reductase: a pitfall. FEBS Lett 89:26–28. https://doi.org/10.1016/ 0014-5793(78)80514-6 23. Pick E, Charon J, Mizel D (1981) A rapid densitometric microassay for nitroblue tetrazolium reduction and application of the microassay to macrophages. J Reticuloendothel Soc 30:581–593 24. Tan AS, Berridge MV (2000) Superoxide produced by activated neutrophils efficiently reduces the tetrazolium salt, WST-1 to produce a soluble formazan: a simple colorimetric assay for measuring respiratory burst activation and for screening anti-inflammatory agents. J Immunol Methods 238:59–68. https://doi.org/10.1016/ s0022-1759(00)00156-3 25. Berridge MV, Tan AS (1998) Trans-plasma membrane electron transport: a cellular assay for NADH- and NADPH-oxidase based on extracellular, superoxide-mediated reduction of the sulfonated tetrazolium salt WST-1. Protoplasma 205:74–82. https://doi.org/ 10.1007/BF01279296 26. Seredenina T et al (2015) A subset of N-substituted phenothiazines inhibits NADPH oxidases. Free Radic Biol Med 86:239–249. https://doi.org/10.1016/j.freeradbiomed.2015.05.023 27. Poinas A, Gaillard J, Vignais P, Doussiere J (2002) Exploration of the diaphorase activity of neutrophil NADPH oxidase. Eur J Biochem 269:1243–1252. https://doi.org/10.1046/j.1432-1033. 2002.02764.x 28. Auclair C, Voisin E (1985) Nitroblue tetrazolium reduction. In: Greenwald RA (ed) CRC handbook of methods for oxygen radical research. CRC Press, Boca Raton, pp 123–132. https://doi.org/10. 1201/9781351072922

J. Zielonka and M. Juric 29. Li J, Guillory RJ (1997) Purified leukocyte cytochrome b558 incorporated into liposomes catalyzes a cytosolic factor dependent Diaphorase activity. Biochemistry 36:5529–5537. https://doi.org/ 10.1021/bi963013r 30. Münzel T, Afanas’ev IB, Kleschyov AL, Harrison DG (2002) Detection of superoxide in vascular tissue. Arterioscler Thromb Vasc Biol 22:1761–1768. https://doi.org/10.1161/01.atv. 0000034022.11764.ec 31. Li Y, Zhu H, Kuppusamy P, Roubaud V, Zweier JL, Trush MA (1998) Validation of lucigenin (bis-N-methylacridinium) as a chemilumigenic probe for detecting superoxide anion radical production by enzymatic and cellular systems. J Biol Chem 273:2015– 2023. https://doi.org/10.1074/jbc.273.4.2015 32. Wardman P, Burkitt MJ, Patel KB, Lawrence A, Jones CM, Everett SA, Vojnovic B (2002) Pitfalls in the use of common luminescent probes for oxidative and Nitrosative stress. J Fluoresc 12:65–68. https://doi.org/10.1023/A:1015363220266 33. Spasojević I, Liochev SI, Fridovich I (2000) Lucigenin: redox potential in aqueous media and redox cycling with O-2 Production1. Arch Biochem Biophys 373:447–450. https://doi.org/10. 1006/abbi.1999.1579 34. Faulkner K, Fridovich I (1993) Luminol and lucigenin as detectors for O2•-. Free Radic Biol Med 15:447–451. https://doi.org/10. 1016/0891-5849(93)90044-u 35. Liochev SI, Fridovich I (1997) Lucigenin (bis-Nmethylacridinium) as a mediator of superoxide anion production. Arch Biochem Biophys 337:115–120. https://doi.org/10.1006/ abbi.1997.9766 36. Maskiewicz R, Sogah D, Bruice TC (1979) Chemiluminescent reactions of lucigenin. 1. Reactions of lucigenin with hydrogen peroxide. J Am Chem Soc 101:5347–5354. https://doi.org/10. 1021/ja00512a040 37. Janiszewski M, Souza HP, Liu X, Pedro MA, Zweier JL, Laurindo FR (2002) Overestimation of NADH-driven vascular oxidase activity due to lucigenin artifacts. Free Radic Biol Med 32:446– 453. https://doi.org/10.1016/s0891-5849(01)00828-0 38. Rezende F et al (2016) Unchanged NADPH oxidase activity in Nox1-Nox2-Nox4 triple knockout mice: what do NADPHstimulated Chemiluminescence assays really detect? Antioxid Redox Signal 24:392–399. https://doi.org/10.1089/ars.2015.6314 39. Rezende F et al (2017) Cytochrome P450 enzymes but not NADPH oxidases are the source of the NADPH-dependent lucigenin chemiluminescence in membrane assays. Free Radic Biol Med 102:57–66. https://doi.org/10.1016/j.freeradbiomed. 2016.11.019 40. Vilim V, Wilhelm J (1989) What do we measure by a luminoldependent chemiluminescence of phagocytes? Free Radic Biol Med 6:623–629. https://doi.org/10.1016/0891-5849(89)90070-1 41. Daiber A et al (2004) Measurement of NAD(P)H oxidase-derived superoxide with the luminol analogue L-012. Free Radic Biol Med 36:101–111. https://doi.org/10.1016/j.freeradbiomed.2003.10.012 42. Nishinaka Y, Aramaki Y, Yoshida H, Masuya H, Sugawara T, Ichimori Y (1993) A new sensitive chemiluminescence probe, L-012, for measuring the production of superoxide anion by cells. Biochem Biophys Res Commun 193:554–559. https://doi.org/10. 1006/bbrc.1993.1659 43. Imada I, Sato EF, Miyamoto M, Ichimori Y, Minamiyama Y, Konaka R, Inoue M (1999) Analysis of reactive oxygen species generated by neutrophils using a chemiluminescence probe L-012. Anal Biochem 271:53–58. https://doi.org/10.1006/abio.1999.4107 44. Takac I et al (2011) The E-loop is involved in hydrogen peroxide formation by the NADPH oxidase Nox4. J Biol Chem 286:13304– 13313. https://doi.org/10.1074/jbc.M110.192138 45. Cheng G, Ritsick D, Lambeth JD (2004) Nox3 regulation by NOXO1, p47phox, and p67phox. J Biol Chem 279:34250– 34255. https://doi.org/10.1074/jbc.M400660200

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

46. Ueno N, Takeya R, Miyano K, Kikuchi H, Sumimoto H (2005) The NADPH oxidase Nox3 constitutively produces superoxide in a p22phoxdependent manner: its regulation by oxidase organizers and activators. J Biol Chem 280:23328–23339. https://doi.org/10. 1074/jbc.M414548200 47. Al Ghouleh I et al (2013) Aquaporin 1, Nox1, and Ask1 mediate oxidant-induced smooth muscle cell hypertrophy. Cardiovasc Res 97:134–142. https://doi.org/10.1093/cvr/cvs295 48. Navas Díaz A, González García JA, Lovillo J (1997) Enhancer effect of fluorescein on the luminol-H2O2-horseradish peroxidase chemiluminescence: energy transfer process. J Biolumin Chemilumin 12:199–205. https://doi.org/10.1002/(SICI)10991271(199707/08)12:43.0.CO;2-U 49. Lee ES et al (2016) Nanoparticles based on quantum dots and a luminol derivative: implications for in vivo imaging of hydrogen peroxide by chemiluminescence resonance energy transfer. Chem Commun (Camb) 52:4132–4135. https://doi.org/10.1039/ c5cc09850e 50. Zhang N, Francis KP, Prakash A, Ansaldi D (2013) Enhanced detection of myeloperoxidase activity in deep tissues through luminescent excitation of near-infrared nanoparticles. Nat Med 19:500– 505. https://doi.org/10.1038/nm.3110 51. Zielonka J, Lambeth JD, Kalyanaraman B (2013) On the use of L-012, a luminol-based chemiluminescent probe, for detecting superoxide and identifying inhibitors of NADPH oxidase: a reevaluation. Free Radic Biol Med 65:1310–1314. https://doi.org/10. 1016/j.freeradbiomed.2013.09.017 52. Misra HP, Squatrito PM (1982) The role of superoxide anion in peroxidase-catalyzed chemiluminescence of luminol. Arch Biochem Biophys 215:59–65. https://doi.org/10.1016/0003-9861 (82)90278-8 53. Sundqvist T (1991) Bovine aortic endothelial cells release hydrogen peroxide. J Cell Physiol 148:152–156. https://doi.org/10.1002/ jcp.1041480118 54. Khan P et al (2014) Luminol-based chemiluminescent signals: clinical and non-clinical application and future uses. Appl Biochem Biotechnol 173:333–355. https://doi.org/10.1007/s12010-0140850-1 55. Teranishi K, Shimomura O (1997) Coelenterazine analogs as chemiluminescent probe for superoxide anion. Anal Biochem 249:37–43. https://doi.org/10.1006/abio.1997.2150 56. Teranishi K (2007) Luminescence of imidazo[1,2-a]pyrazin-3 (7H)-one compounds. Bioorg Chem 35:82–111. https://doi.org/ 10.1016/j.bioorg.2006.08.003 57. Nakano M, Sugioka K, Ushijima Y, Goto T (1986) Chemiluminescence probe with Cypridina luciferin analog, 2-methyl-6-phenyl3,7-dihydroimidazo[1,2-a]pyrazin-3-one, for estimating the ability of human granulocytes to generate O2-. Anal Biochem 159:363– 369. https://doi.org/10.1016/0003-2697(86)90354-4 58. Sugioka K, Nakano M, Kurashige S, Akuzawa Y, Goto T (1986) A chemiluminescent probe with a Cypridina luciferin analog, 2-methyl-6-phenyl-3,7-dihydroimidazo[1,2-a]pyrazin-3-one, specific and sensitive for O2- production in phagocytizing macrophages. FEBS Lett 197:27–30. https://doi.org/10.1016/ 0014-5793(86)80291-5 59. Lucas M, Solano F (1992) Coelenterazine is a superoxide anionsensitive chemiluminescent probe: its usefulness in the assay of respiratory burst in neutrophils. Anal Biochem 206:273–277. https://doi.org/10.1016/0003-2697(92)90366-f 60. Mitani M, Yokoyama Y, Ichikawa S, Sawada H, Matsumoto T, Fujimori K, Kosugi M (1994) Determination of horserdish peroxidase concentration using the chemiluminescence of Cypridina luciferin analogue, 2 methyl-6-(p-methyoxyphenyl)-3,7dihydroimidazo[1,2-a]pyrazin-3-one. J Biolumin Chemilumin 9: 355–361. https://doi.org/10.1002/bio.1170090602

337

61. Fujimori K, Nakajima H, Akutsu K, Mitani M, Sawada H, Nakayama M (1993) Chemiluminescence of Cypridina luciferin analogues. Part 1. Effect of pH on rates of spontaneous autoxidation of CLA in aqueous buffer solutions. J Chem Soc Perkin Trans 2:2405–2409. https://doi.org/10.1039/P29930002405 62. Akutsu K, Nakajima H, Katoh T, Kino S, Fujimori K (1995) Chemiluminescence of Cipridina luciferin analogues. Part 2. Kinetic studies on the reaction of 2-methyl-6-phenylimidazo [1,2-a]pyrazin-3(7H)-one (CLA) with superoxide: hydroperoxyl radical is an actual active species used to initiate the reaction. J Chem Soc Perkin Trans 2:1699–1706. https://doi.org/10.1039/ P29950001699 63. Fujimori K et al (1998) Chemiluminescence of Cypridina luciferin analogs. Part 3. MCLA Chemiluminescence with singlet oxygen generated by the retro-Diels-Alder reaction of a naphthalene Endoperoxide. Photochem Photobiol 68:143–149. https://doi.org/10. 1111/j.1751-1097.1998.tb02481.x 64. Teranishi K (2007) Development of imidazopyrazinone red-chemiluminescent probes for detecting superoxide anions via a chemiluminescence resonance energy transfer method. Luminescence 22:147–156. https://doi.org/10.1002/bio.939 65. Sekiya M, Umezawa K, Sato A, Citterio D, Suzuki K (2009) A novel luciferin-based bright chemiluminescent probe for the detection of reactive oxygen species. Chem Commun. https://doi.org/10. 1039/b903751a 66. Teranishi K, Nishiguchi T (2004) Cyclodextrin-bound 6-(4-methoxyphenyl)imidazo[1,2-alpha+/-]pyrazin-3(7H)-ones with fluorescein as green chemiluminescent probes for superoxide anions. Anal Biochem 325:185–195. https://doi.org/10.1016/j.ab. 2003.10.042 67. Michalski R, Zielonka J, Hardy M, Joseph J, Kalyanaraman B (2013) Hydropropidine: a novel, cell-impermeant fluorogenic probe for detecting extracellular superoxide. Free Radic Biol Med 54:135–147. https://doi.org/10.1016/j.freeradbiomed.2012.09.018 68. Robinson KM et al (2006) Selective fluorescent imaging of superoxide in vivo using ethidium-based probes. Proc Natl Acad Sci U S A 103:15038–15043. https://doi.org/10.1073/pnas.0601945103 69. Zielonka J et al (2008) Cytochrome c-mediated oxidation of hydroethidine and Mito-hydroethidine in mitochondria: identification of homo- and heterodimers. Free Radic Biol Med 44:835–846. https://doi.org/10.1016/j.freeradbiomed.2007.11.013 70. Zielonka J, Kalyanaraman B (2010) Hydroethidine- and MitoSOXderived red fluorescence is not a reliable indicator of intracellular superoxide formation: another inconvenient truth. Free Radic Biol Med 48:983–1001. https://doi.org/10.1016/j.freeradbiomed.2010. 01.028 71. Kalyanaraman B, Dranka BP, Hardy M, Michalski R, Zielonka J (2014) HPLC-based monitoring of products formed from hydroethidine-based fluorogenic probes--the ultimate approach for intra- and extracellular superoxide detection. Biochim Biophys Acta 1840:739–744. https://doi.org/10.1016/j.bbagen.2013.05.008 72. Zhao H, Joseph J, Fales HM, Sokoloski EA, Levine RL, VasquezVivar J, Kalyanaraman B (2005) Detection and characterization of the product of hydroethidine and intracellular superoxide by HPLC and limitations of fluorescence. Proc Natl Acad Sci U S A 102: 5727–5732. https://doi.org/10.1073/pnas.0501719102 73. Zielonka J, Vasquez-Vivar J, Kalyanaraman B (2008) Detection of 2-hydroxyethidium in cellular systems: a unique marker product of superoxide and hydroethidine. Nat Protoc 3:8–21. https://doi.org/ 10.1038/nprot.2007.473 74. Zielonka J, Sarna T, Roberts JE, Wishart JF, Kalyanaraman B (2006) Pulse radiolysis and steady-state analyses of the reaction between hydroethidine and superoxide and other oxidants. Arch Biochem Biophys 456:39–47. https://doi.org/10.1016/j.abb.2006. 09.031

338 75. Michalski R et al (2020) Oxidation of ethidium-based probes by biological radicals: mechanism, kinetics and implications for the detection of superoxide. Sci Rep 10:18626. https://doi.org/10. 1038/s41598-020-75373-2 76. Tang PC et al (2008) MyD88-dependent, superoxide-initiated inflammation is necessary for flow-mediated inward remodeling of conduit arteries. J Exp Med 205:3159–3171. https://doi.org/10. 1084/jem.20081298 77. Fernandes DC, Gonçalves RC, Laurindo FRM (2017) Measurement of superoxide production and NADPH oxidase activity by HPLC analysis of Dihydroethidium oxidation. In: Touyz RM, Schiffrin EL (eds) Hypertension: methods and protocols. Springer, New York, pp 233–249. https://doi.org/10.1007/978-1-4939-66257_19 78. Fernandes DC, Wosniak J Jr, Pescatore LA, Bertoline MA, Liberman M, Laurindo FR, Santos CX (2007) Analysis of DHE-derived oxidation products by HPLC in the assessment of superoxide production and NADPH oxidase activity in vascular systems. Am J Phys Cell Physiol 292:C413–C422. https://doi.org/ 10.1152/ajpcell.00188.2006 79. Seredenina T et al (2016) Evaluation of NADPH oxidases as drug targets in a mouse model of familial amyotrophic lateral sclerosis. Free Radic Biol Med 97:95–108. https://doi.org/10.1016/j. freeradbiomed.2016.05.016 80. Kalinovic S et al (2021) Detection of extracellular superoxide in isolated human immune cells and in an animal model of arterial hypertension using hydropropidine probe and HPLC analysis. Free Radic Biol Med 168:214–225. https://doi.org/10.1016/j. freeradbiomed.2021.03.041 81. Michalski R, Michalowski B, Sikora A, Zielonka J, Kalyanaraman B (2014) On the use of fluorescence lifetime imaging and dihydroethidium to detect superoxide in intact animals and ex vivo tissues: a reassessment. Free Radic Biol Med 67:278– 284. https://doi.org/10.1016/j.freeradbiomed.2013.10.816 82. Zielonka J, Hardy M, Kalyanaraman B (2009) HPLC study of oxidation products of hydroethidine in chemical and biological systems: ramifications in superoxide measurements. Free Radic Biol Med 46:329–338. https://doi.org/10.1016/j.freeradbiomed. 2008.10.031 83. Rios N, Radi R, Kalyanaraman B, Zielonka J (2020) Tracking isotopically labeled oxidants using boronate-based redox probes. J Biol Chem 295:6665–6676. https://doi.org/10.1074/jbc.RA120. 013402 84. Zielonka J, Zielonka M, Kalyanaraman B (2019) HPLC-based monitoring of oxidation of Hydroethidine for the detection of NADPH oxidasederived superoxide radical anion. Methods Mol Biol 1982:243–258. https://doi.org/10.1007/978-1-4939-9424-3_ 14 85. Zielonka J, Zielonka M, Cheng G, Hardy M, Kalyanaraman B (2019) High-throughput screening of NOX inhibitors. Methods Mol Biol 1982:429–446. https://doi.org/10.1007/978-1-49399424-3_25 86. Zielonka J et al (2012) Global profiling of reactive oxygen and nitrogen species in biological systems: high-throughput real-time analyses. J Biol Chem 287:2984–2995. https://doi.org/10.1074/ jbc.M111.309062 87. Zielonka J, Hardy M, Kalyanaraman B (2021) Chapter 13 spin trapping. In: Nitroxides: synthesis, properties and applications. The Royal Society of Chemistry, London, pp 482–518. https://doi.org/ 10.1039/9781788019651 88. Rosen GM, Britigan BE, Halpern HJ, Pou S (1999) Free radicals: biology and detection by spin trapping. Oxford University Press 89. Ouari O, Hardy M, Karoui H, Tordo P (2011) Recent developments and applications of the coupled EPR/spin trapping technique (EPR/ST). Electron Paramagn Reson 22:1–40. https:// doi.org/10.1039/9781849730877

J. Zielonka and M. Juric 90. Hardy M et al (2009) Improving the trapping of superoxide radical with a beta-cyclodextrin- 5-diethoxyphosphoryl-5-methyl-1pyrroline-N-oxide (DEPMPO) conjugate. Chemistry (Easton) 15: 11114–11118. https://doi.org/10.1002/chem.200901342 91. Vásquez-Vivar J et al (1998) Superoxide generation by endothelial nitric oxide synthase: the influence of cofactors. Proc Natl Acad Sci U S A 95:9220–9225. https://doi.org/10.1073/pnas.95.16.9220 92. Vasquez-Vivar J, Martasek P, Hogg N, Karoui H, Masters BS, Pritchard KA Jr, Kalyanaraman B (1999) Electron spin resonance spin-trapping detection of superoxide generated by neuronal nitric oxide synthase. Methods Enzymol 301:169–177. https://doi.org/ 10.1016/s0076-6879(99)01080-0 93. Isogai Y, Shiro Y, Nasuda-Kouyama A, Iizuka T (1991) Superoxide production by cytochrome b558 purified from neutrophils in a reconstituted system with an exogenous reductase. J Biol Chem 266:13481–13484. https://doi.org/10.1016/S0021-9258(18) 92720-1 94. Abbas K, Babic N, Peyrot F (2016) Use of spin traps to detect superoxide production in living cells by electron paramagnetic resonance (EPR) spectroscopy. Methods 109:31–43. https://doi. org/10.1016/j.ymeth.2016.05.001 95. Abbas K, Hardy M, Poulhes F, Karoui H, Tordo P, Ouari O, Peyrot F (2015) Medium-throughput ESR detection of superoxide production in undetached adherent cells using cyclic nitrone spin traps. Free Radic Res 49:1122–1128. https://doi.org/10.3109/10715762. 2015.1045504 96. Besson E et al (2019) Embedding cyclic nitrone in mesoporous silica particles for EPR spin trapping of superoxide and other radicals. Analyst 144:4194–4203. https://doi.org/10.1039/ c9an00468h 97. Goldstein S, Samuni A, Hideg K, Merenyi G (2006) Structureactivity relationship of cyclic nitroxides as SOD mimics and scavengers of nitrogen dioxide and carbonate radicals. J Phys Chem A 110:3679–3685. https://doi.org/10.1021/jp056869r 98. Krishna MC, Russo A, Mitchell JB, Goldstein S, Dafni H, Samuni A (1996) Do nitroxide antioxidants act as scavengers of O2-. Or as SOD mimics? J Biol Chem 271:26026–26031. https://doi.org/10. 1074/jbc.271.42.26026 99. Kundu K, Knight SF, Willett N, Lee S, Taylor WR, Murthy N (2009) Hydrocyanines: a class of fluorescent sensors that can image reactive oxygen species in cell culture, tissue, and in vivo. Angew Chem Int Ed Eng 48:299–303. https://doi.org/10.1002/ anie.200804851 100. Sadlowski CM, Maity S, Kundu K, Murthy N (2017) Hydrocyanines: a versatile family of probes for imaging radical oxidants in vitro and in vivo. Mol Sys Des Eng 2:191–200. https:// doi.org/10.1039/C7ME00014F 101. Kato M, Marumo M, Nakayama J, Matsumoto M, YabeNishimura C, Kamata T (2016) The ROS-generating oxidase Nox1 is required for epithelial restitution following colitis. Exp Anim 65:197–205. https://doi.org/10.1538/expanim.15-0127 102. Saeedi BJ, Chandrasekharan B, Neish AS (2019) Hydro-Cy3mediated detection of reactive oxygen species in vitro and in vivo. Methods Mol Biol 1982:329–337. https://doi.org/10. 1007/978-1-4939-9424-3_20 103. Samuni U, Samuni A, Goldstein S (2021) Cyclic Hydroxylamines as monitors of Peroxynitrite and superoxide-revisited. Antioxidants (Basel) 11(1):40. https://doi.org/10.3390/antiox11010040 104. Dikalov SI, Polienko YF, Kirilyuk I (2018) Electron paramagnetic resonance measurements of reactive oxygen species by cyclic hydroxylamine spin probes. Antioxid Redox Signal 28:1433– 1443. https://doi.org/10.1089/ars.2017.7396 105. Maeda H et al (2005) A Design of Fluorescent Probes for superoxide based on a nonredox mechanism. J Am Chem Soc 127:68–69. https://doi.org/10.1021/ja047018k

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

106. Maeda H et al (2007) Design of a practical fluorescent probe for superoxide based on protection–deprotection chemistry of fluoresceins with benzenesulfonyl protecting groups. Chemistry – A European J 13:1946–1954. https://doi.org/10.1002/chem. 200600522 107. Lu X, Chen Z, Dong X, Zhao W (2018) Water-soluble fluorescent probe with dual mitochondria/lysosome Targetability for selective superoxide detection in live cells and in zebrafish embryos. ACS Sensors 3:59–64. https://doi.org/10.1021/acssensors.7b00831 108. Hu JJ et al (2015) Fluorescent probe HKSOX-1 for imaging and detection of endogenous superoxide in live cells and in vivo. J Am Chem Soc 137:6837–6843. https://doi.org/10.1021/jacs.5b01881 109. Ji K, Shan J, Wang X, Tan X, Hou J, Liu Y, Song Y (2021) Rational design of near-infrared fluorescent probes for superoxide anion radical: enhancement of self-stability and sensitivity by selfimmolative linker. Free Radic Biol Med 167:36–44. https://doi.org/ 10.1016/j.freeradbiomed.2021.02.029 110. Xu K, Liu X, Tang B (2007) A Phosphinate-based red fluorescent probe for imaging the superoxide radical anion generated by RAW264.7 macrophages. Chembiochem 8:453–458. https://doi. org/10.1002/cbic.200600392 111. Xu K, Liu X, Tang B, Yang G, Yang Y, An L (2007) Design of a Phosphinate-Based Fluorescent Probe for superoxide detection in mouse peritoneal macrophages. Chem Eur J 13:1411–1416. https:// doi.org/10.1002/chem.200600497 112. Liu X, Tian X, Xu X, Lu J (2018) Design of a phosphinate-based bioluminescent probe for superoxide radical anion imaging in living cells. Luminescence 33:1101–1106. https://doi.org/10. 1002/bio.3515 113. Maeda H (2008) Which are you watching, an individual reactive oxygen species or Total oxidative stress? Ann N Y Acad Sci 1130: 149–156. https://doi.org/10.1196/annals.1430.012 114. Winterbourn CC (2013) Chapter one - the biological chemistry of hydrogen peroxide. In: Cadenas E, Packer L (eds) Methods in enzymology, vol 528. Academic, pp 3–25. https://doi.org/10. 1016/B978-0-12-405881-1.00001-X 115. Nelson DP, Kiesow LA (1972) Enthalpy of decomposition of hydrogen peroxide by catalase at 25° C (with molar extinction coefficients of H2O2 solutions in the UV). Anal Biochem 49: 474–478. https://doi.org/10.1016/0003-2697(72)90451-4 116. Noble RW, Gibson QH (1970) The reaction of ferrous horseradish peroxidase with hydrogen peroxide. J Biol Chem 245:2409–2413. https://doi.org/10.1016/S0021-9258(18)63167-9 117. Forman HJ (2007) Use and abuse of exogenous H2O2 in studies of signal transduction. Free Radic Biol Med 42:926–932. https://doi. org/10.1016/j.freeradbiomed.2007.01.011 118. Koppenol WH, Vissers MC, Hampton MB, Kettle AJ (2017) Hydrogen peroxide, a molecule with a Janus face: its history, chemistry, and biology. In: Hydrogen peroxide metabolism in health and disease. CRC Press, pp 3–16. https://doi.org/10.1201/ 9781315154831 119. Zielonka J, Marcinek A, Adamus J, Gȩbicki J (2003) Direct observation of NADH radical cation generated in reactions with one-electron oxidants. J Phys Chem A 107:9860–9864. https:// doi.org/10.1021/jp035803y 120. Gȩbicki J, Marcinek A, Zielonka J (2004) Transient species in the stepwise interconversion of NADH and NAD +. Acc Chem Res 37: 379–386. https://doi.org/10.1021/ar030171j 121. LeBel CP, Ischiropoulos H, Bondy SC (1992) Evaluation of the probe 2′,7′-dichlorofluorescin as an indicator of reactive oxygen species formation and oxidative stress. Chem Res Toxicol 5:227– 231. https://doi.org/10.1021/tx00026a012 122. Keston AS, Brandt R (1965) The fluorometric analysis of ultramicro quantities of hydrogen peroxide. Anal Biochem 11:1–5. https:// doi.org/10.1016/0003-2697(65)90034-5

339

123. Royall JA, Ischiropoulos H (1993) Evaluation of 2′,7′-Dichlorofluorescin and Dihydrorhodamine 123 as fluorescent probes for intracellular H2O2 in cultured endothelial cells. Arch Biochem Biophys 302:348–355. https://doi.org/10.1006/abbi. 1993.1222 124. Rota C, Chignell CF, Mason RP (1999) Evidence for free radical formation during the oxidation of 2′-7′-dichlorofluorescin to the fluorescent dye 2′-7′-dichlorofluorescein by horseradish peroxidase: Possible implications for oxidative stress measurements. Free Radic Biol Med 27:873–881. https://doi.org/10.1016/s08915849(99)00137-9 125. Folkes LK, Patel KB, Wardman P, Wrona M (2009) Kinetics of reaction of nitrogen dioxide with dihydrorhodamine and the reaction of the dihydrorhodamine radical with oxygen: implications for quantifying peroxynitrite formation in cells. Arch Biochem Biophys 484:122–126. https://doi.org/10.1016/j.abb.2008.10.014 126. Wrona M, Wardman P (2006) Properties of the radical intermediate obtained on oxidation of 2′,7′-dichlorodihydrofluorescein, a probe for oxidative stress. Free Radic Biol Med 41:657–667. https://doi. org/10.1016/j.freeradbiomed.2006.05.006 127. Marchesi E, Rota C, Fann YC, Chignell CF, Mason RP (1999) Photoreduction of the fluorescent dye 2′-7′-dichlorofluorescein: a spin trapping and direct electron spin resonance study with implications for oxidative stress measurements. Free Radic Biol Med 26:148–161. https://doi.org/10.1016/s0891-5849(98)00174-9 128. Chignell CF, Sik RH (2003) A photochemical study of cells loaded with 2′,7′-dichlorofluorescin: implications for the detection of reactive oxygen species generated during UVA irradiation. Free Radic Biol Med 34:1029–1034. https://doi.org/10.1016/s08915849(03)00022-4 129. Li Y et al (2015) Thioxo-dihydroquinazolin-one compounds as novel inhibitors of myeloperoxidase. ACS Med Chem Lett 6: 1047–1052. https://doi.org/10.1021/acsmedchemlett.5b00287 130. Smith Susan ME et al (2012) Ebselen and congeners inhibit NADPH oxidase 2-dependent superoxide generation by interrupting the binding of regulatory subunits. Chem Biol 19: 752–763. https://doi.org/10.1016/j.chembiol.2012.04.015 131. Dao VT-V et al (2020) Isoform-selective NADPH oxidase inhibitor panel for pharmacological target validation. Free Radic Biol Med 148:60–69. https://doi.org/10.1016/j.freeradbiomed.2019. 12.038 132. Lissi E, Pascual C, Del Castillo MD (1994) On the use of the quenching of luminol luminescence to evaluate sod activity. Free Radic Biol Med 16:833–837. https://doi.org/10.1016/0891-5849 (94)90200-3 133. Mohanty JG, Jaffe JS, Schulman ES, Raible DG (1997) A highly sensitive fluorescent micro-assay of H2O2 release from activated human leukocytes using a dihydroxyphenoxazine derivative. J Immunol Methods 202:133–141. https://doi.org/10.1016/s00221759(96)00244-x 134. Zhou M, Diwu Z, Panchuk-Voloshina N, Haugland RP (1997) A stable nonfluorescent derivative of Resorufin for the Fluorometric determination of trace hydrogen peroxide: applications in detecting the activity of phagocyte NADPH oxidase and other oxidases. Anal Biochem 253:162–168. https://doi.org/10.1006/abio.1997.2391 135. Debski D et al (2016) Mechanism of oxidative conversion of Amplex(R) red to resorufin: pulse radiolysis and enzymatic studies. Free Radic Biol Med 95:323–332. https://doi.org/10.1016/j. freeradbiomed.2016.03.027 136. Gorris HH, Walt DR (2009) Mechanistic aspects of horseradish peroxidase elucidated through single-molecule studies. J Am Chem Soc 131:6277–6282. https://doi.org/10.1021/ja9008858 137. Borbely G et al (2010) Small-molecule inhibitors of NADPH oxidase 4. J Med Chem 53:6758–6762. https://doi.org/10.1021/ jm1004368

340 138. Hirano K et al (2015) Discovery of GSK2795039, a novel small molecule NADPH oxidase 2 inhibitor. Antioxid Redox Signal 23: 358–374. https://doi.org/10.1089/ars.2014.6202 139. Summers FA, Zhao B, Ganini D, Mason RP (2013) Photooxidation of Amplex red to resorufin: implications of exposing the Amplex red assay to light. Methods Enzymol 526:1–17. https://doi.org/10. 1016/B978-0-12-405883-5.00001-6 140. Zhao B, Ranguelova K, Jiang J, Mason RP (2011) Studies on the photosensitized reduction of resorufin and implications for the detection of oxidative stress with Amplex red. Free Radic Biol Med 51:153–159. https://doi.org/10.1016/j.freeradbiomed.2011. 03.016 141. Zhao B, Summers FA, Mason RP (2012) Photooxidation of Amplex red to resorufin: implications of exposing the Amplex red assay to light. Free Radic Biol Med 53:1080–1087. https:// doi.org/10.1016/j.freeradbiomed.2012.06.034 142. Balvers WG, Boersma MG, Vervoort J, Rietjens IM (1992) Experimental and theoretical study on the redox cycling of resorufin by solubilized and membrane-bound NADPH-cytochrome reductase. Chem Res Toxicol 5:268–273. https://doi.org/10.1021/ tx00026a019 143. Dutton DR, Reed GA, Parkinson A (1989) Redox cycling of resorufin catalyzed by rat liver microsomal NADPH-cytochrome P450 reductase. Arch Biochem Biophys 268:605–616. https://doi. org/10.1016/0003-9861(89)90328-7 144. Maeda H, Matsu-ura S, Senba T, Yamasaki S, Takai H, Yamauchi Y, Ohmori H (2000) Resorufin as an electron acceptor in glucose oxidasecatalyzed oxidation of glucose. Chem Pharm Bull (Tokyo) 48:897–902. https://doi.org/10.1248/cpb.48.897 145. Votyakova TV, Reynolds IJ (2004) Detection of hydrogen peroxide with Amplex red: interference by NADH and reduced glutathione autooxidation. Arch Biochem Biophys 431:138–144. https:// doi.org/10.1016/j.abb.2004.07.025 146. Serrano J, Jové M, Boada J, Bellmunt MJ, Pamplona R, PorteroOtín M (2009) Dietary antioxidants interfere with Amplex red-coupledfluorescence assays. Biochem Biophys Res Commun 388:443–449. https://doi.org/10.1016/j.bbrc.2009.08.041 147. Reszka KJ, Wagner BA, Burns CP, Britigan BE (2005) Effects of peroxidase substrates on the Amplex red/peroxidase assay: antioxidant properties of anthracyclines. Anal Biochem 342:327–337. https://doi.org/10.1016/j.ab.2005.04.017 148. Donkó Á et al (2009) Detection of hydrogen peroxide by lactoperoxidase-mediated dityrosine formation. Free Radic Res 43:440–445. https://doi.org/10.1080/10715760902859069 149. Panus PC, Radi R, Chumley PH, Lillard RH, Freeman BA (1993) Detection of H2O2 release from vascular endothelial cells. Free Radic Biol Med 14:217–223. https://doi.org/10.1016/0891-5849 (93)90013-k 150. Bylund J, Björnsdottir H, Sundqvist M, Karlsson A, Dahlgren C (2014) Measurement of respiratory burst products, released or retained, during activation of professional phagocytes. In: Quinn MT, DL FR (eds) Neutrophil methods and protocols. Humana Press, Totowa, NJ, pp 321–338. https://doi.org/10.1007/978-162703-845-4_21 151. Ruch W, Cooper PH, Baggiolini M (1983) Assay of H2O2 production by macrophages and neutrophils with homovanillic acid and horse-radish peroxidase. J Immunol Methods 63:347–357. https://doi.org/10.1016/s0022-1759(83)80008-8 152. Werner E (2003) Determination of cellular H O production. Sci STKE 2003:PL3. https://doi.org/10.1126/stke.2003.168.pl3 153. Boveris A, Martino E, Stoppani AO (1977) Evaluation of the horseradish peroxidase-scopoletin method for the measurement of hydrogen peroxide formation in biological systems. Anal Biochem 80:145–158. https://doi.org/10.1016/0003-2697(77)90634-0

J. Zielonka and M. Juric 154. Corbett JT (1989) The scopoletin assay for hydrogen peroxide. A review and a better method. J Biochem Biophys Methods 18:297– 307. https://doi.org/10.1016/0165-022x(89)90039-0 155. Pick E, Keisari Y (1980) A simple colorimetric method for the measurement of hydrogen peroxide produced by cells in culture. J Immunol Methods 38:161–170. https://doi.org/10.1016/00221759(80)90340-3 156. Pick E, Mizel D (1981) Rapid microassays for the measurement of superoxide and hydrogen peroxide production by macrophages in culture using an automatic enzyme immunoassay reader. J Immunol Methods 46:211–226. https://doi.org/10.1016/00221759(81)90138-1 157. Pick E (1986) Microassays for superoxide and hydrogen peroxide production and nitroblue tetrazolium reduction using an enzyme immunoassay microplate reader. Methods Enzymol 132:407–421. https://doi.org/10.1016/s0076-6879(86)32026-3 158. Pacquelet S, Lehmann M, Luxen S, Regazzoni K, Frausto M, Noack D, Knaus UG (2008) Inhibitory action of NoxA1 on dual oxidase activity in airway cells*. J Biol Chem 283:24649–24658. https://doi.org/10.1074/jbc.M709108200 159. Martyn KD, Frederick LM, von Loehneysen K, Dinauer MC, Knaus UG (2006) Functional analysis of Nox4 reveals unique characteristics compared to other NADPH oxidases. Cell Signal 18:69–82. https://doi.org/10.1016/j.cellsig.2005.03.023 160. Arnold RS et al (2001) Hydrogen peroxide mediates the cell growth and transformation caused by the mitogenic oxidase Nox1. Proc Natl Acad Sci 98:5550–5555. https://doi.org/10.1073/ pnas.101505898 161. Morand S, Ueyama T, Tsujibe S, Saito N, Korzeniowska A, Leto TL (2009) Duox maturation factors form cell surface complexes with Duox affecting the specificity of reactive oxygen species generation. FASEB J 23:1205–1218. https://doi.org/10.1096/fj. 08-120006 162. Lippert AR, Van de Bittner GC, Chang CJ (2011) Boronate oxidation as a bioorthogonal reaction approach for studying the chemistry of hydrogen peroxide in living systems. Acc Chem Res 44:793– 804. https://doi.org/10.1021/ar200126t 163. Zielonka J, Sikora A, Hardy M, Joseph J, Dranka BP, Kalyanaraman B (2012) Boronate probes as diagnostic tools for real time monitoring of peroxynitrite and hydroperoxides. Chem Res Toxicol 25:1793–1799. https://doi.org/10.1021/tx300164j 164. Sikora A et al (2020) Boronate-based probes for biological oxidants: a novel class of molecular tools for redox biology. Front Chem 8:580899. https://doi.org/10.3389/fchem.2020. 580899 165. Dickinson BC, Peltier J, Stone D, Schaffer DV, Chang CJ (2011) Nox2 redox signaling maintains essential cell populations in the brain. Nat Chem Biol 7:106–112. https://doi.org/10.1038/ nchembio.497 166. Woolley JF et al (2012) H2O2 production downstream of FLT3 is mediated by p22phox in the endoplasmic reticulum and is required for STAT5 signalling. PLoS One 7:e34050. https://doi.org/10. 1371/journal.pone.0034050 167. Brewer TF, Garcia FJ, Onak CS, Carroll KS, Chang CJ (2015) Chemical approaches to discovery and study of sources and targets of hydrogen peroxide redox signaling through NADPH oxidase proteins. Annu Rev Biochem 84:765–790. https://doi.org/10.1146/ annurev-biochem-060614-034018 168. Reis J et al (2020) A closer look into NADPH oxidase inhibitors: validation and insight into their mechanism of action. Redox Biol 32:101466. https://doi.org/10.1016/j.redox.2020.101466 169. Sikora A, Zielonka J, Lopez M, Joseph J, Kalyanaraman B (2009) Direct oxidation of boronates by peroxynitrite: mechanism and implications in fluorescence imaging of peroxynitrite. Free Radic Biol Med 47:1401–1407. https://doi.org/10.1016/j.freeradbiomed. 2009.08.006

20

Methods to Measure Reactive Oxygen Species Production by NADPH Oxidases

170. Zielonka J, Sikora A, Joseph J, Kalyanaraman B (2010) Peroxynitrite is the major species formed from different flux ratios of co-generated nitric oxide and superoxide: direct reaction with boronate-based fluorescent probe. J Biol Chem 285:14210–14216. https://doi.org/10.1074/jbc.M110.110080 171. Michalski R, Zielonka J, Gapys E, Marcinek A, Joseph J, Kalyanaraman B (2014) Real-time measurements of amino acid and protein hydroperoxides using coumarin boronic acid. J Biol Chem 289:22536–22553. https://doi.org/10.1074/jbc.M114. 553727 172. Truzzi DR, Augusto O (2017) Influence of CO2 on Hydroperoxide metabolism. In: Vissers MC, Hampton M, Kettle AJ (eds) Hydrogen peroxide metabolism in health and disease. CRC Press, Boca Raton, FL, pp 81–99. https://doi.org/10.1201/9781315154831 173. Grzelakowska A et al (2022) Water-soluble cationic boronate probe based on coumarin imidazolium scaffold: synthesis, characterization, and application to cellular peroxynitrite detection. Free Radic Biol Med 179:34–46. https://doi.org/10.1016/j. freeradbiomed.2021.12.260 174. Zielonka J, Sikora A, Podsiadly R, Hardy M, Kalyanaraman B (2021) Identification of Peroxynitrite by profiling oxidation and nitration products from mitochondria-targeted Arylboronic acid. Methods Mol Biol 2275:315–327. https://doi.org/10.1007/978-10716-1262-0_20 175. Smulik R et al (2014) Nitroxyl (HNO) reacts with molecular oxygen and forms peroxynitrite at physiological pH. Biological Implications. J Biol Chem 289:35570–35581. https://doi.org/10. 1074/jbc.M114.597740 176. Grzelakowska A et al (2021) Two-photon fluorescent probe for cellular peroxynitrite: fluorescence detection, imaging, and identification of peroxynitrite-specific products. Free Radic Biol Med 169:24–35. https://doi.org/10.1016/j.freeradbiomed.2021.04.011 177. Lippert AR, Keshari KR, Kurhanewicz J, Chang CJ (2011) A hydrogen peroxide-responsive hyperpolarized 13C MRI contrast agent. J Am Chem Soc 133:3776–3779. https://doi.org/10.1021/ ja111589a 178. Xie X et al (2016) Rational design of an α-Ketoamide-based nearinfrared fluorescent probe specific for hydrogen peroxide in living systems. Anal Chem 88:8019–8025. https://doi.org/10.1021/acs. analchem.6b01256 179. Abo M, Urano Y, Hanaoka K, Terai T, Komatsu T, Nagano T (2011) Development of a highly sensitive fluorescence probe for

341

hydrogen peroxide. J Am Chem Soc 133:10629–10637. https://doi. org/10.1021/ja203521e 180. Zhang K-M et al (2015) A coumarin-based two-photon probe for hydrogen peroxide. Biosens Bioelectron 64:542–546. https://doi. org/10.1016/j.bios.2014.09.073 181. Xu K, Liu F, Wang H, Wang S, Wang L, Tang B (2009) Sulfonatebased fluorescent probes for imaging hydrogen peroxide in living cells. Sci China Ser B Chem 52:734–740. https://doi.org/10.1007/ s11426-009-0109-9 182. Xu K, Tang B, Huang H, Yang G, Chen Z, Li P, An L (2005) Strong red fluorescent probes suitable for detecting hydrogen peroxide generated by mice peritoneal macrophages. Chem Commun:5974–5976. https://doi.org/10.1039/B512440A 183. Maeda H et al (2004) Fluorescent probes for hydrogen peroxide based on a non-oxidative mechanism. Angew Chem Int Ed 43: 2389–2391. https://doi.org/10.1002/anie.200452381 184. Quimbar ME, Davis SQ, Al-Farra ST, Hayes A, Jovic V, Masuda M, Lippert AR (2020) Chemiluminescent measurement of hydrogen peroxide in the exhaled breath condensate of healthy and asthmatic adults. Anal Chem 92:14594–14600. https://doi.org/ 10.1021/acs.analchem.0c02929 185. Guo H, Aleyasin H, Dickinson BC, Haskew-Layton RE, Ratan RR (2014) Recent advances in hydrogen peroxide imaging for biological applications. Cell Biosci 4:64. https://doi.org/10.1186/ 2045-3701-4-64 186. Lee D et al (2007) In vivo imaging of hydrogen peroxide with chemiluminescent nanoparticles. Nat Mater 6:765–769. https://doi. org/10.1038/nmat1983 187. Yang J et al (2017) Oxalate-curcumin-based probe for micro- and macroimaging of reactive oxygen species in Alzheimer’s disease. Proc Natl Acad Sci 114:12384–12389. https://doi.org/10.1073/ pnas.1706248114 188. Lukyanov KA, Belousov VV (2014) Genetically encoded fluorescent redox sensors. Biochim Biophys Acta 1840:745–756. https:// doi.org/10.1016/j.bbagen.2013.05.030 189. Roma LP, Deponte M, Riemer J, Morgan B (2018) Mechanisms and applications of redox-sensitive green fluorescent protein-based hydrogen peroxide probes. Antioxid Redox Signal 29:552–568. https://doi.org/10.1089/ars.2017.7449 190. Kostyuk AI et al (2020) In vivo imaging with genetically encoded redox biosensors. Int J Mol Sci 21(21):8164. https://doi.org/10. 3390/ijms21218164

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

21

Christopher M. Dustin, Eugenia Cifuentes-Pagano, and Patrick J. Pagano

Abstract

The NADPH oxidase family of proteins are professional reactive oxygen species (ROS)-generating enzymes that generate superoxide (O2-) by transferring electrons from NADPH across cellular membranes to O2. Comprised of seven members (NOX 1-5 and DUOX 1-2), these enzymes exist at the intersection of physiological and pathological signaling. While the NOXs are well known to contribute to homeostatic, physiological signaling, they are perhaps more infamously known as drivers of disease phenotypes in multiple cells and tissues. Given this fact, significant effort has been directed toward developing inhibitors for the NOXs with varying degrees of success. This difficulty is due to the high degree of homology in their core catalytic domains, making isoform selectivity difficult. While selective inhibitors do exist, they are unfortunately often plagued by a lack of specificity or potency, hampering their therapeutic potential. Therefore, current trends in NOX inhibitor research are aimed at generating NOX inhibitors that are both potent and isoform-specific. In this chapter, we will detail the current state of NOX inhibitors and their critical features, while also concluding with the outlook of NOX inhibitor development moving forward. Keyword

NADPH oxidase · NADPH oxidoreductase · NOX · Oxidase · Small molecule · Peptide · Peptidic · Drugs · Translational · Clinical · Reactive oxygen · Reactive nitrogen · Selective inhibitors

C. M. Dustin · E. Cifuentes-Pagano · P. J. Pagano (✉) Department of Pharmacology and Chemical Biology, Vascular Medicine Institute, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA e-mail: [email protected]

1

Introduction

The NADPH Oxidases (NOXs) are a family of multicomponent membrane-bound oxidases that have the unique function of specifically generating reactive oxygen species (ROS) [1– 3]). As “professional” ROS-generating enzymes, NOXs are critical for many cellular processes and are known to be involved in various diseases featuring abnormalities in their expression and/or activation. Recent years have seen a greater focus on the role of ROS imbalances, i.e. ROS generation vs. scavenging, in disease, with efforts initially being made to obliterate ROS across the board. However, much of the challenge is whether or not this is practicable or desired, in addition to the appreciation that ROS are bona fide signaling mediators that subserve vital organismal functions [4]. Indeed, attempts at clinical trials in the 1990s without the expert guidance of redox biologists and chemists ill-advisedly centered on antioxidant vitamins whose oxidant scavenging properties and effectiveness were equivocal at best. It came as no surprise, therefore, to many basic scientists that those regimens yielded little to no clinical benefit [5]. That is, among the multiple explanations given for the disillusioning failure of those trials, was that any anticipated, salient benefits of those interventions were offset by an inadvertent blockade of fundamental signaling roles of ROS [6] and by extension NOXs whose attendant ROS generation was very likely indiscriminately inhibited by antioxidant administration [7]. To date, there is still no conclusive evidence that generalized antioxidant therapy is the answer to treating disease. Therefore, at the vanguard of inaccurately named “antioxidant” therapies in recent years have been technologies aimed at addressing either the specific sources of ROS within cells or blocking/modulating the ability of those sources to act on downstream signaling effectors. With respect to the NOXs, the main focus has been on the former, as the NOXs expressly produce ROS. However, while there has been much effort toward developing early-stage candidate inhibitors for the NOXs, there has been

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_21

343

344

C. M. Dustin et al.

relatively little advancement with respect to therapeutics that have made it out of exploratory/preclinical stages, with only two FDA-approved molecules in trials as NOX inhibitors (Setanaxib and APX-115, see respective sections for corresponding trial codes). Additionally, the search for isoform-selective inhibitors has been formidable and is still ongoing. To date, there have been no clinical trials to our knowledge centered on targeting one NOX isoform with an isoform-specific inhibitor. In this chapter, we will discuss the major challenges of developing isoform-selective NOX inhibitors, as well as the overall landscape of NOX inhibitors to date. Finally, we will discuss the potential utility, feasibility, and shortcomings of targeting specific NOX isoforms, and what this may mean for the future of NOX research and drug development. Herein, we provide a brief overview of the NOXs as is pertinent to the finer points of NOX inhibitor design, while a more in-depth overview of NOX structure/function is provided elsewhere in this book.

2

The NADPH Oxidases: A Brief Overview as a Prelude to Therapeutic Modalities

As is more thoroughly described elsewhere in this volume (in particular, Chap. 9 by A. van der Vliet), the NOX family consists of seven members, NOX 1-5 and dual oxidases (DUOX) 1-2, all of which specifically generate superoxide (O2-) and/or hydrogen peroxide (H2O2) [8], the latter being the result of either superoxide dismutasemediated conversion of O2- to H2O2, spontaneous dismutation, or a posited direct production of H2O2 from O2 by select NOXs [9, 10]. NOXs are predominantly localized to plasma membranes and vectorially produce O2- and thus H2O2 to the outside of the cell [1], whose fate, in turn, is largely determined by aquaporin-mediated transport or diffusion across membranes [11–13]. In the case of the prototype NOX2 in phagocytes, the fate of ROS is the phagosome by virtue of plasmalemmal involution upon microbial engulfment. However, other NOX isoforms can be localized to various subcellular organelles including the mitochondrion [14], endoplasmic reticulum [15], endosome/redoxosome [16, 17] and nucleus [18]. All NOX isoforms share a similar core structure in which there is a centralized double hemecontaining transmembrane subunit, for which the individual oxidase is named, responsible for catalyzing the transfer of electrons from NADPH to extracellular O2 [1]. This is facilitated by a C-terminal cytosolic dehydrogenase domain containing NADPH and FAD binding sites. In addition to the core catalytic hemoprotein, there are multiple auxiliary and cytosolic subunits that are differentially required for stability (p22phox) and activation, depending on the specific NOX in question, with some requiring multiple cytosolic subunits for

proper organization and function while others requiring relatively few or none (Fig. 21.1) (extensively reviewed in [8]). NOXs were initially presumed to be expressly involved in ROS production solely for host-defense purposes as with NOX2 oxidase (vide infra, see also Chap. 2 by J.T. Curnutte and A.I. Tauber) [19]. However, concurrent with the discovery of multiple homologues and their wide tissue distribution, as well as the evolving landscape of our knowledge pertaining to the multifaceted roles of ROS within cellular systems, a broader picture of NOX signaling began to take shape. The first discovered and therefore prototypical NOX (later named NOX2) was found in neutrophils [20]. Given its role in ROS production during phagocyte responses, this led to the initial classification of NOX2 as a phagocytosisassociated host defense enzyme [21–23]. The full NOX2 complex requires multiple subunits for its proper organization and function alongside the main catalytic subunit. The first and most critical subunit is p22phox (also known by cytochrome b-558 alpha chain or its gene name CYBA), a transmembrane subunit responsible for binding and stabilizing NOX2, as well as binding the cytosolic organizer subunits in the NOX complex. NOX2 also requires an organizer subunit p47phox (also named neutrophil cytosolic factor 1, NCF1). p47phox exists in a latent, autoinhibited conformation unless and until it is phosphorylated by various protein kinases, of which protein kinase C (PKC) appears foremost. Phosphorylation causes a conformational change that disengages an autoinhibitory domain [24]. This “open” conformation exposes internal tandem SH3 domains (supergroove) that allow direct binding of p47phox to a polyproline motif on p22phox and the concomitant translocation of subunits that associate with p47phox [24]. One such subunit is p67phox (also termed NCF2), commonly known as the activating subunit. p67phox binds to p47phox and is simultaneously translocated to the membrane following p47phox phosphorylation, where its subsequent interaction with NOX2 is essential for activation [25]. p40phox (NCF4) is similarly translocated alongside p47phox and p67phox [1] though its role as a positive or negative regulator remains controversial. Finally, NOX2 requires the recruitment and activation of the small GTPases Rac1 and/or Rac2, which bind p67phox[1, 25]. Once these integral components are assembled and in place, NOX2 is active and capable of producing O2- at capacity. Later discoveries would reveal that NOX2 was not solely expressed in phagocytic cells, and that there were multiple other NOX family members distributed in varied combinations and quantity across other tissues [26]. These different isoforms (so referred eponymously by their core hemoprotein) were revealed to have a differential requirement for cytosolic subunits. The NOX1 isoform was the first NOX2 homolog described and is the most similar in

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

NOX2

NOX1

p22

345

NOX4

p22

p22

Rac1/2

Rac1/2

P67 p

P47

hox

phox

NoxO1

NoxA1

Poldip2 D FA

D FA

D FA

H

NA

DP

NA

H DP

NA

DUOX1/2

NOX5 Calmodulin

Ca 2+

Tks4/5

H DP

DUOXA1/A2

Ca2+

Ca2+

Ca 2+

Ca 2+ Ca 2+

HSP90 H DP NA

D FA

D FA

H DP

NA

Fig. 21.1 Organization and structure of the mammalian NADPH Oxidases. The seven mammalian NOXs share a similar, dual hemecontaining transmembrane catalytic domain. This subunit catalyzes the transfer of electrons from NADPH via FAD, bound by the cytosolic dehydrogenase tail, across membranes to the terminal acceptor molecular O2, generating O2- that is rapidly dismutated to H2O2. The canonical NOX, NOX2, is comprised of multiple subunits required for activation. These include the transmembrane stabilizing subunit p22phox and cytosolic p47phox (“organizer” subunit), p67phox (“activator” subunit) and the small GTPase Rac1/2. NOX2 may bind an additional cytosolic subunit p40phox (not shown) that is not required for activation. The canonical NOX1 complex is similar in organization to NOX2, however utilizing distinct organizer and activator proteins (NOXO1 and NOXA1,

respectively). NOX4 only requires p22phox though it has been shown to bind multiple cytosolic proteins, such as Poldip2 and TKS4/5, that modulate activity. NOX5 is unique in its organization as it requires no subunits, aside from the core catalytic domain, whereas it contains EF-hand domains on its cytosolic N-terminus rendering it Ca2+-sensitive. NOX5 is also known to bind modulatory cytosolic proteins, i.e. calmodulin and HSP90. Finally, DUOX1 and DUOX2 are likely the most distinct of the family members. These proteins, while possessing EF-hand domains, require no canonical subunits for activation; however, they contain an extracellular peroxidase homology domain and a transmembrane maturation factor (DUOXA1/DUOXA2) that are required for proper membrane translocation and enzymatic activity. Created with BioRender.com

sequence and composition to NOX2 [27], in that the roles its subunits subserve are remarkably similar, save for a key difference. While NOX1 requires p22phox and Rac1, as well as organizer and activator subunits like NOX2, the canonical NOX1 engages two distinct proteins that fulfill the organizer and activator roles. That is, NADPH oxidase activator 1 (NOXA1) supplants p67phox of NOX2 for the activator role and NADPH oxidase organizer 1 (NOXO1) stands in for p47phox of NOX2 for the organizer function in NOX1 [28]. Interestingly, while these two proteins fulfill similar roles to p67phox and p47phox and possess high structural and functional similarities with their NOX2-associated counterparts, they differ in critical ways [1]. Importantly, NOXO1 does not possess an autoinhibitory loop, conceivably explaining the widely observed constitutive activity of the NOX1 complex and a greater dependence on transcription and translation [29]. On the other hand, despite a low sequence homology between p47phox and NOXO1 that might

predict their individual fidelity to canonical NOXs 2 & 1, respectively, there is also evidence that NOX1 can exist in hybrid form (in which its organizing subunit is p47phox instead of NOXO1), further complicating its regulation. Interestingly, only the combination of NOXO1 and NOXA1 produced constitutively active NOX1, whereas the scenario in which either NOXO1/p67phox or NOXA1/p47phox are employed as organizer and activator, no constitutive activity is observed [29]. NOX3, the least-widely distributed of the NOX enzymes, similarly has a requirement for p22phox, though its cytosolic subunit requirements are less clear. It has been reported that NOX3 is capable of utilizing NOXO1, NOXA1, p47phox, and/or p67phox, with other evidence that NOX3 solely requires NOXO1 for activation [1, 30–33]. In contrast to the other members of the NOX family thus far discussed, NOX4 ostensibly only requires p22phox [34]. Given this ostensible unrequisite need for cytosolic subunits, it is generally thought that NOX4 is constitutively

346

C. M. Dustin et al.

active and, therefore, transcriptionally regulated [34]. That notwithstanding, NOX4 can be regulated by direct phosphorylation [35]. Intriguingly, and distinct to most other NOX enzymes, NOX4 is purported to expressly generate H2O2 [9, 10] instead of O2-, making it somewhat unique among the NOX family. Finally, NOX5 does not ostensibly require cytosolic subunits or p22phox and its structure is distinctly different from the other NOX enzymes [36]. Specifically, NOX5 has a cytosolic EF-hand domain on its N-terminus that binds calcium and interacts with a region in the dehydrogenase domain, known as the regulatory EF-Hand binding domain (REFBD), as well as a calmodulin-binding region C-terminal to the REFBD, both interactions being required for activation [36]. The dual oxidases (DUOX1 and DUOX2, also called thyroid oxidases or THOX) have 2 cytosolic EF-hand domains, requiring calcium for activation and production of H2O2 [37]. In addition to the EF-hand domains, the DUOXs are characterized by an extracellular peroxidase homology domain, named for its high sequence similarity to members of the peroxidase family, with an as-of-yet undetermined function in human DUOXs. Additionally, unlike NOX5, DUOX1/2 require the expression of specific maturation factors DUOXA1/2 for proper plasma membrane localization, as well as full enzymatic activity [37].

3

Challenges of Druggable NOX inhibition

From a fundamental standpoint, NOX inhibitors facilitate our understanding and appreciation of NOXs’ roles in biology and disease and are the most translatable proof-of-concept and platform on which to design drugs. While genetic methods such as RNAi or knockout models in mice/rats help lay the crucial foundation to our understanding of the processes that are NOX-dependent, they have at least one important limitation. That is, eradication of an entire NOX subunit is expected to elicit a wide array of untoward effects by virtue of that subunit’s myriad heretofore unknown and yet-to-be theorized interactions. Thus, despite the fact that molecular and genetic approaches are quite often considered the gold standard for expanding our knowledge of the NOXs in rodent biology and disease, they are, in fact, rather useless from the vantage point of assessing a “druggable” target. That is not to say that targeted genetic modalities, (e.g. CRISPR) that can alter specific deleterious interactions, per se, do not hold great promise. Indeed, they do. However, multiple hurdles must be overcome for these newer cuttingedge strategies to be applicable. Until that time, our focus must remain on druggable small molecule antagonists, decoys, and inhibitors of the NOXs for clinical use. It is also our contention that if the field is to progress on this path, isoform-selective and potent inhibitors should remain a primary objective.

That notwithstanding, there are multiple features of the NOXs and the currently available inhibitors that make this a challenging endeavor. The first issue with assessing NOX activity and NOX inhibition is the general dearth of specific inhibitors. While the drive for NOX inhibitors has powered significant progress in recent years, lack of selectivity is a concern that is unfortunately still extant. The predominant issue is that the NOXs share a high level of homology, in addition to an overlap in subunit utilization [1]. For example, targeting the highly homologous catalytic heme sites in CYBB is futile since such a strategy would by design “pan” inhibit multiple NOXs which could extend to highly conserved sites in other hemoproteins [3, 38, 39]. Furthermore, subunit interactions that are conserved across NOX systems (e.g. p47phox: p22phox in the canonical NOX2 and hybrid NOX1 systems) could hamper attempts at achieving selectivity. In other words, the hybrid NOX1 system engages p47phox (in the place of NOXO1 [40]) via binding to p22phox, and thus may be blocked by p47phox-targeted inhibitors. Compounding this issue is the fact that there are generally multiple NOXs expressed within any given cell type and tissue, as well as some proteins with similar features. On the other hand, if it is found, for instance, that the canonical NOX2 and hybrid NOX1 systems elicit the same or similar deleterious disease outcomes, versatility may be desired. Similarly, many NOX inhibitors are highly promiscuous and display significant off-target effects. As such, while these inhibitors in general may be highly potent and efficacious in a reductionist setting, exploitation of such a strategy is irrational if numerous off-target effects are found in vivo [41]. Another concern with respect to NOX inhibitor specificity is whether the molecule in question is even an inhibitor at all. Multiple initially endorsed inhibitors have later been deemed non-specific ROS scavengers that do not necessarily block NOX activation, per se [39, 42, 43]. Such a broad approach, while touted by multiple research and industrial entities, is even more troublesome as it brings us right back to where we started with the conundrum of eradicating vital ROS. As technology develops and the search for new inhibitors continues, there are still two major deficiencies that render refinement of NOX inhibitors difficult: limited structural data and limited options for well-characterized and precise, high throughput cell-free assays. Further to those points, one of the most formidable challenges that scientists face when attempting to generate potent and specific inhibitors for the NOXs is a paucity of wide-ranging structural data. In drug development contexts, structural data are often crucial for modifying and refining lead compounds for desired effects through structure activity relationship (SAR) screening and optimization [44]. What SAR aims to achieve is an informed alteration of compound structure with the goal of improving various properties, such as potency, efficacy, safety, and solubility [44]. To date, the body of structural data available for the NOXs is scarce as a

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

consequence of the complexity of some isoforms, issues with subunit-replete enzyme purification, and the general complexity of determining the structure of transmembrane proteins [45]. In terms of active NOX isozymes, there is currently only full crystal structure data for C. stagnale NOX5 transmembrane and dehydrogenase domains (PDB: 5O0T, 5O0X) [45]. Two other groups also recently published Cryo-EM data showing the murine DUOX1 complexed with DUOXA1 (PDB: 6WXR, 6WXU, 6WXV) [46], as well as the human DUOX1-DUOXA1 complex under high and low concentrations of Ca2+ (PDB: 7D3E, 7D3F) [47]. NOX5 and DUOX1 do not require p22phox or seemingly other cytosolic subunits (aside from DUOXA1 in the case of DUOX1), making them a logical starting point for NOX structure determination [36, 37]). While no full structures exist for the remaining isoforms, a few partial structures have been published. For example, the NOX2 NADPH-binding domain has been crystallized and reposited in the Protein Data Bank (PDB: 3A1F, unpublished). Additionally, some preliminary electron microscopy evidence for the NOX2 flavocytochrome b (NOX2 and p22phox) interaction has been published as a demonstration of NOX-focused Cryo-EM methodology [48]. There are reported structures for some of the subunits as well, albeit incomplete. To date p47phox (PDB: 1NG2) [49], p40phox (PDB: 2DYB) [50], and multiple incomplete structures for p67phox (e.g. 1HH8 [51] and 1K4U [52], among others) have been published or were added to the PDB without publication. Similarly, there are multiple listed crystal structures of various interactions between essential subunits, such as the interaction between p67phox and Rac (PDB:1E96) [53], which is highly coveted structural information for inhibitor design. Additionally, there are two structures of p47phox: p22phox interaction, one revealed with X-ray crystallography (PDB: 1OV3) [49] and another with solution NMR (PDB: 1WLP) [54]. Similarly, incomplete structures of NOXA1 (PDB: 7CFZ, SH3 domain crystal structure, reposited but unpublished) and an NOXO1 (PDB: 2 l73, PX domain NMR structure, reposited but unpublished) are accessible. However, to date, there are still no full structures for the entire NOX1-4 complexes, as well as the core transmembranal subunits. This is likely due to the aforementioned difficulty in purifying an intact NOX subunit itself, in addition to the high degree of difficulty in purifying a replete and active complex with all essential subunits in place. For example, in the case of the replete NOX2 complex, the purification would require the requisite pivotal phosphorylation of p47phox, coalescence of p47phox, p40phox, p67phox, and Rac with NOX2 and p22phox association with prior to purification of this complex in its intact, active state, followed by their retention in this state during structural determination. While there are some structures available from which to work, there is still a long way to go for complete structures of both the

347

multiple individual NOX subunits, as well as the complexes as a whole, to bring a full 3-dimensional visualization of the family of enzymes into focus. While structural information is critical for enhancing our understanding of the NOXs and for developing inhibitors, there is, as mentioned, another factor that complicates the mission. Currently, our ability to quickly, efficiently, and definitively assay for authentic NOX activity in a high throughput manner is limited – a necessity for evaluating inhibitory specificity. Most testing to date has been executed in heterologous reconstituted cell lines. However, “cell-free” assays are also employed wherein cells are lysed prior to measurement and membrane fractions containing the NOX cytochromes are differentially centrifuged and sedimented, which can then be followed by supplementation with recombinant cytosolic subunits, and measurement of activity by direct NADPH addition [55]. These analyses have the potential to be high throughput and great strides have been made in utilizing these assays [56]. However, how faithfully they reproduce how activation of these enzymes occurs in vivo, and thus how effective inhibitors defined by such assays are, are still not clear. For one, consistent molar concentrations of subunits across these recapitulated systems should always be attempted. Controlling for all of these factors, the assays are only as good as the ROS-detection probes employed. Moreover, assays that prove disrupted NOX assembly in vivo, as well as ex vivo [57], are necessary.

4

NOX Targeting Strategies: Advantages and Drawbacks

When addressing the matter of selective NOX targeting, there are a variety of ways in which the problem can be approached. For example, Fig. 21.2 outlines the various interactions that can potentially be, and have been, exploited by NOX inhibitors. However, it is not as simple as targeting these interactions in many cases. Specifically, the chief concern should be the overall objective for the inhibitor molecule in question. Namely, will the inhibitor be a pan-NOX inhibitor or selective for one or more isoforms. The latter generally proves to be much more difficult as there is a high degree of homology among the NOX isoforms. Assuredly, pan-NOX inhibition does not come without its drawbacks. Some of the best pan-NOX inhibitor strategies are plagued by significant off-target effects causing a host of responses within cells [3]. On the other hand, discriminate targeting of the different NOXs proves to be a formidable challenge; and the most fruitful strategies appear to be by rational design, i.e. aimed to target select peptide sequences of the NOXs themselves with a goal of inhibition or antagonism of precise binding domains in cytosolic subunits germane to a particular NOX. Unfortunately, despite the potential and specificity of peptidic

348

Fig. 21.2 Sites of interaction within the NOX structure that are potential druggable targets. Known interaction sites within the canonical NOX complex for potential targeting. 1. Interaction between Rac1/2 and the activator subunit (p67phox/NOXA1). 2. Intermolecular interaction between the activator subunit and core NOX protein. 3. Interaction of the C-terminal dehydrogenase tail with the core NOX protein. 4. Interaction of the organizer subunit (p47phox/NOXO1) with the polybasic B-loop of the catalytic domain. 5. Stabilizing interaction of the NOX catalytic domain with p22phox. 6. Requisite interaction between the activator and organizer subunits for proper translocation and complex assembly. 7. Interaction of the SH3 domain-containing “supergroove” region of the organizer subunit with p22phox. Created with BioRender.com

inhibitors, there is still significant resistance to the notion of peptides as drugs, often derived from a broadly held bias that peptides are not stable in the gut and thus not orally bioavailable. That bias is increasingly being demystified and debunked by the advancements in formulation chemistry and nanotechnology [58, 59]. One of the more common NOX inhibition strategies is targeting aimed at the cytosolic dehydrogenase tail, as this is the main initiation site of electron transfer [60]. As described in more detail elsewhere, this tail contains both NADPHbinding, as well as the FAD-binding domains, which are critical components allowing the transfer of electrons to the transmembrane heme groups. Multiple NOX inhibitors target these domains, with particular focus on the FAD binding domain; and, as expected, they are among the most effective and commonly used inhibitors described to date [3]. However, there is a major drawback to targeting these sites, namely with respect to the potential off-target effects. Targeting the FAD-binding domains, for example, has the particular drawback of targeting other flavoproteins within the proteome including other ROS-generating enzyme systems, i.e. xanthine oxidase [61], cytochrome P450 isozymes [62] and endothelial nitric oxide synthase (eNOS) [63]. This poses a serious problem outside of ROS generation, because critical “housekeeping” proteins are also

C. M. Dustin et al.

flavoproteins, such as the thioredoxin-reducing protein thioredoxin reductase, which is critical in multiple antioxidant pathways [64]. Similarly, given that major ROS-generating enzymes such as xanthine oxidase can also be inhibited by targeting FAD, the interpretation of what appears to be NOX inhibition may be a false positive. Given the importance of these other proteins in normal cellular function and disease, this could become a major issue if not accounted for. Along the same lines, it does not allow for a specific identification of which NOX isoform is responsible for a particular ROS generation and attendant phenotype that is being observed. Despite this serious shortcoming, many groups continue to use molecules that target FAD to this day. It should therefore be asserted that studies cannot claim compounds to be selective NOX inhibitors without first demonstrating that the numerous off-target effects are not contributing to the observed inhibitory effect. Indeed, this is likely a similar case for the targeting of the NADPH-binding domain, given the broad utilization of NADPH as a substrate in antioxidant and metabolic enzymes. It is worth noting, however, that there may be scenarios in which broader NOX or even other oxidase inhibition is desirable such as in a hyper-reactive inflammatory reaction that subsumes the induction and harmful effects of multiple oxidases. The next most obvious target might be the blockade of electron transfer to molecular O2 via the transmembrane hemoprotein core. While this may be the most straightforward and logical theoretical strategy, there are fundamental considerations that deem this pursuit illogical. For example, the high degree of homology among NOXs with respect to the functionally conserved hemoprotein should raise a red flag with respect to the wider NOX family as well as other related hemoproteins [1]. Thus, targeting the heme core is expected to yield multiple off target effects across enzyme and organ systems, with many proteins engaging the like for electron transfer. Given the potential for high levels of off target binding, it is generally not exploited as a design strategy. And, while there have been some NOX inhibitors described to target the catalytic portion of the domain directly, there are few specific descriptions of how these inhibitors work mechanistically. With respect to selectivity, one rational strategy that is employed for the purpose of designing NOX inhibitors is disruption of cytosolic subunit interactions with the core NOX subunit, or with each other, thereby blocking the activation of the complex through blocking assembly. Perhaps the most logical of these would be strategies focused on blocking the interaction of the activator subunits (p67phox and NOXA1) with their respective NOX subunits. Generally, this strategy involves blocking the interaction of a core membrane subunit with the activation domain of p67phox/NOXA1, thus blocking NOX activation [65]. This strategy has

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

previously been applied to the design of peptidic inhibitors mimicking these specific sequences, and certain small molecule inhibitors have been speculated to block this interaction as well. Similar strategies can be applied to the organizer subunits (p47phox/NOXO1), which bind the other cytosolic subunits and translocate with them to the membrane-bound NOX/p22phox complex. This then indirectly blocks the ability of the activator subunits to associate with the catalytic subunit by inhibiting complexation. Specifically, there are multiple strategies that may be put forward here. One would be to block the association of the organizer with the polybasic region of the NOX subunit B-loop, which then prevents organizer association [66]. Another would be to target the p22phox/organizer subunit interaction by blocking the organizer’s SH3-domain interface with the proline-rich region of p22phox [67, 68]. Similarly, it may be possible to block the SH3 domain-mediated interaction of p47phox and p67phox prior to translocation, which would then inhibit proper translocation of p67phox alongside p47phox following phosphorylation, although no current inhibitors exploit this strategy to our knowledge. There is also the possibility of targeting the interaction of the catalytic core NOX subunit with p22phox, which is critical for membrane localization and stability, an interaction that has been shown to be influenced by specific amino acids. To our knowledge, there have not yet been any inhibitors described that target this interaction, and selectivity would be dubious given the dependence of NOX1, NOX2, NOX3, and NOX4 on p22phox and the degree of homology among them. Finally, blockade of the interaction of the small GTPase Rac1/2 with p67phox via targeting the tetracopeptide repeats that facilitate this interaction is a plausible approach that has been employed as well [69]. It should be noted, however, that a focus on the interaction between Rac and the activator subunit should strictly focus on only blocking this interaction, as blocking Rac activity itself is likely to have detrimental consequences given the manifold NOX-independent signaling pathways in which Racs are involved [70]. In aggregate, subunit association and complex assembly is of the most preferred methods of NOX inhibition that only stands to increase in popularity as more structural information becomes available elucidating the myriad interaction sites of these isozymes. To date, peptidic and small molecule inhibitors have been designed to target a number of these interactions. There are approaches to indirect inhibition of NOX signaling that may be practical. One such option is suppression of effector molecules that induce NOX activity, such as kinases. For instance, with respect to NOX2 oxidase, it is well-established that its organizing subunit, p47phox, is generally considered to be dependent on PKC-mediated phosphorylation to ensure proper unfurling of its autoinhibitory loop, although other kinases have been implicated in p47phox regulation as well [24, 71]. The displacement of this

349

autoinhibitory loop exposes the SH3 domains and allows for p47 binding to the p22phox C-terminus. In this way, it is reasonable to assume that inhibition of PKC would inhibit canonical NOX2 activation. Similarly, it has been demonstrated that PKC-mediated phosphorylation may regulate NOX1 association with NOXA1 [72]. Conversely, it has been suggested that cyclic AMP-dependent protein kinase/ protein kinase A (PKA)-mediated phosphorylation of NOXA1 can inhibit NOX1 through increased association with 14-3-3 proteins and ensuing blockade of subunit association [73]; thus, agonists for PKA may elicit a NOX1inhibitory effect. The latter has been observed with NOX4 as well, in the context of protein kinase Fyn-mediated tyrosine phosphorylation of the NOX4 dehydrogenase domain and resultant disruption of the interaction between NOX4 and p22phox [35]. There is also the potential for targeting other upstream signaling pathways. For example, NOX5 and DUOX1/2 are predominantly activated by Ca2+ binding to their EF-hand domains [74]. Therefore, it would stand to reason that limiting calcium mobilization by, for example, inhibiting phospholipase C-mediated liberation of inositol triphosphate or blocking of various G protein coupled receptors (GPCRs) that induce this activation, could conceivably inhibit NOX5 and/or DUOX1/2 to a greater degree than the other NOXs [75]. As we continue to learn about the juxtaposition of NOXs and calcium uptake, and their distribution in a signalosome, we may be better positioned to “dial” down activity of one NOX over another. Finally, there is the possibility of addressing NOX signaling and the resultant ROS through various antioxidant-based approaches, in either a generalized or targeted manner. These various approaches, however, are expectedly not as desirable for NOX inhibition, as there is a high degree of possible off-target inhibition that could be detrimental. While the aforementioned strategies would be expected to inhibit NOX activation or reduce the main products of NOX activity, the concern would be the wide-ranging blockade of downstream pathways. Case-in-point, previously termed NOX inhibitors Shionogi-1 and -2 were later found to inhibit PKC [76] whose substrates are numerous. In a similar vein, generalized antioxidant treatment has been demonstrated to evoke many different drawbacks and provide little to no salutary effects in clinical trials, and is therefore considered to be ineffective when compared to specific targeting of ROS sources [5].

5

Pan-NOX Inhibitors (Including Those with More-Focused Selectivity)

The intent of this chapter is to evaluate and describe the current state of isoform-selective NOX inhibitors. Still, it is beneficial to discuss the classic pan-NOX inhibitors. Indeed,

350

pan inhibitors were some of the first NOX inhibitors available and are still widely used today as an effective, albeit often misguided, confirmation of NOX activity in experiments. The most well-known of these inhibitors is diphenyleneiodonium (DPI) which is a general inhibitor of NOXs that irreversibly targets the FAD-binding site in the dehydrogenase domain [77]. DPI is potent, boasting an IC50 in the low nanomolar range, and has been used as a NOX inhibitor since its identification in the late 1980s [78]. However, in that time, it has been proved non-specific as it is capable of irreversibly targeting other flavoproteins, as mentioned above. More recent work describes the development of DPI-related analogs for improved solubility and efficacy, however relative selectivity of these analogs with respect to other NOXs and FAD-requisite enzymes remains nebulous [79]. Indeed, terms such as “broad specificity” and “selective for multiple NOXs” used ubiquitously in the field are oxymorons and should be avoided. The issue of promiscuity of inhibition likewise plagues the NOX inhibitor 4-(2-aminoethyl)-benzenesulphonyl fluoride (AEBSF), which was demonstrated to be an effective NOX inhibitor that was shown to prevent association of p67phox and p47phox with the catalytic subunit of NOX2 in cell-free preparations, but shows off-target serine protease inhibition [80]. Given the ubiquity and importance of serine proteases, it is not particularly enticing as a therapeutic molecule. In addition to the aforementioned shortcomings of iodonium compounds, a broadly used pan-NOX inhibitor for probing NOX activity is the plant-derived catechol apocynin. Originally isolated from P. Kurroa, apocynin had been shown to be an inhibitor of neutrophil oxidative burst and, consequently, of NOX activity, acting through inhibition of assembly via blockade of p47phox translocation [81]. Interestingly, it was later shown that apocynin is, in fact, a pro-drug. That is, there is a requirement for apocynin to dimerize by way of various peroxidases, predominantly myeloperoxidase in neutrophils, to effectively inhibit NOX [81]. Interestingly, this would also require MPO [82] or another cell-specific isozyme like vascular peroxidase 1 (VPO1) [83] in the parenchyma. Moreover, with time it was demonstrated that apocynin (or its metabolite) was not, in fact, a direct inhibitor of NOX themselves. Rather, it was demonstrated to be a general ROS scavenger under some conditions and concentrations which, in turn, significantly tempered confidence in its NOX-inhibitory effects [42]. To complicate matters, it has been demonstrated that apocynin, like DPI, is capable of inhibiting multiple other oxidases [84]. Despite this, apocynin continues to be used, and has, in fact, been employed to suppress NOX activity in pre-clinical/clinical disease models. But, given the delineated concerns, the true participation of NOX in these results are called into serious question.

C. M. Dustin et al.

It has been suggested that the synthetic polyphenol S17834 may more accurately satisfy the definition of a pan-NOXi, as it was shown to inhibit general NOX activity in vascular endothelial cells (reported IC50 9.8 ± 1.3 – 56 ± 9.6 μM, depending on the cell type) [85]. While it was originally illustrated that it has seemingly no effects on oxidant signals directly [85], S17834 was indeed exhibited to have off-target effects, the most notable of which being activation of AMP-activated protein kinase, which is involved in many critical metabolic processes [86, 87]. Two additional agents classified as pan-NOX inhibitors are the triazolo pyrimidine derivatives VAS2870 and VAS3947. These agents reportedly target the NOX through a covalent interaction originally thought to prevent complex formation [88], yet recent findings suggest that they act through a dehydrogenase domain-targeted cysteine modification that is capable of blocking electron transport [39]. In either case, VAS2870 and VAS3947 are widely thought to be bona fide NOX inhibitors, in contrast to DPI and apocynin, as they do not demonstrate inhibition of non-NOX ROS-producing enzymes or display intrinsic antioxidant activity. This efficacy of bona fide NOX inhibitors VAS2870 and VAS3947 has been demonstrated in multiple cell-free and whole cell settings with various NOXs, using both pre-assembled NOX and NOX activated with a stimulus, successfully inhibiting ROS production with a reported IC50 of 10.6 μM in cell-free preparations and a reported IC50 between 0.77 and 2.0 μM in whole cell assays for VAS2870 [76, 89, 90] and between 2.0 and 13.0 μM for VAS3947 (depending on the specific NOX expression pattern of the cell line used) [91, 92]. All the same, this notion is unfortunately obscured by the observed absence of semi-recombinant NOX2 inhibition by Gatto and colleagues, perhaps suggesting that VAS2870 may instead inhibit upstream effectors of the NOX, [76] or be a pro-drug converted to a more active metabolite yet-to-be revealed. Furthermore, there has been some evidence that VAS2870 and VAS3947 engender off-target effects on other proteins’ Cys residues, most notably the demonstrated off-target binding to the RyR1 channel [93]. Given the wide-ranging importance of cysteines in biological processes, there is some concern that collateral Cys-alkylation could disrupt other cellular processes. Adding to that concern, it has been suggested that VAS2870 causes cellular toxicity at concentrations between 2.5 and 4.2 μM [41], as measured by calcein-AM, which may be a significant drawback of using the VAS compounds for NOX inhibition given that the reported IC50 in whole cell- and cell-free systems is >two-fold higher [89]. In sharp contrast to the findings by Gatto and colleagues, the contention that VAS compounds are non-specific NOX inhibitors is called into question by recent data from Reis and colleagues, suggesting that the VAS compounds are bona fide NOX inhibitors acting

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

through specific covalent alkylation of a cysteine residue within the dehydrogenase domain, shared by all NOXs and ferredoxin NADP+ reductases and involved in the binding of the nicotinamide ring of NADPH [39]. Recent developments have brought to the fore another putative pan-NOX inhibitor: APX-115. APX-115 is a pyrazole derivative [94] identified using high-throughput screening, resulting in the discovery of the formerly-termed Ewha-18,278 (later acquired by AptaBio, Inc. and renamed APX-115) [94]. Ewha-18,278 displayed near complete inhibition of induced cellular ROS production in bone marrowderived macrophages stimulated with receptor activator of nuclear factor kappa-B ligand (RANKL) at 10-20 μM, as well as inhibition of NOX1 (Ki: 1.08 μM), NOX2 (Ki: 0.57 μM), and NOX4 (Ki: 0.63 μM) in a reconstituted system. These values were reported as Ki (a measure of inhibitor affinity for an enzyme) rather than IC50, which can be calculated from Ki when enzyme concentration is taken into account. Ewha18,278 was also predicted to target the NADPH binding site in silico, pointing to a distinct mechanism of action [94]. APX-115 was reportedly capable of inhibiting NOX5 in a mouse model of diabetic nephropathy [95]. Importantly, APX-115 displayed no xanthine oxidase or glucose oxidase inhibitory activity at biologically effective concentrations, suggesting efficacy delimited to NOX isozymes, and minimal to no measurable ROS scavenging activity [94]. Since its initial discovery and proved inhibition of NOX-dependent signaling in osteoporosis, APX-115 has been utilized by multiple groups to disrupt NOX in other contexts. Importantly, following preclinical pharmacokinetic studies demonstrating significant absorption of APX-115 in mouse models, two recent clinical trials have been initiated to test APX-115 in the treatment of diabetic nephropathy in human subjects (Trial ID: NCT04534439) and inflammation underlying COVID-19-induced pneumonia (Trial ID: NCT04880109).

5.1

Celastrol

Celastrol, a plant extract derived from Celastraceae, is a similar case. While it has been demonstrated to inhibit NOX1 (IC50: 0.41 ± 0.20 μM) and NOX2 (0.59 ± 0.34 μM) fairly potently via binding to either p47phox or NOXO1, it has also shown significant inhibitory capacity of NOX4 and NOX5 at roughly ten-fold higher concentrations[96]. However, the mechanism of the latter inhibition is unknown. While Celastrol has been applied in various contexts as a NOX inhibitor [97, 98], significant concerns have been raised regarding its potential off-target effects, weakening its potential as a therapeutic (reviewed in [99]), in addition to recently determined inhibition of

351

xanthine oxidase activity, assay interference, non-specific radical scavenging properties [39].

6

and

NOX1

Targeted inhibition of NOX1, along with NOX2, has logically been at forefront of attempted isoform-specific NOX inhibition given the relatively intricate constitution of these two enzyme complexes in the NOX family. Indeed, the intricacy of numerous important interactions among subunits allows for perhaps a more facile disruption of the assembled and latent, activatable enzyme complex. Unlike NOX2, however, inhibition of NOX1 may not carry the same potential caveats, namely alteration of phagocyte-mediated host defenses. Limited evidence has characterized the presence of NOX1 in phagocytes [100, 101], albeit ascribing hostdefense function of lesser magnitude. Aside from the somewhat delimited pan-NOX inhibitor APX-115 described above, the only other drug class to our knowledge that has been approved for clinical trials are dual target NOX1/NOX4 inhibitors (vide infra). To fully appreciate the relevance of NOX1 inhibition and its potential therapeutic value, it is critical to review the various diseases in which NOX1 has been implicated. Indeed, in the colon, where NOX1 is most abundantly expressed [27], predominantly in colonic epithelial cells [102], it has been associated with multiple diseases. For example, it has been observed that NOX1 expression is elevated in colon cancers [103], as well as in lesions associated with inflammatory diseases such as ulcerative colitis and Crohn’s Disease [104]. Thus, NOX1 attenuation, mutation or depletion may one-day prove useful in the treatment of colon cancer and inflammation of the gut. Intriguingly, NOX1 has also been demonstrated to be genetically defective in certain inflammatory diseases of the colon, such as inflammatory bowel disease, suggesting a more protective role [105, 106], indicating that NOX1 targeting must be carefully approached. NOX1 is also decidedly abundant in smooth muscle cells throughout the vasculature, and, at least in rodents, plays a critical role in vascular smooth muscle cell (SMC) proliferation [107, 108]. Likewise, NOX1 contributes to angiotensin II-dependent impairment of vasodilatation and induction of hypertension [109–111]. SMC and endothelial cell (EC) hyperproliferation, hallmarks of vascular remodeling, are observed to be NOX1-mediated in vascular diseases such as pulmonary arterial hypertension [40, 112, 113], as well as restenosis and atherosclerosis[114, 115]. In addition, vascular NOX1 is known to play a role in angiogenesis, which contributes to the growth of tumors [116]. The downside of this, of course, is that blockade of angiogenesis in non-cancerous tissue might be expected to give rise to widespread ischemia and tissue dysfunction.

352

C. M. Dustin et al.

NOX1-mediated ROS are implicated in ischemic retinopathies, specifically playing a role in the promotion of retinal neovascularization, a process that can be stimulated by localized tissue hypoxia during retinal ischemia and can ultimately lead to vision impairment [117]. With respect to inflammatory signaling, NOX1 is present in macrophages, where it may play a role in macrophage differentiation from monocytes, as well as so-called alternative activation of macrophages (also called M2 polarization), a process by which macrophages adopt a more anti-inflammatory and tissue remodeling-focused phenotype vs the classical (M1) inflammatory phenotype [100]. On the other hand, macrophage NOX1 has also been observed to control liver inflammation and macrophage-driven hepatocellular carcinoma [118]. Moreover, NOX1 signaling can play a significant role in the underlying redox-mediated inflammatory mechanisms leading to liver fibrosis following chronic liver disease and hepatocellular injury [119]. In the lung, NOX1 is expressed within the airway epithelium where it is implicated in hyperoxia-induced acute lung injury by way of triggering higher levels of apoptosis, and consequential degeneration of alveoli [120, 121]. NOX1 is also known to play a critical role in the progression of diabetes mellitus, by inducing increased atherosclerosis in response to vascular oxidative stress [122]. Finally, NOX1 is shown to promote increased ROS and redox signaling leading to neuronal dysfunction and death, which purportedly participate in neurological disorders including Parkinson’s Disease [123, 124]. Owing to the multiple disease processes in which NOX1 is involved, there has been considerable effort expended developing NOX1 inhibitors. However, while there are multiple inhibitors capable of reducing NOX1 activity, there are still few inhibitors that specifically target NOX1 (Fig. 21.3).

6.1

Setanaxib and other GKT Derivatives

Perhaps the most-well known of the small molecule NOX inhibitors is the pyrazolopyridine dione Setanaxib (also known as GKT137831/GKT-831), originally identified as a NOX4 inhibitor that exhibited robust inhibition of NOX1, along with its forerunner compound GKT136901 [125, 126]. Setanaxib holds the distinction of being one of the few NOX inhibitors to reach clinical trials to date. In addition to its NOX1i properties (Ki: 140 ± 40 nm), Setanaxib exhibits significant potency and efficacy for NOX4 (Ki: 110 ± 30 nM) as measured in cell-free assays [126]. GKT136901 demonstrates similar potency in cell-free assays with Ki values of 165 nM (±5 nM) for NOX4 and 160 nM (±10 nM) for NOX1 [125] but additionally has been described as selective and direct scavenger of peroxynitrite [127]. Importantly, both inhibitors demonstrated

comparatively weak inhibition of NOX2 (values in the μM range) and partial inhibition of NOX5 (Setanaxib: 410 ± 100 nM; GKT136901: 450 ± 10 nM, both expressed in Ki [126, 128]) with no demonstrable inhibition of xanthine oxidase activity, indicating that they are neither ROS scavengers nor disruptors of ubiquitous flavoenzyme activity like DPI [125, 126]. These two compounds have been utilized to probe the impact of NOX activity in a variety of settings, each propounding their therapeutic potential. For example, when Setanaxib was first described by Aoyama and colleagues, it was reported to prevent progression of CCl4-induced liver fibrosis by inhibiting NOX1 as well as NOX4 [126]. Setanaxib was also used to inhibit NOX1 activity in diabetes mellitus, preventing macrophagemediated inflammation and the formation of atherosclerotic plaques in diabetic mice [122]. On the other hand, GKT136901 has been demonstrated to prevent tumor growth through inhibition of NOX1-mediated ROS, preventing the induction of endothelial migration as a consequence of ROS-mediated peroxisome proliferator activated receptor alpha (PPAR-α) inhibition [116]. Similarly, with respect to the lung, GKT136901 was demonstrated to inhibit STAT3mediated cell death following hyperoxia, a response related to acute respiratory distress syndrome (ARDS), which has previously been associated with NOX1 [120, 121]. Furthermore, NOX1 inhibition by Setanaxib prevented neovascularization, inflammation and vaso-obliteration following ischemia, preventing oxygen-induced retinopathy development, predominantly by hampering VEGF production [117]. With the success of Setanaxib in preclinical NOX1focused diseases, as well the success in treating NOX4focused diseases that will be covered later, Genkyotex Inc. has initiated Phase 2 clinical trials with Setanaxib for primary biliary cholangitis (NCT03226067), Type 2 diabetes/diabetic nephropathy (NCT02010242), as well as an investigatorinitiated trial for type 1 diabetes with albuminuria (U11111187-2609) [129]. More recently, a next generation NOX1 inhibitor termed GKT-771 has been introduced which yielded initial promising preclinical results in the context of hepatocellular carcinoma [130] and HIV-tat-induced endothelial dysfunction, with potential use in the treatment of HIV-related cardiovascular disease [131]. To our knowledge, however, there have been no public pronouncements on GKT-771 following planned phase I clinical trials in 2017. Notwithstanding nascent clinical benefits, there have been questions raised recently regarding the nature of the observed NOX inhibition in these studies, i.e. concerns about whether these inhibitors are indeed NOX inhibitors, per se. For instance, recent studies illustrate interference with hydrogen peroxide detection by Amplex Red, which the authors posit may have confounded assertions of NOX inhibition [39, 41]. In fact, Augsburger and colleagues demonstrated

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

353

Fig. 21.3 NOX1 inhibitors. A visual representation of currently available NOX1 inhibitors and their binding sites. Isoformspecific inhibitors are depicted in black, pan-NOX inhibitors are depicted in red and inhibitors that target at least 2 isoforms are depicted in green. Arrows indicate confirmed or predicted sites of interaction, whereas inhibitors listed with a “?” are not characterized as possessing a known or predicted binding site. Created with BioRender.com

that GKT136901 and Setanaxib may not actually inhibit NOX activity directly when comparing the Amplex Red results with HRP-independent assays, citing high background interference with the Amplex Red assay and an inability to inhibit the signal produced by WST-1 or coumarin boronic acid assays [41]. Consequently, the authors indicated that previously observed results of GKT treatment in preclinical models may be a consequence of undefined redox-related mechanisms [41].

6.2

ML171 and ML090

A small molecule compound originally touted as a NOX1selective inhibitor is the phenothiazine inhibitor ML171, identified by high-throughput cell-based screen [132]. ML117 was initially shown to be NOX1-specific and relatively potent, with an IC50 ranging between 0.129 μM and 0.25 μM, depending on the cell type, and was thought to target the catalytic subunit [132]. That contention alone should have raised considerable concern. Additionally, ML171 showed promising functional utility in preventing invadopodia formation in colon cancer cells [132]. Following those studies, it was shown that ML171-mediated inhibition of NOX1 could prevent the progression of thrombus formation via inhibition of NOX1 in both mouse- and humanderived platelets [133] in addition to inhibition of NOX1-

mediated Syk activation controlling platelet aggregation [134]. Moreover, recent findings indicate that the effect involves cooperative NOX1 and protein disulfide isomerase (PDI) signaling, a consequence attenuated by simultaneous inhibition of both NOX1 and PDI, with marginal effect of inhibiting either protein alone [135]. And, more recently, ML171 was observed to inhibit NOX1 in extracellular vesicles generated from activated platelets, preventing propagative platelet activation [136]. With respect to the vasculature and hypertension, ML171 attenuated the rise in blood pressure in mice in response to vasoconstrictor hormone angiotensin-II, further tendering that NOX1-derived ROS may play a role in mitochondria/cyclooxygenase-2 mediated hypertension [137]. However, recent studies reveal that ML171 will also directly interfere with commonly employed ROS detection methods, e.g. Amplex Red and luminol-based assays [43, 138]. What’s more, ML171 may also inhibit other NOX isoforms and xanthine oxidase, albeit at a roughly ten-fold higher concentration than that required for NOX1 inhibition [43, 99]. As such, it has been suggested that ML171 does not qualify as a NOX-selective inhibitor [138]. Toward the same end, a structurally related inhibitor, ML090, was developed to target NOX1 [139], and utilized to inhibit NOX1 in homocysteine thiolactone-induced vasoconstriction [140]. That being asserted and despite its initially purported NOX1 specificity [139], recent evidence suggests that ML090 more potently inhibits NOX5 when

354

C. M. Dustin et al.

compared to NOX1 and NOX4 [43]. Similarly, ML090 significantly interfered with ROS detection assays, and was revealed as a possible H2O2 scavenger calling its validity as an inhibitor into question [39]. Despite these issues, ML171 has recently been used in multiple studies to probe for NOX1 function - as such, those results should be interpreted with caution.

6.3

NOS31

The most recent small molecule NOX1 inhibitor to be characterized, NOS31, was described by Yamamoto and colleagues in 2018. NOS31 was identified as a NOX1 inhibitor though a screen of microbial metabolites, revealing the naturally occurring 2,3-disubstituted-2,3-dihydrobenzofuran termed NOS31 [141]. Utilizing a NOX1 overexpressing HEK293 cell line, NOS31 displayed an IC50 for NOX1 of 2 μM. A ten-fold higher concentration inhibited NOX4, while inhibition of other NOX isoforms required much higher concentrations [141]. On this basis and having not displayed any background radical scavenging activity or xanthine oxidase inhibition, NOS31 was deemed a potential NOX1 inhibitor [141]. In the same study, NOS31 inhibited growth of hepatic carcinoma cells, a cancer linked to overproduction and aberrant activation of NOX1 as well as the growth of gastric cancer cells overexpressing NOX1 [141]. While there is still no further work regarding NOS31 and its application toward other NOX1-dependent diseases, and more rigorous information regarding its level of interference with common biochemical assays aside from luminescence and/or impact on cell viability is necessary, NOS31 appears to hold some promise as effective in a milieu in which NOX1 is the dominant NOX.

6.4

NOXA1ds

While discussion of inhibitors as drugs typically are fixated on small molecule inhibitors, there have been notable developments with respect to peptidic inhibitors within NOX research and their delivery systems. Indeed, as multiple inhibitory peptides have been described in the context of probing specific structural features and interactions of NOX subunits through random sequence phage display and “peptide walking” strategies[142–144], some peptides have been expressly chosen for the purpose of inhibition. Such is the case for NOXA1ds, a peptidic inhibitor postulated by our group to inhibit NOX1 oxidase by blocking the interaction between the catalytic subunit and the activator subunit NOXA1 [65] (ds referring to “docking sequence”). NOXA1ds was deliberately designed to achieve its selective inhibitory effect by utilizing a partial sequence of the

purported activation domain homologous to p67phox in NOXA1 and a flanking non-homologous region. The goal was to block the canonical assembly of NOX1 oxidase [65], and the result was a peptide inhibitor that was specific for NOX1 with an IC50 of 20 nM, showing no inhibition of NOX2, NOX4, or NOX5 in related heterologous systems [65]. Importantly, to rule out any non-specific inhibition by the peptide, a sequence-scrambled peptide was tested, which failed to inhibit NOX1 activity [65]. Furthermore, the same study demonstrated that NOXA1ds was capable of permeating cell membranes, in addition to inhibiting hypoxia-induced O2- production by NOX1, further establishing a role of NOXA1ds as a potent, cell and tissue penetrating selective NOX1 inhibitor [65]. To date, NOXA1ds has been used to “dial down” NOX1 activity in multiple pathologies. Namely, our group has utilized NOXA1ds to probe the contribution of NOX1 in endothelial proliferation and migration underlying pulmonary hypertension [40, 112], as well as in cellular senescence and the inflammatory milieu arising with age [145, 146], and stretch-induced ROS generation leading to altered vascular smooth muscle-dependent remodeling/dysfunction [147]. Other groups have utilized NOXA1ds in other vascular contexts. For example, NOXA1ds applied to vascular smooth muscle cells isolated from spontaneously hypertensive rats revealed that NOX1 plays a role in ROS generation and protein oxidation/stress responses [148] as well as in vascular endothelial growth factor (VEGF) inhibitor-induced ROS production in hypertension [149]. In a similar manner, NOXA1ds demonstrated the capacity to inhibit NOX1dependent ROS production enabled by protein disulfide isomerase a1 overexpression in VSMCs with implications for vascular remodeling [150]. NOXA1ds was employed to inhibit NOX1-dependent peroxynitrite formation in a pulmonary hypertension model, revealing an antagonistic effect of NOX1 on endothelial transient receptor potential cation channel subfamily V member 4 (TRPV4) and a resultant increase in pulmonary artery pressure [151]. Two recent studies utilizing NOXA1ds demonstrated that inhibition of NOX1 in obesity models can ameliorate ROS production and increase vasodilatation in renal arteries isolated from an obese Zucker rat model [152], as well as attenuate hypertension by inhibiting peroxynitrite formation and its attendant antagonism of TRPV4 channels in high fat diet-fed mice [153]. Further in vivo evidence for the efficacy of NOXA1ds was provided in two recent publications studying the impact of NOX1 in experimental ileocolitis, of which there was a reduction in crypt cell apoptosis in response to NOX1 inhibition [154], and our laboratory’s study demonstrating an unexpected suppression of age-related inflammatory/ senescent phenotypes following NOXA1ds administration [146]. Finally, NOXA1ds was shown to inhibit ROS production in mutant KRAS-expressing colon carcinoma,

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

suggesting a potential role for NOXA1ds in disrupting NOX1 signaling in colon cancer [155]. Overall, these studies suggest a broad utility with respect to NOXA1ds.

6.5

NF02

Another inhibitory peptide targeting NOX1 has also recently been described by Mousslim and colleagues, termed NF02 [156]. NF02 was uncovered in a screen of peptides from the active site of bacterial dehydrogenases and displayed an ability to inhibit NOX1 activity with an IC50 of 16.7 ± 5.4 μM in an intact HT29 cell assay with no cytotoxicity, background ROS scavenging, or inhibition of NOX2dependent signaling [156]. Further, the group demonstrated NF02’s ability to inhibit wound closure and spheroid invasion at slightly higher concentrations (circa 50 μM) in colorectal cancer cells [156]. However, as members of the same group note in a later review, NF02 has only been tested against NOX1 and NOX2, and has not yet been tested for inhibition of NOX4 or NOX5 [157]. To date, there has been no further work, to our knowledge, with this peptide. Still, it could prove to be a promising alternative to other peptidebased inhibitors currently available.

7

NOX2

Of the 7 NOX isoforms in mammals, not surprisingly the majority of inhibitor-focused research has been centered around the most well-studied NOX2. In line with NOX1, NOX2 activity is also highly dependent on its association with its cytosolic subunits, providing numerous attractive targets for isoform-specific targeting and rational inhibitor design strategies. However, unlike NOX1 targeting, NOX2 targeting is burdened with the unfortunate downside of possible host defense inhibition given NOX2’s critical role in phagocyte-mediated host defenses. While many clinician scientists remain wary of pursuing NOX2 as a target on account of concerns of inducing a phenotype similar to chronic granulomatous disease (CGD), the degree of NOX2 inhibition required to elicit this phenotype is unexpectedly high. Indeed, CGD patients with neutrophils producing low residual levels of ROS still present with significantly lesssevere disease and manageable symptoms, which (as described by the authors) can be as low as 1% residual activity without significant impact to mortality [158]. Owing to this high bar for deleterious inhibition, the therapeutic window for NOX2 inhibitors to abrogate parenchymal NOX2 in disease appears quite large. NOX2 is recognized as playing a role in a wide range of diseases across multiple tissue types, making it attractive for targeted therapies. For example, one of the most widely

355

studied areas of NOX2 research involves NOX2 in the vasculature and in vascular disease [159]. Further to that point, NOX2 has been shown to play a role in hypertension, in response to angiotensin II signaling, both with respect to the induction of vascular remodeling [160] and the induction of inflammation or medial hypertrophy [161, 162]. Additionally, NOX2 is involved in reducing NO bioavailability via its conversion to peroxynitrite by its O2- generation and/or the induction of nitric oxide synthase (NOS) uncoupling [163], leading to elaboration of ROS production. NOX2 is known to be elevated in atherosclerosis [164] and has been shown to play a key role in the formation of atherosclerotic plaques [165]. Indeed, Judkins and co-workers showed that endothelial NOX2-derived O2- plays a direct role in limiting NO bioavailability (reportedly the result of either conversion to peroxynitrite or eNOS uncoupling) in high fat diet apolipoprotein E deficient (ApoE-/-) mice, a common atherosclerosis model, due to higher levels of NOX2 expression and activation, leading to increased early aortic lesions [166]. However, similar to its role in hypertension, NOX2’s involvement has been challenged. For instance, NOX2 overexpression in ApoE-/- did not influence systolic blood pressure or plaque formation, despite evidence of increased endothelial O2- and macrophage recruitment [167]. With respect to injury, NOX2 has been implicated in ischemic and ischemia-reperfusion (I/R) damage, both with respect to cardiac, as well as neurological, pathologies. For example, in myocardial I/R and myocardial infarction, higher levels of NOX2 ensue; in contrast, I/R NOX2 KO mice are characterized by lower myocardial infarct size, as well as lower levels of inflammation, myocardial remodeling, and ROS/oxidative tissue damage compared to wildtype [168, 169]. In similar fashion, neurological I/R injury can be heavily influenced by NOX2. For instance, in the case of ischemic stroke, NOX2 played a significant role in the elevation of ROS and ROS-mediated tissue damage whereas inhibition or knockout of NOX2 led to smaller infarct sizes and improvement in cerebrovascular function [170–172]. NOX2 may play a role in a wider field of neurological diseases not related to vascular function, e.g. NOX2 has been suggested to play a role in Alzheimer’s Disease progression, as NOX2 subunit levels and NOX2 activity increase as Alzheimer’s Disease progresses [173–175], as well as Parkinson’s disease, wherein NOX2 has been demonstrated to play a direct role in cytotoxicity and Parkinson’s disease-associated protein activation in dopaminergic neurons [176] and microglia [177]. With respect to pulmonary diseases, there is also a significant contribution of NOX2 signaling as well. For example, NOX2 has been linked to macrophage-mediated alveolar destruction in chronic obstructive pulmonary disease [178]. NOX2 has also been implicated in pulmonary hypertension. In particular, it has been shown that NOX2 KO mice

356

C. M. Dustin et al.

Fig. 21.4 NOX2 inhibitors. A visual representation of the currently available NOX2 inhibitors and their binding sites. Isoform-specific inhibitors are depicted in black, pan-NOX inhibitors are depicted in red, and inhibitors that inhibit at least 2 isoforms are depicted in green. Red “blocking” arrows indicate confirmed or predicted sites of interaction whereas inhibitors listed with a “?” are not characterized as possessing a known or predicted binding site. Created with BioRender.com

fail to develop classic features of pulmonary hypertension, including increased pulmonary artery pressure and increased vessel remodeling and right ventricular hypertrophy [179]. NOX2 could also upregulate ROS in conjunction with mitochondrial signaling and NOX4, wherein activation of mitochondrial ROS production induces downstream NOX2 expression both in vivo and in isolated human pulmonary artery endothelial cells [180]. However, the manner by which NOX2 participates in this process is unclear [181], i.e. it may not be transcriptionally upregulated in pulmonary hypertension as is the case for NOX1 and NOX4, instead it appears to more classically conform to paradigmatic posttranslational activation [182]. Given these broad diseasemanifesting roles, inhibition of NOX2 has been of intense interest within the NOX field for some time. As such, NOX2 has received the most attention regarding inhibitor development (Fig. 21.4).

7.1

NOX2ds-tat

The peptide inhibitor NOX2ds-tat (formerly gp91ds-tat) was the first described isoform-specific NOX inhibitor, preferentially targeting NOX2 over the other NOX isoforms. NOX2ds-tat was first described in 2001 and was designed in an effort to create a cell-permeant agent that would inhibit NOX2 with the goal of probing the contribution of angiotensin II (ANGII)-mediated NOX2 activation and O2- production to blood pressure elevation and vascular dysfunction [66]. While other peptides have been characterized to inhibit

NOX2 in cell free systems, they were designed for the purpose of probing critical interactions involved in NOX assembly and are as-of-yet not thoroughly tested for isoform selectivity and for in vivo effectiveness to our knowledge [142–144]. To elaborate, the chimeric peptide NOX2ds-tat was specifically generated based on a positive hit in a phage display library screen observed to most effectively inhibit NOX2 [144]. The “NOX2ds” portion of the peptide mimics an 11 amino acid stretch of the rodent cytosolic B-loop of NOX2 (altered by one amino acid) at which the organizing subunit p47phox “docks”, ergo ds for docking sequence. Additionally, NOX2ds-tat contains a small portion of HIV tat protein, which aids with cellular internalization, allowing NOX2ds-tat to reach the cytosol where p47phox is localized under latent conditions [66]. Indeed, NOX2ds-tat was capable of specifically blocking ANGII-mediated O2- production in both in vitro assays utilizing aortic sections, and in ex vivo in aortic rings isolated from mice injected with NOX2ds-tat directly. Importantly, little if any collateral inhibition of xanthine oxidase and no radical scavenging activity was observed. Moreover, no adverse reactions in vivo were observed when the inhibitor was infused for up to 7 days, and the peptide was found to be stable in aqueous buffer solution (in minipumps in vivo) for that duration. Additionally, ~35% maximum inhibition of O2- was detected when intact human neutrophils were pretreated with the peptide at twice the concentration of the maximally effective concentration of 50 μM used on mouse vascular cells and tissue, suggesting perhaps a differential and fortuitous inability to equitably inhibit human neutrophil NOX2 [66]. We surmised

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

that NOX2ds-tat was degraded by exoproteases on the surface of neutrophils and thus less bioactive upon traversing the plasma membrane, but this has yet to be tested. NOX2ds-tat was indeed able to significantly lower ANGII-mediated blood pressure elevation in comparison to a peptide control, indicating that this inhibition is sequence-specific to NOX2ds-tat and not properties of the amino acid constituents themselves [66]. Our group later demonstrated specificity of NOX2ds peptide in the blockade of p47phox in the NOX2 oxidase, and thus only inhibits canonical NOX2 (not the hybrid NOX1) determining the IC50 to be 0.74 μM in a cell-free system [183]. Following its initial discovery, NOX2ds-tat has been exploited in a multitude of contexts, both in vitro and in vivo. In addition to much of our own work exploring the contribution of NOX2 to various vascular diseases such as hypertension [66], smooth muscle hypertrophy, and neointimal hyperplasia associated with disease in vivo [184], others have shown that NOX2ds-tat can abrogate NOX2’s contribution to not only hypertension, but other critical processes in diseases such as diabetes [185], atherosclerosis [165], Alzheimer’s Disease[174, 175], Parkinson’s Disease [57], pathological platelet activation [186], stroke [175, 187], and ischemia-reperfusion injury[188], among others.

7.2

GSK2795039

GSK2795039 is a 7-azaindol derivative that was initially described by Hirano and colleagues at GlaxoSmithKline, the result of a high throughput screening NOX2i assay [189]. GSK2795039 was demonstrated to be highly specific for NOX2 and showed little background radical scavenging or inhibition of xanthine oxidase activity, as compared to DPI. Additionally, GSK2795039 was demonstrated to have an IC50 between 0.269 and 0.660 μM in different cell-free NOX2 assays, depending on the assay used, with minimal inhibitory effects on xanthine oxidase activity [189]. Whole cell assays using L-012 and Oxyburst revealed similar IC50 values between 0.182 μM and 0.186 μm, respectively [189]. Despite some reported electron donor activity that can suppress Amplex Red’s ability to detect ROS, GSK2795039 displayed relatively strong NOX2 selectivity with only weak inhibition of the other NOX isoforms utilizing the WST-1 assay [189]. The IC50 for NOX2 was indicated as 2.88 μM (WST-1 assays) and < 100 μM for other NOXs. However, later findings indicated that GSK2795039 may also weakly inhibit NOX4 [39]. Further, GSK2795039 crossed the blood brain barrier, was orally bioavailable, and possessed the capacity for NOX2 inhibition in vivo [189]. The authors reconstituted recombinant cytosolic NOX2 subunits to explore the mechanism of action for

357

GSK2795039, ultimately demonstrating that GSK2795039 competitively inhibits NOX2 by blocking NADPH binding [189]. To date, there have been no reported clinical trials initiated using GSK2795039, but a recent study describing structural changes to the compound in response to incubation with microsomal and cytosolic fractions collected from mouse, rat, and human livers revealed notable metabolism of the molecule that may be responsible for its rapid clearance [190]. Despite this, there has been some reported preclinical use of the molecule reported by multiple groups. In vitro, GSK2795039 can inhibit ROS production in microglia and prevent neuronal death following iron release and inflammation [191], and has the capacity to suppress proliferation and migration of vascular smooth muscle cells [192]. In vivo, GSK2795039 exhibited promising protective effects in a preclinical model of traumatic brain injury [193] and prevented inflammation and vascular damage triggered by ROS in response to e-cigarette vapor [194]. More recently, studies suggested that its NOX2 blockade can prevent β-amyloid-dependent features of Alzheimer’s disease and epileptic seizure in mice [195], and can stave off ROS-mediated thrombosis in a mouse model of heparininduced thrombocytopenia [196]. Modifications of GSK2795049 to prevent rapid clearance were suggested by Pradilha and colleagues [190], and clinical trials are expected.

7.3

CYR5099

First unveiled in a report by Liu and colleagues in 2019, CYR5099 is a benzo[f]indol-9(4H)-one derivative of a NOX2 inhibitor screen hit [197]. Utilizing a variety of neutrophil activators, CYR5099 was deemed highly efficacious at inhibiting isolated neutrophil O2- production with an IC50 between 2.80 ± 0.27 μM and 5.19 ± 0.49 μM, depending on the stimulus used, with minimal to no cytotoxicity and direct ROS-scavenging effects up to 15 μM [197]. Further analyses showed that, based on molecular docking analysis, CYR5099 purportedly targets the FAD-binding region of the catalytic subunit, and displays no inhibition of NOX1 activity [197]. Further, it was shown that in vivo administration of CYR5099 prevented ROS-mediated inflammation and paw injury in an arthritis model [197]. In aggregate, these results suggest that CYR5099 could hold promise as a NOX2 inhibitor, however its true specificity is still unknown given that only inhibition profiles of NOX1 were compared and contrasted against NOX2. Given limited reported information on this molecule, more evidence will be required to elevate the potential for CYR5099.

358

7.4

C. M. Dustin et al.

PHOX-I1 and PHOX-I2

PHOX-I1 and PHOX-I2 are two small molecule inhibitors that, to our knowledge, are the only inhibitors that target the interaction between Rac1/2 and p67phox [69]. This was achieved visualizing X-ray crystallography structures of p67phox and Rac1 in complex and assessing for points of contact and discrepancies among isolated p67phox crystal structures [69]. The investigators utilized an automated screening program to assess compounds for binding capacity at the p67phox-Rac1 interface, identifying the molecule that was termed Phox-I1, and showing its capacity for NOX2 blockade in cell-based assays (IC50: 3 μM in dHL-60 cells, 8 μM in isolated neutrophils) without ancillary antioxidant activity [69]. Curiously, Phox-I1 was incapable of inhibiting PMA-induced ROS, which the authors reasoned might have arisen from Phox-I1 inhibiting NOX2 through alternate mode of activation [69]. Through SAR analyses, the authors identified a structural analog Phox-I2 (IC50: 1 μM in dHL-60 cells, 6 μM in neutrophils) [69]. Neither inhibitor was capable of inhibiting NOX4, indicating a modicum of relative effectiveness, yet the authors did not test for NOX1, NOX5, no less DUOX, inhibition. It is important to note at this juncture that the omission of DUOX in screens is a criticism that is broadly applicable across compound testing, including our own. Interestingly, the ability of this compound to inhibit the Rac: NOXA1 complex was not interrogated [69]. With that stated, multiple studies in recent years have utilized Phox-I2 as a NOX2 inhibitor in in vitro studies [198– 201]. Neither Phox-I compound has been evaluated in vivo to our knowledge.

7.5

Ebselen

Ebselen has long been known to act as an antioxidant and glutathione peroxidase mimic. In 2012, Smith and colleagues provided compelling evidence for ebselen congeners inhibiting the assembly of p47phox and p22phox utilizing a number of assays, including fluorescence polarization, which they developed to test this assembly in vitro [68]. They went on to show that ebselen analogs can inhibit Nox2 activity in neutrophils, which requires p47phox binding to p22phox, at approximately the same IC50 concentration at which they inhibit binding and at 2 orders of magnitude lower than concentrations required for its peroxidase activity [68]. The latter supports divergent chemical properties of the analogs above and beyond ebselen’s peroxidase activity. On the other hand, the authors posit that as a consequence of the homology between the NOX2 cytoplasmic proteins and those of NOX1 and NOX3, ebselen and its analogs might

disrupt the assembly of those enzyme complexes like that with Nox2 [68]. Still, the study is lauded for a thorough discussion of the limitations on specificity and a deep investigation of the mechanism of action of the agents studied. With this knowledge in mind, one might consider the predominant NOX isoform expressed in a disease and apply the congener cautiously. While ebselen has been tested in clinical trials for disease treatment, in many cases under the name SPI-1005, there have been no reported trials utilizing ebselen with the express purpose of targeting NOX.

7.6

Perhexiline

Perhexiline is an approved drug commonly used to treat heart disease for its properties as a vasodilator. As such, it is primarily used for the treatment of angina. While the evidence for perhexiline as a NOX-focused therapeutic is scarce, perhexiline did inhibit NOX2 in both isolated neutrophils, as well as in ex vivo preparations of aorta and other cardiac tissues, with an observed IC50 ranging between 2.3 μM and 26.2 μM [202]. Further evidence supported perhexiline as a NOX2i as it could inhibit both NOX2 in whole cells and semi-recombinant assays [76]. A recent manuscript by Reis and colleagues proved that perhexiline inhibits NOX4 with perhaps greater efficacy than with NOX2, and with no effect on NOX5 dehydrogenase activity [39]. It is important to note that, as a therapeutic, perhexiline is not prescribed outside of Australia and New Zealand. Importantly in that regard, perhexiline has the potential to cause adverse effects in individuals that have a “poor metabolizer” phenotype – a consequence of deficiency in the main perhexiline metabolizing enzyme cytochrome p450 2D6, which can result in a perhexiline half-life of up to 40 days [203, 204]. Given this fact, as well as the potential for perhexiline to inhibit carnityl palmitoyltransferase as an off-target effect in NOX2-focused treatments, more rigorous studies are plainly needed to determine the potential of perhexiline as a NOX2-focused therapeutic [203].

7.7

Naloxone

Evidence for naloxone, an opioid receptor antagonist commonly administered to patients to treat opioid overdose, as an inhibitor of NOX2 activity is limited. Naloxone was found to inhibit O2– production in glia without effect on xanthine oxidase activity at an isomer-dependent IC50 between 1.96 μM and 2.52 μM [205]. The authors present evidence that naloxone binds to the catalytic subunit of NOX2, and, in turn, blocks p47phox translocation [205]. Importantly, the

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

authors noted that the concentration of naloxone necessary for NOX inhibition is substantially higher than what is used for opioid receptor antagonism. Indisputably, this calls the therapeutic application of naloxone as a NOX2i into question [205]. Later evidence comparing various NOX2 inhibitors demonstrated that naloxone can inhibit NOX2 in cell-based assays. Even so, its potency is significantly diminished in semi-recombinant preparations suggesting that its action is indirect [189]. Naloxone was reportedly incapable of inhibiting ROS generation in a mouse paw inflammation assay when compared to GSK2795039 [189].

7.8

CPP11G and CPP11H

High-throughput screening of a ~ 600 compounds from a redox-related small molecule library, Cifuentes-Pagano and colleagues discovered a bridged tetrahydroisoquinoline compound as a lead molecule [206]. The plate-based screening platform employed was optimized to provide a stable and highly sensitive NOX2 system. Exploiting SAR and synthetic chemistry to modify a sidechain of the second aromatic ring, two molecules were identified that displayed inhibition of NOX2 relative to other vascular NOX isoforms in COS and HEK cell-based assays. The study utilized 2 independent ROS probes and no alteration of xanthine oxidase activity or ROS scavenging was exhibited [206]. Two active compounds named CPP11G and CPP11H (formerly 11G and 11H in [206]) had IC50 values for of 20 μM ± 1.9 for CPP11G and 32 μM ± 1.9 for CPP11H, respectively, in intact cells [206]. Further work from our group demonstrated that these two compounds were capable of blocking NOX2 assembly with p47phox in isolated endothelial cells in response to TNF-α stimulation [67]. Through computational analysis and structural mapping, a binding site for CPP11G and CPP11H predicted blockade of the well-characterized SH3 supergroove of p47phoxinteraction with the proline rich region of p22phox C-terminus [67]. CPP11G and CPP11H abrogated PMA-induced translocation of p47phox and binding to NOX2 in the plasma membrane in Cos7 cells overexpressing the NOX2 complex components (COS-phox cells). These inhibitors also suppressed TNFα-induced human aortic endothelial cell ROS generation, inhibited both mitogen-activated protein kinase (MAPK) and NFkB activation/nuclear localization, and reduced monocyte adhesion. The latter was consistent with a demonstrable effect of the compounds to suppress EC vascular cell adhesion molecule 1 (VCAM-1) and intercellular adhesion molecule 1 (ICAM-1). In vivo, an i.v. bolus dose of 15 mg/kg prior abolished the stimulatory effect of TNFα on aortic ROS, and inflammatory indicators as well as endothelial dysfunction and major declines in hind-limb blood flow [67]. Further, CPP11G permeated the blood brain barrier following a single

359

oral dose in a rat model of Parkinson’s disease, inhibiting NOX2 assembly and activation in brain tissue [57]. While these are very promising results, more work is needed to ensure efficacy, delineate PK/PD and safety in these and other preclinical models of disease.

7.9

Other Potential NOX2 Inhibitors

Further to the development of NOX2 inhibitors, a recent manuscript by Solbak and colleagues detailed multiple other new compounds designed to inhibit p47phox-p22phox via binding to the SH3 domains of p47phox, similar to the proposed mechanism of action of previously reported ebselen-congeners and the CPP compounds [207]. Utilizing fluorescence polarization, thermal shift, and surface plasmon resonance, the authors were able to identify molecules capable of binding to p47phox [207]. Based on these analyses, the authors cleverly characterized dimeric analogs of the initial screen hits that were capable of binding both p47phox SH3 domains with Ki values between 20-32 μM, of which they note one of the dimers displays a similar Ki value to the p22phox proline-rich domain peptide assessed in their hands [207]. The analyses, however, did not include measurements of NOX activity, and all results are based on p47phox binding in vitro. Therefore, while promising, significant work remains to be undertaken to ensure the specificity and efficacy of these molecules for NOX2 vs. other NOXs and oxidoreductases in vitro and in vivo.

8

NOX4

NOX4 is among the most well-studied NOX isoforms, and the first described, to significantly deviate from the canon of NOX2 complexity and organization. As mentioned earlier, and extensively covered elsewhere in this text (see Chap. 12 by L. Hecker et al.), NOX4 does not require cytosolic subunits for its activation, seemingly only p22phox as a stabilizing subunit. However, there is evidence that the polymerase delta-interacting protein 2 (POLDIP2) positively regulates NOX4 function [208], in addition to other cytosolic proteins such as protein disulfide isomerase (PDI) and tyrosine kinase substrate 4/5 (TKS4/5) [209], though these proteins are not considered obligatory for NOX4 activity. As such, NOX4 is generally thought to be constitutively active and regulated by expression [34], with select evidence that post-translational modifications, e.g. phosphorylation, alter NOX4 activity [35]. In alignment with the other NOXs, NOX4 has been implicated in multiple diseases. Quite uniquely, despite its activation in disease, the role of NOX4 is not always deleterious. In comparison to the generally held assumptions regarding NOX1 and NOX2 in disease,

360

C. M. Dustin et al.

the role of NOX4 has been considered protective in broad contexts be it via a hormetic effect or by engendering a pro-differentiated cell phenotype [210]. As is very likely the case with all NOX isoforms and ROS in general, the magnitude and localization of the specific NOX and its derived ROS signal appear to be the critical determinant of whether a particular signaling event is beneficial or injurious. NOX4 (originally Renox) is highly expressed in the kidney, and, not surprisingly, has been demonstrated to participate in the etiology of a number of kidney diseases [1, 211]. For example, one of the most well-described is its injurious role in diabetic nephropathy, in which NOX4 signaling leads to ROS-induced dysfunction and inflammation, extracellular matrix deposition, hypertrophy, and apoptosis [212]. It should also be noted that, within the kidney and heart, NOX4 may also subserve a salutary role [210, 213– 215], and thus its inhibition is ill-advised if not approached carefully. NOX4 has been suggested to play a critical role in the development/progression of various cardiovascular pathologies including heart failure [216], stroke [217], and pulmonary hypertension [182, 218, 219]. Intriguingly, there is also evidence that NOX4 is not involved in these processes [220], or may play a beneficial role [210]. In the case of atherosclerosis, there is evidence that reductions in NOX4 may contribute to disease progression [221]. NOX4 is also well known to contribute to fibrosis in multiple tissues, predominantly pulmonary fibrosis [222, 223] and cardiac fibrosis [216, 224]. NOX4 effects fibrotic disease generally downstream of transforming growth factor-beta (TGF-β) [225]. This feature makes NOX4 inhibition an attractive

target for fibrosis treatment, as current options are lacking and these diseases are often irreversible. While peptidic inhibitors for NOX4 inhibition have seen mixed success [226, 227], and therefore have not emerged as a viable inhibition strategy, current advancements in small molecule targeting show promise for use in potential therapeutic strategies (Fig. 21.5).

Fig. 21.5 NOX4 inhibitors. A visual representation of the current NOX4 inhibitors and their binding sites. Isoform-specific inhibitors are depicted in black, pan-NOX inhibitors are depicted in red, and inhibitors that target at least 2 isoforms are depicted in green. Red

“blocking” arrows indicate confirmed or predicted sites of interaction, whereas inhibitors listed with a “?” are not characterized as possessing a known or predicted binding site. Created with BioRender.com

8.1

Setanaxib and other GKT Derivatives

Setanaxib and the related GKT compounds, as previously stated, are dual inhibitors targeting both NOX1 and NOX4 with similar potency. In fact, the precursor compound GKT136901 was first described in an effort to find novel NOX4 inhibitors for the treatment of idiopathic pulmonary fibrosis (IPF), and the inhibition of NOX1 was discovered serendipitously [125]. Similarly, later studies from Genkyotex describing Setanaxib (at the time, referred to by its development name GKT137831) revealed a significant contribution of NOX4 to experimental liver fibrosis alongside NOX1, and an ability of the molecule to inhibit both enzymes ultimately demonstrating Setanaxib’s ability to prevent liver fibrosis in mice effected by TGF-β and ANGII signaling [126]. Incidentally, as mentioned previously, Setanaxib also exhibited potency and efficacy toward NOX5 in vitro, albeit to a lesser degree [126]. Since then, the GKT molecules have been employed to target NOX4 in a variety of other pathologies, eventually leading to some NOX4-focused clinical trials. For example, Setanaxib has

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

been utilized to extensively study NOX4’s role in multiple fibrosis models. NOX4 inhibition with GKT136901 and Setanaxib has been proposed as a potential therapeutic for treating pulmonary fibrosis, with studies demonstrating efficacy of NOX4 inhibition downstream of TGF-β in both in vitro and in vivo models and in the commonly used bleomycin-induced fibrosis model [228], preventing epithelial death in vitro [229]. Similar findings were made with two GKT136901-related compounds, referred to as Compounds 87 and 88 by Laleu and colleagues [125], showing a significant reduction of in vivo fibrotic gene expression in lungs, as well as an in vitro reduction in myofibroblast accumulation [230]. Setanaxib was also shown to partially revert age-related senescent and apoptosis-resistant phenotypes in ex vivo idiopathic pulmonary fibrosis (IPF) fibroblasts, in addition to reverting fibrosis and improving survival in vivo following bleomycin treatment of mice [223]. Further, administration of Setanaxib alongside viral delivery of hepatocyte growth factor promoted post-bleomycin regeneration and engraftment of alveolar epithelial cells in mice [231]. A potential caveat, however, is that bleomycin treatment as a pulmonary fibrosis model has the downside of spontaneously resolving itself over time [232]. As such, results regarding fibrosis reversion using this model in any context should be interpreted with caution, as progressive nature of the human disease is not fully replicated in the rodent, and the true efficacy of NOX4 treatments in a human model awaits results from ongoing phase II clinical trials (Trial ID: NCT03865927) [233]. With respect to hepatic fibrosis, the GKT compounds have also seen success in inhibition of NOX4. For instance, Setanaxib reduced profibrotic gene expression and hepatocyte apoptosis in a mouse model of liver fibrosis, preventing or reverting fibrotic development [234]. Similar results were obtained regarding both NOX1 and NOX4 in isolated primary hepatic stellate cells [119]. NOX4 inhibition with Setanaxib prevented fibrosis, reduced inflammation, and improved insulin sensitivity in the livers of mice fed a fast food-based diet in a model of non-alcoholic steatohepatitis, suggesting a role for NOX4 inhibition in diabetic responses as well as fibrosis [235]. Indeed, in terms of diabetes and kidney disease, Setanaxib can prevent ROS-induced glomerular cell hypertrophy in genetically diabetic mice [236]. Setanaxib similarly reduced kidney damage by glomerular alterations and ECM deposition, as well as inflammation, in a streptozocininduced diabetes model, and similarly reduced profibrotic markers in isolated human podocytes [237]. Similar effects were found in a genetic db/db mouse model using GKT136901. These compounds have also been applied to NOX4 in vascular diseases, per se. For example, Setanaxib averted hallmarks of hypoxic pulmonary hypertension in mice, including NOX4-dependent right ventricular

361

hypertrophy and vessel wall thickening. As well, it prevented alterations in response to peroxisome proliferator-activated receptor gamma (PPARγ) [219]. NOX4 inhibition by Setanaxib suppressed ANGII-induced cardiac fibrosis and hypertrophy in mice [224], and inhibited primary cardiac fibroblast proliferation via inhibition of IL-18 in vitro [238]. Overall, the GKT compounds appear to have strong potential as NOX therapeutics, despite requiring validation in clinical trials and the above mentioned characterizations regarding NOX selectivity [41].

8.2

GLX351322, GLX481372, GLX7013114, and GLX481304

NOX4 inhibitor GLX351322 was shown to inhibit NOX4 with an IC50 of 5 μM in NOX4-overexpressing cells, demonstrating a ten-fold higher IC50 for NOX2 and no antioxidant activity [239]. GLX351322 restored insulin sensitivity in both a mouse model of type 2 diabetes and in isolated human islet cells, and offset high glucose-induced cell death [239]. SAR of GLX351322 gave rise to two more inhibitors GLX481372 and GLX7013114, wherein it was found that while GLX481372 had an IC50 for NOX4 of 0.68 μM, it was similarly potent toward NOX5 (IC50: 0.6 μM) [200]. In contrast, GLX7013114 appeared NOX4 specific (IC50: 0.3 μM) and safeguarded the viability of human islet cells exposed to high glucose [200]. GLX481304 equally inhibits NOX4 and NOX2 with a reported IC50 of 1.25 μM. On the other hand, GLX7013144 boasted selectivity for NOX4 [240]. The study, however, tested relative selectivity of GLX481304 for NOX2/NOX4 over NOX1 only, i.e. not testing NOX5 despite its presence in the vasculature. Nevertheless, GLX481304 could significantly reduce ROS, as well as improve cardiac contractility upon hypoxia/reoxygenation of isolated mouse hearts, in addition to improving measured developed pressure following ischemia-reperfusion injury [240]. Clearly, more studies are required to fully test and characterize these inhibitors and their utility.

8.3

ACD042(Grindelic Acid) and ACD084

ACD042 (also known as grindelic acid) and ACD084 are two plant-derived small molecules that were originally identified from a screen of edible plant metabolites [241]. The study was undertaken succeeding findings that certain plant extracts were capable of suppressing NOX4 expression in isolated cells [242]. Utilizing HEK cells overexpressing NOX4, ACD042 and ACD084 were observed to selectively inhibit NOX4 with IC50 values of 2.06 ± 0.76 μM and 3.08 ± 2.77 μM, respectively, compared to the other isolated compounds that showed inhibitory action against NOX2 or

362

C. M. Dustin et al.

NOX5 [241]. Moreover, ACD042 was found to be capable of directly inhibiting activity in the isolated NOX4 dehydrogenase domain, suggesting that it targets either the NADPH- or FAD-binding domains [241]. To date, there has not been work characterizing these molecules for efficacy in vivo, and their specificity for NOX4 over NOX1 was absent from the screening and analysis. That notwithstanding, the potential for viable NOX4-selective inhibitors derived from edible sources warrants attention.

8.4

Fulvene-5

Fulvene-5 is a fulvene derivate that was originally synthesized and utilized as a NOX inhibitor in a mouse model of endothelial tumor growth following earlier successes with triphenylmethane dyes [243]. The authors demonstrated that Fulvene-5 was capable of inhibiting both NOX2- and NOX4-mediated ROS by 40% in whole cell models, in addition to suppressing both tumor-associated gene expression, as well as tumor growth following injection of cells into mice [243]. However, initial studies did not probe the specificity of Fulvene-5 for NOX2/NOX4, in addition to not assessing the antioxidant capabilities [243]. Further work showed efficacy of Fulvene-5 in inhibition of NOX4mediated tumor responses following radiation [244], inhibition of NOX4-mediated ROS and cardiac arrhythmia in zebrafish models [245], as well as inhibition of NOX4mediated tumor development in a mouse model of ataxia telangiectasia [246]. Despite these findings, Fulvene-5 has not been significantly utilized as a NOX inhibitor aside from these select studies. This is likely due to the aforementioned lack of data regarding isoform inhibition (or lack thereof) outside of NOX2/NOX4, antioxidant properties, or off-target effects.

8.5

NOX4 Inhibitors in the Pipeline

While there have been significant advances in NOX4 inhibitor research, there is still more to be done, with some newer prospects on the horizon. To name a few, a recent study by Xu and colleagues designed a NOX4 inhibitor panel utilizing a pharmacophore-based approach due to the lack of available structural data [247]. Their findings revealed a group of sulfonylurea compounds that did not scavenge ROS and could ostensibly inhibit NOX4-dependent α-smooth muscle actin expression in aortic smooth muscle cells with IC50 values between 0.5 and 27 μM [247]. It remains, however, to be seen how these inhibitors perform in cell-free assays that more directly assess NOX4 activation as well as NOX4 specificity [247]. The use of a protein expression-focused

assay for assessing NOX4 activity is not a direct measure of activity. Recently patented IP introduced a highly selective NOX4 inhibitor intended for use in fibrotic disorders, UANox048, which can significantly inhibit direct NOX4 activity as well as α-smooth muscle actin expression absent of ROS scavenging activity [248]. However, there is no accompanying published data to our knowledge for this molecule, or its related compounds, aside from patent information. Finally, the inhibitor M13 is characterized as a potent and selective NOX4 inhibitor (IC50: 0.01 μM) capable of impeding ischemia-reperfusion-induced hyperpermeability in brain endothelial cells [43].

9

NOX5

NOX5 is unique in its organization as a NOX isoform because it does not require auxiliary NOX subunits. It is activated by Ca2+ binding to its EF hand domains and was originally detected in the testes, lymph nodes, and spleen [249]. Indeed, NOX5 regulation involves interaction of the Ca2+-binding protein calmodulin with a specific calmodulin consensus binding domain on its C-terminus, which sensitizes NOX5 to Ca2+[250]. NOX5 can directly bind heat shock protein 90 (HSP90), which stabilizes NOX5 and positively influences activity [251]. It has been demonstrated that NOX5 activation requires phosphorylation downstream of certain kinases such as PKC [252] or calcium/calmodulindependent protein kinase II (CAMKII) [253]. NOX5 activity can then induce the downstream activation of effector proteins such as extracellular signal-regulated kinase (ERK) [254], feeding into numerous signaling processes. Herein we provide a brief overview for the purposes of discussing NOX5’s therapeutic potential (Fig. 21.6). When it comes to an emphasis on NOXs in disease, NOX5 is arguably the most poorly studied. Nevertheless, multiple models utilizing cell culture and NOX5-expressing mice have demonstrated an impact of NOX5 in disease. For example, NOX5 reportedly plays a role in the pathology of multiple vascular cell types, including the vascular endothelium [254] and vascular smooth muscle [255]. Further, NOX5 is implicated in heart failure and cardiac hypertrophy, whereby induction of NOX5 by ANGII signaling can induce cardiomyocyte hypertrophy in vitro, in addition to showing increased hypertrophic responses in transgenic NOX5 mice subjected to ANGII or pressure overload [256]. NOX5 has been strongly associated with coronary artery disease wherein coronary artery NOX5 expression is elevated in individuals with advanced atherosclerosis and neointimal lesions [257]. In that vein, NOX5 has been shown to be highly upregulated in abdominal aortic aneurism [258] and can contribute to hypertension through increased NOX5

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

363

ML090

? NOX5

?

Calmodulin

Melittin

Ca 2+

Ca 2+

Ca 2+ Ca 2+

GLX481372

HSP90

? Peptides 1 and 3

NA

H DP

VAS2870 VAS3947

D FA

? ?

Celastrol

GKT136901

Setanaxib

Gedunin

DPI

Fig. 21.6 NOX5 inhibitors. A visual representation of the current NOX5 inhibitors and their binding sites. Isoform-specific inhibitors are depicted in black, pan-NOX inhibitors are depicted in red, and inhibitors that target at least 2 isoforms are depicted in green. Arrows

indicate confirmed or predicted sites of interaction while inhibitors listed with a “?” are not characterized as possessing a known or predicted binding site. Created with BioRender.com

expression inducing vascular remodeling and hypercontractility [259], as well as through elicited NOS uncoupling [260, 261]. NOX5 has similarly been shown to play a significant role in diabetic nephropathy through induction of albuminuria and disruption of podocyte barrier integrity [261, 262]; it has been implicated in multiple cancers with evidence pointing toward a role for NOX5 in prostate cancer cell survival and proliferation [263], promoting signaling that can lead to esophageal cancer [264, 265], in addition to being highly expressed in both melanoma and breast cancer [266].

sites [268]. The interaction between the EF hand domain and the C-terminal domain was found to be abrogated by melittin and thus NOX5-mediated superoxide production [268]. Given the nature of melittin as a calmodulin and calmodulin binding site inhibitor, however, there are significant concerns regarding unintended inhibition of other processes that are dependent on calmodulin signaling.

9.1

Peptide Inhibitors of NOX5

Early inhibitors of NOX5 were postulated from peptides utilized to probe the specific binding sites between its N-terminal EF hand domain and C-terminal dehydrogenase domain. Studies revealed that specific sequences of the cytoplasmic dehydrogenase domain (denoted “Peptide 1” and “Peptide 3”) significantly inhibited NOX5-derived superoxide generation by targeting an autoinhibitory region termed the regulatory EF-Hand binding domain (REFBD) [267]. These peptides have not been utilized to our knowledge as inhibitors of NOX5 outside of this study. Another peptide that has shown promise with respect to NOX5 inhibition is melittin, a peptide found in bee venom that has been shown to inhibit calmodulin and bind calmodulin-binding

9.2

ML090

ML090 was initially identified as a NOX1-targeted small molecule alongside ML171. However, it has recently been suggested that ML090 may inhibit NOX5 with greater efficacy (81%, Emax) but with similar potency as with NOX1 and NOX4 with a NOX5 IC50 of 10 nM, vs IC50 values of 25 nM and 20 nM for NOX1 and NOX4 respectively [43]. Indeed, ML090 reduced NOX5-dependent brain endothelial permeability in response to hypoxia/reoxygenation [269]. More recently, ML090 was utilized to inhibit NOX5, and downstream cycloogenase-2 and prostaglandin E2 (PGE2), in aortic endothelial cells [270, 271], however no further studies have demonstrated ML090-mediated inhibition of NOX5 and altered phenotypes in vivo. These results must be assessed with the aforementioned issues regarding ML090 and its potential NOX-independent activities in common assays [39].

364

9.3

C. M. Dustin et al.

Gedunin

The most recently described (as of the writing of this chapter) NOX5 inhibitor is Gedunin, a molecule known to inhibit the critical NOX5 modulator HSP90. Gedunin was modeled to bind with strong affinity to the C-terminal HSP90-binding domain of NOX5 prevent HSP90 binding and destabilize it [272]. Celastrol reportedly binds and inhibits NOX5 in a similar way. Assessing Gedunin in an in vitro model of hyperglycemia, given the known contribution of NOX5 to diabetic complications, the authors demonstrated that Gedunin prevented elevated glucose-induced hemolysis, hemoglobin glycosylation, and lipid peroxidation [272]. Similarly, Gedunin was capable of rescuing glucose-mediated inhibition of red blood cell antioxidant defenses, which the authors reasoned is downstream of NOX5-mediated ROS, or might be the result of an antioxidant effect of Gedunin [272]. Overall, while this study is relatively new, there may be potential in repurposing Gedunin as a NOX5 inhibitor.

10

DUOX1 and DUOX2

The DUOXs, sometimes (albeit rarely) referred to as NOX6 and NOX7, are two NOX family members that deviate from the classical NOX organization significantly. While they contain the classic transmembrane, dual heme catalytic core and dehydrogenase domain, they are similar to NOX5 in that they also have EF hand domains, likewise requiring Ca2+ for proper activation [37]. Additionally, DUOXs have an extracellular peroxidase-like domain, so-named for its sequence homology with peroxidase enzymes [37]. The DUOXs do not require any of the other classical NOX regulatory components to function; however, they do require a specific transmembrane maturation factor, DUOXA1 or DUOXA2, which allows proper plasma membrane translocation and enzyme function [37]. DUOXs are most highly expressed within the thyroid where they play a critical role in regulating thyroid hormone synthesis [37]. Moreover, they are found in other tissues, predominantly the epithelia of the lungs and intestine/colon, with some expression also identified within the skin and various inflammatory/immune cells [273]. Like other members of the NOX family, DUOXs can contribute to various host defense processes when functioning normally, in addition to various wound healing responses [274]. However, there are potential harmful consequences of overabundant DUOX signaling, albeit the precise role for DUOX in disease is still not clear. For instance, DUOXs are highly expressed in airway epithelial cells, and DUOX1, per se, appears to drive inflammatory responses and wound healing in these cells [275]. Somewhat perplexingly, the role of DUOX1 in cancer is contradictory across different cancer types. The evidence predominantly points to a loss of

DUOX1 in lung cancer, leading to more invasive and migratory cellular phenotypes downstream of epithelial to mesenchymal transition [276, 277]. In other cancers, in contrast, DUOX1 is overexpressed [278]. With respect to DUOX1 overactivation, DUOX1 may play a specific role in the pathogenesis and development of allergic asthma, as both acute and chronic allergen exposure induces DUOX1-dependent inflammatory responses, as well as increased DUOX1 expression in chronic contexts [279–284]. DUOX1-deficient mice show tempered allergic inflammation compared to wild type controls, suggesting a role for DUOX1 in the development of type 2 inflammation [285]. However, somewhat confusingly, DUOX1 may also be downregulated in age-related emphysema/chronic obstructive pulmonary disease (COPD) and associated inflammatory responses [286]. DUOX2, on the other hand, plays a different role in most cases. That is, DUOX2 is predominantly overactivated and/or overexpressed in various cancers [278]. Similarly, it has been seen that increased airway inflammation can be downstream of aberrantly activated DUOX2 [287, 288], which can lead to further inflammatory issues, with select evidence pointing toward DUOX2 working in conjunction with inducible NOS to produce reactive nitrogen species and 3-nitrotyrosine and increased inflammation in severe asthmatics [289], as well as playing a key role in the induction of type 1 inflammation [285], rather than type 2. DUOXs, predominantly DUOX2, have also been implicated in renal function and diabetic nephropathy, specifically being involved in mesangial cell-mediated ROS production [290]. DUOXs (predominantly DUOX2) are similarly involved in inflammatory diseases of the gut and bowels, including Crohn’s disease and other inflammatory bowel diseases, however in both cases it appears that the functional outcome is lowered DUOX2 activity, leading to disease progression [291]. Despite the potential contribution of DUOX1 and DUOX2 to various diseases, no inhibitors have been developed to specifically target either isoform (Fig. 21.7). With respect to existing NOX inhibitors, there is relatively little information regarding their ability to inhibit the DUOXs. One exception appears to be ML171 which is capable of strongly inhibiting DUOX1-mediated epithelial responses in vitro and in vivo in a model of acute allergen challenge [273, 279]. Alternatively, it has been demonstrated that VAS2870 can inhibit DUOX1 and DUOX2 following PMA-mediated activation of COS-7 cells transfected with DUOX1/DUOXA1 or DUOX2/ DUOXA2 complexes [292], while also mimicking observed phenotypes in DUOX mutant-derived thyroid hormone deficiencies in a zebrafish model, suggesting that VAS2870 was inducing the phenotype through DUOX1 inhibition [292]. VAS2870 was demonstrated to inhibit DUOX1mediated wound healing, further lending evidence to the use of VAS2870 as a DUOX1 inhibitor [293]. In terms of

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

365

ML171 ?

DUOX1/2

DUOXA1/A2

Ca2+ 2+

Ca

VAS2870

D FA

?

NA

H DP

DPI Fig. 21.7 DUOX1/2 inhibitors. A visual representation of the current DUOX1/2 inhibitors and their binding sites. Pan-NOX inhibitors are depicted in red and inhibitors that target at least 2 isoforms are depicted

potential future inhibitors, recent findings demonstrated that DUOX1 may be inhibited by cysteine alkylation [294], and that the critical association of DUOX1 and DUOXA1 may require intermolecular disulfide bonding [295]. Taking these two findings together, it has been suggested that this binding region may be a potential site for inhibitor targeting. Indeed, a patent has been filed for the specific use of synthetic peptides for the express purpose of blocking this interaction and subsequently inhibiting NOX activity [296], yet it still remains to be seen if this is a viable strategy for DUOX1 inhibition. Owing to the advent of Cryo-EM structures of DUOX1/DUOXA1 complexes [46, 47], as well as the possibility of X-ray crystallography structures on the heels of recent NOX5 structures [45], advances in the development of DUOX inhibitors are expected as new structural data is obtained.

in green. Arrows indicate confirmed or predicted sites of interaction whereas inhibitors listed with a “?” are not characterized as possessing a known or predicted binding site. Created with BioRender.com

11

Conclusions and Perspectives: Is There a Need for Isoform-Specific NOX Inhibitors?

As evidenced by the contents of the previous sections and summarized in Table 21.1, a monumental amount of work has gone into the development of NOX inhibitors over the past 20–30 years, and the push is only gaining momentum as resources and tools become more sophisticated. A formidable hurdle of obtaining detailed structural information of the NOXs is slowly being surmounted and, as a consequence, we are likely to see significant advancements in inhibitor design. Despite continuing advances in this regard, the question remains of whether isoform-specific inhibitors are superior to “multilateral”-NOX targeting or pan-inhibition. Arguing against selective inhibition is therapeutic potential for simultaneous inhibition of multiple NOX isoforms extant in a disease phenotype. When evaluating the different NOXs that contribute to disease, it becomes eminently clear that there are pathologies wherein multiple NOXs are

366

C. M. Dustin et al.

Table 21.1 NOX Inhibitors—Isoform Specificity and Targeted Sites/Interactions

DPI S17834 VAS2870 VAS3947 Celastrol

IC50 (μM) Nox1 Nox2 0.2 0.1 ND ND 10.6 12 2 0.41 0.59

Nox3 ND ND ND ND ND

Nox4 0.1 ND

Nox5 0.02 ND

13 2.79

ND 3.31

APX-115 GKT136901

1.08 0.16 (ki)

ND ND

GKT-831 (Setanaxib)

0.14 (ki)

GKT-771

0.45 (ki) 0.41 (ki) Inactive

Inhibitor

ML171 ML090 NOS31 NoxA1ds NF02 Nox2ds-tat GSK2795039 CYR5099 Phox-I1 Phox-I2 Ebselen/congeners Perhexiline Naxolone CPP11G CPP11H Fulvene-5 GLX351322 GLX481372 GLX7013114 GLX481304 ACD042 (Grindelic acid) ACD084 Melittin Peptide1 Peptide3 ML090 Gedunin

0.063 (ki) 0.25 0.36 2 0.019 16.7 Inactive >100 Inactive ND ND 0.15

0.57 1.53 (ki) 1.75 (ki) Inactive

Inactive

0.63 0.17 (ki) 0.11 (ki) 4.3 (ki)

5 ≥ 10 >40 Inactive Inactive 0.74 40 ND ND ND >100 ND ND ND ND

5 ≥ 10 >28.7 Inactive ND Inactive >100 ND Inactive Inactive Inactive

>40 Inactive ND ND >100 ND ND ND 0.7

ND ND >100 >100 ND ND 7 Inactive Inactive ND

2.3 2.0/2.5 20 32 >5 40 16 Inactive 1.25 >20

ND ND ND ND ND ND 3.2 Inactive Inactive ND

ND ND >100 >100 >5 5 0.68 0.3 1.25 2.06

ND ND >100 >100 ND ND 0.57 Inactive Inactive >20

ND

>5.0

ND

3.08

>5.0 0.1 30 30 0.01 ND

0.025 ND

Inactive ND

ND

ND ND

0.02 ND

ND

Other Enzymes XO, NOS

Targeted Interactions

References

FAD site

[78] [85] [89] [91] [96]

p47phox (NOXO1)p22phox

[94] [125], [127], [128] [126] [130] Xo(5.5) XO (3.5) Nox1-NOXA1 Nox2(B-loop)-p47phox XO (29) Rac1/2-p67phox Rac1/2-p67phox p47phox, p67phox translocation p47phox p47phox-p22phox p47phox-p22phox

EF-hands EF-hands-REFBD EF-hands-REFBD Hsp90 site

[132] [139] [141] [65] [156] [66], [183] [189] [197] [69] [69] [68] [202] [205] [206] [206] [243] [239] [200] [200] [200], [240] [241] [241] [268] [267] [267] [43] [272]

NOX Inhibitors clustered by color and corresponding to their purported NOX target. Gray = Pan inhibitors; Yellow = NOX1; Pink = NOX2; Green = NOX4; Blue = NOX5. ND = Not Determined. XO = Xanthine Oxidase. NOS = Nitric Oxide Synthase. REFBD = Regulatory EF-hand Binding Domain. Hsp90 = Heat Shock Protein 90. Bold numerical values indicate relative selectivity for the corresponding NOX a

upregulated and deemed causal to the phenotype. For example, in the case of diabetes and diabetic nephropathy, there is evidence for the upregulation and injurious effects of NOX4 and NOX5, among other NOXs. Inhibition of only one of these may not lead to effectual treatment if the other NOXs

supplant or supersede participation in the deleterious outcome. In that case, multi- or pan-inhibition may be justified. Careful assessment of the short and long-term repercussions, however, of multilateral inhibition is stochastically more complex than singular inhibition. For one, it is likely not

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

the case that each NOX is causal to the triggering of precisely the same signaling cascade and phenotype in each context. Moreover, the durable inhibition of one or more NOX(s) is predicted to give rise to negative consequences for the patient as the role of each isoform may evolve from deleterious to salutary effects, and vice versa, over time. While this may also be true for selective NOX inhibition, the consequences of singular inhibition are predictably less complex. On the other hand, in some scenarios in which the situation is exceedingly less complex (i.e. NOX isoforms are unidirectional in their negative outcomes) pharmacologists and medicinal chemists may deem multilateral blockade as desirable. Indeed, the argument for selective inhibition is accentuated by an increasing awareness that not all NOXs contribute to a disease phenotype, and that each NOX at various time points in the life of a patient might be expected to play physiological or homeostatic roles [1]. Intuitively, NOX blockade early in the life of a patient or inadvertently in utero, could plausibly bear disastrous effects. One example is the well-described role of NOX4 in differentiation and selfrenewal, as well as the plausible pro-tumorigenic effects of its inhibition [297, 298]. Moreover, in disease, inhibition of NOX2 and NOX4 simultaneously in atherosclerosis may be predictably foreseen as detrimental, or at least a net-negative, as NOX4 has been shown to have antiatherosclerotic effects while NOX2 has primarily been shown to promote atherosclerosis. Similarly, in the case of NOX5 and the DUOXs, they are well known to play a role in sperm motility, and thyroid hormone synthesis, respectively. Therefore, treatment with an inhibitor that concomitantly blocks those NOXs may be detrimental for reproductive and metabolic/immune function. That being said, as with any drug class, titration of drugs to therapeutic levels of singular NOX inhibition can be expected to mitigate these untoward effects and be considerably less problematic from a pharmacological standpoint. Considering the abovementioned challenges, it stands to reason that the most logical path forward from a scientific and drug development perspective is the selective targeting of a particular NOX at a time, be it small molecule or peptidic in nature. It would seem that a concerted focus on one NOX during drug development substantially reduces the likelihood of off-target effects for pharmacologists and medicinal chemists alike. That is, optimization of a compound using structure activity relationships and rational design with considerations for potency, efficacy, toxicity and safety is expected to yield “cleaner” drugs for use in the clinic. That is not to say that, once developed, formulations compounding one or more NOX drugs ought not be considered. In conclusion, expanding momentum toward optimizing drugs for a single NOX is likely to bear the ripest fruit over

367

the long term. Such approaches would enable researchers to tailor formulations toward diseases individually and temporally. With ever-evolving technology available to scientists, the advances in structural information that have been and are yet-to-be made, and the combined knowledge of over 50 years of research, prodigious discoveries regarding NOX inhibition strategies are on the horizon. Furthermore, creative improvements in nanotechnologies and delivery strategies could soon obviate the too often-disparaged druggability of, by nature and design, supremely selective peptide and peptidomimetics. Acknowledgements Funds were provided by NIH Grant Nos. R01HL142248, R01HL079207, R01HL112914, T32GM008424, and American Heart Association 18TPA34170069 (P.J.P.) and T32HL110849 (C.M.D).

References 1. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87(1):245–313. https://doi.org/10.1152/physrev.00044.2005 2. Csányi G, Taylor WR, Pagano PJ (2009) NOX and inflammation in the vascular adventitia. Free Radic Biol Med 47(9):1254–1266. https://doi.org/10.1016/j.freeradbiomed.2009.07.022 3. Cifuentes-Pagano E, Meijles DN, Pagano PJ (2014) The quest for selective nox inhibitors and therapeutics: challenges, triumphs and pitfalls. Antioxid Redox Signal 20(17):2741–2754. https://doi.org/ 10.1089/ars.2013.5620 4. Sies H, Jones DP (2020) Reactive oxygen species (ROS) as pleiotropic physiological signalling agents. Nat Rev Mol Cell Biol 21(7):363–383. https://doi.org/10.1038/s41580-020-0230-3 5. Ghezzi P, Jaquet V, Marcucci F, Schmidt HHHW (2017) The oxidative stress theory of disease: levels of evidence and epistemological aspects. Br J Pharmacol 174(12):1784–1796. https://doi. org/10.1111/bph.13544 6. Cifuentes ME, Pagano PJ (2006) Targeting reactive oxygen species in hypertension. Curr Opin Nephrol Hypertens 15(2):179–186 7. Holmstrom KM, Finkel T (2014) Cellular mechanisms and physiological consequences of redox-dependent signalling. Nat Rev Mol Cell Biol 15(6):411–421. https://doi.org/10.1038/nrm3801 8. Lambeth JD (2004) NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol 4(3):181–189. https://doi.org/10.1038/ nri1312 9. Takac I, Schröder K, Zhang L, Lardy B, Anilkumar N, Lambeth JD et al (2011) The E-loop is involved in hydrogen peroxide formation by the NADPH oxidase Nox4. J Biol Chem 286(15):13304–13313. https://doi.org/10.1074/jbc.M110.192138 10. Nisimoto Y, Diebold BA, Cosentino-Gomes D, Lambeth JD (2014) Nox4: a hydrogen peroxide-generating oxygen sensor. Biochemistry 53(31):5111–5120. https://doi.org/10.1021/bi500331y 11. Al Ghouleh I, Frazziano G, Rodriguez AI, Csányi G, Maniar S, St Croix CM et al (2013) Aquaporin 1, Nox1, and Ask1 mediate oxidant-induced smooth muscle cell hypertrophy. Cardiovasc Res 97(1):134–142. https://doi.org/10.1093/cvr/cvs295 12. Hara-Chikuma M, Chikuma S, Sugiyama Y, Kabashima K, Verkman AS, Inoue S et al (2012) Chemokine-dependent T cell migration requires aquaporin-3-mediated hydrogen peroxide uptake. J Exp Med 209(10):1743–1752. https://doi.org/10.1084/ jem.20112398

368 13. Bienert GP, Schjoerring JK, Jahn TP (2006) Membrane transport of hydrogen peroxide. Biochim Biophys Acta 1758(8):994–1003. https://doi.org/10.1016/j.bbamem.2006.02.015 14. Kuroda J, Ago T, Matsushima S, Zhai P, Schneider MD, Sadoshima J (2010) NADPH oxidase 4 (Nox4) is a major source of oxidative stress in the failing heart. Proc Natl Acad Sci U S A 107(35):15565–15570. https://doi.org/10.1073/pnas.1002178107 15. Laurindo FRM, Araujo TLS, Abrahão TB (2014) Nox NADPH oxidases and the endoplasmic reticulum. Antioxid Redox Signal 20(17):2755–2775. https://doi.org/10.1089/ars.2013.5605 16. Li Q, Harraz MM, Zhou W, Zhang LN, Ding W, Zhang Y et al (2006) Nox2 and Rac1 regulate H2O2-dependent recruitment of TRAF6 to endosomal interleukin-1 receptor complexes. Mol Cell Biol 26(1):140–154. https://doi.org/10.1128/MCB.26.1.140-154. 2006 17. Oakley FD, Abbott D, Li Q, Engelhardt JF (2009) Signaling components of redox active endosomes: the redoxosomes. Antioxid Redox Signal 11(6):1313–1333. https://doi.org/10.1089/ ars.2008.2363 18. Kuroda J, Nakagawa K, Yamasaki T, Nakamura K, Takeya R, Kuribayashi F et al (2005) The superoxide-producing NAD(P)H oxidase Nox4 in the nucleus of human vascular endothelial cells. Genes Cells 10(12):1139–1151. https://doi.org/10.1111/j. 1365-2443.2005.00907.x 19. Geiszt M, Leto TL (2004) The Nox family of NAD(P)H oxidases: host defense and beyond. J Biol Chem 279(50):51715–51718. https://doi.org/10.1074/jbc.R400024200 20. Babior BM, Lambeth JD, Nauseef W (2002) The neutrophil NADPH oxidase. Arch Biochem Biophys 397(2):342–344. https://doi.org/10.1006/abbi.2001.2642 21. Cross AR, Segal AW (2004) The NADPH oxidase of professional phagocytes--prototype of the NOX electron transport chain systems. Biochim Biophys Acta 1657(1):1–22. https://doi.org/10. 1016/j.bbabio.2004.03.008 22. Harper AM, Dunne MJ, Segal AW (1984) Purification of cytochrome b-245 from human neutrophils. Biochem J 219(2): 519–527. https://doi.org/10.1042/bj2190519 23. Segal AW, Jones OT (1978) Novel cytochrome b system in phagocytic vacuoles of human granulocytes. Nature 276(5687):515–517. https://doi.org/10.1038/276515a0 24. El-Benna J, Dang PM-C, Gougerot-Pocidalo M-A, Marie J-C, Braut-Boucher F (2009) p47phox, the phagocyte NADPH oxidase/NOX2 organizer: structure, phosphorylation and implication in diseases. Exp Mol Med 41(4):217–225. https://doi.org/10.3858/ emm.2009.41.4.058 25. Nauseef WM (2004) Assembly of the phagocyte NADPH oxidase. Histochem Cell Biol 122(4):277–291. https://doi.org/10.1007/ s00418-004-0679-8 26. Brown DI, Griendling KK (2009) Nox proteins in signal transduction. Free Radic Biol Med 47(9):1239–1253. https://doi.org/10. 1016/j.freeradbiomed.2009.07.023 27. Suh Y-A, Arnold RS, Lassegue B, Shi J, Xu X, Sorescu D et al (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401(6748):79–82. https://doi.org/10.1038/43459 28. Panday A, Sahoo MK, Osorio D, Batra S (2015) NADPH oxidases: an overview from structure to innate immunity-associated pathologies. Cell Mol Immunol 12(1):5–23. https://doi.org/10. 1038/cmi.2014.89 29. Bánfi B, Clark RA, Steger K, Krause KH (2003) Two novel proteins activate superoxide generation by the NADPH oxidase NOX1. J Biol Chem 278(6):3510–3513. https://doi.org/10.1074/ jbc.C200613200 30. Bánfi B, Malgrange B, Knisz J, Steger K, Dubois-Dauphin M, Krause KH (2004) NOX3, a superoxide-generating NADPH oxidase of the inner ear. J Biol Chem 279(44):46065–46072. https:// doi.org/10.1074/jbc.M403046200

C. M. Dustin et al. 31. Cheng G, Ritsick D, Lambeth JD (2004) Nox3 regulation by NOXO1, p47phox, and p67phox. J Biol Chem 279(33): 34250–34255. https://doi.org/10.1074/jbc.M400660200 32. Ueno N, Takeya R, Miyano K, Kikuchi H, Sumimoto H (2005) The NADPH oxidase Nox3 constitutively produces superoxide in a p22phox-dependent manner: its regulation by oxidase organizers and activators. J Biol Chem 280(24):23328–23339. https://doi.org/ 10.1074/jbc.M414548200 33. Ueyama T, Geiszt M, Leto TL (2006) Involvement of Rac1 in activation of multicomponent Nox1- and Nox3-based NADPH oxidases. Mol Cell Biol 26(6):2160–2174. https://doi.org/10. 1128/mcb.26.6.2160-2174.2006 34. Martyn KD, Frederick LM, von Loehneysen K, Dinauer MC, Knaus UG (2006) Functional analysis of Nox4 reveals unique characteristics compared to other NADPH oxidases. Cell Signal 18(1):69–82. https://doi.org/10.1016/j.cellsig.2005.03.023 35. Matsushima S, Kuroda J, Zhai P, Liu T, Ikeda S, Nagarajan N et al (2016) Tyrosine kinase FYN negatively regulates NOX4 in cardiac remodeling. J Clin Invest 126(9):3403–3416. https://doi.org/10. 1172/JCI85624 36. Fulton DJR (2009) Nox5 and the regulation of cellular function. Antioxid Redox Signal 11(10):2443–2452. https://doi.org/10. 1089/ars.2009.2587 37. Donko A, Peterfi Z, Sum A, Leto T, Geiszt M (2005) Dual oxidases. Philos Trans R Soc Lond Ser B Biol Sci 360(1464): 2301–2308. https://doi.org/10.1098/rstb.2005.1767 38. Doussiere J, Gaillard J, Vignais PV (1999) The Heme component of the neutrophil NADPH oxidase complex is a target for Aryliodonium compounds. Biochemistry 38(12):3694–3703. https://doi.org/10.1021/bi9823481 39. Reis J, Massari M, Marchese S, Ceccon M, Aalbers FS, Corana F et al (2020) A closer look into NADPH oxidase inhibitors: validation and insight into their mechanism of action. Redox Biol 32: 101466. https://doi.org/10.1016/j.redox.2020.101466 40. Ghouleh IA, Sahoo S, Meijles DN, Amaral JH, de Jesus DS, Sembrat J et al (2017) Endothelial Nox1 oxidase assembly in human pulmonary arterial hypertension; driver of Gremlin1mediated proliferation. Clin Sci (Lond) 131(15):2019–2035. https://doi.org/10.1042/cs20160812 41. Augsburger F, Filippova A, Rasti D, Seredenina T, Lam M, Maghzal G et al (2019) Pharmacological characterization of the seven human NOX isoforms and their inhibitors. Redox Biol 26: 101272. https://doi.org/10.1016/j.redox.2019.101272 42. Heumüller S, Wind S, Barbosa-Sicard E, Schmidt HH, Busse R, Schröder K et al (2008) Apocynin is not an inhibitor of vascular NADPH oxidases but an antioxidant. Hypertension 51(2): 211–217. https://doi.org/10.1161/hypertensionaha.107.100214 43. Dao VT, Elbatreek MH, Altenhöfer S, Casas AI, Pachado MP, Neullens CT et al (2020) Isoform-selective NADPH oxidase inhibitor panel for pharmacological target validation. Free Radic Biol Med 148:60–69. https://doi.org/10.1016/j.freeradbiomed.2019. 12.038 44. Andricopulo AD, Montanari CA (2005) Structure-activity relationships for the design of small-molecule inhibitors. Mini Rev Med Chem 5(6):585–593. https://doi.org/10.2174/ 1389557054023224 45. Magnani F, Nenci S, Millana Fananas E, Ceccon M, Romero E, Fraaije MW et al (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci U S A 114(26):6764–6769. https://doi.org/10.1073/pnas.1702293114 46. Sun J (2020) Structures of mouse DUOX1-DUOXA1 provide mechanistic insights into enzyme activation and regulation. Nat Struct Mol Biol 27(11):1086–1093. https://doi.org/10.1038/ s41594-020-0501-x

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

47. Wu J-X, Liu R, Song K, Chen L (2021) Structures of human dual oxidase 1 complex in low-calcium and high-calcium states. Nat Commun 12(1):155. https://doi.org/10.1038/s41467-020-20466-9 48. Jesaitis AJ, Riesselman M, Taylor RM, Brumfield S (2019) Enhanced Immunoaffinity purification of human neutrophil Flavocytochrome B for structure determination by electron microscopy. In: Knaus UG, Leto TL (eds) NADPH oxidases: methods and protocols. Springer, New York, NY, pp 39–59 49. Groemping Y, Lapouge K, Smerdon SJ, Rittinger K (2003) Molecular basis of phosphorylation-induced activation of the NADPH oxidase. Cell 113(3):343–355. https://doi.org/10.1016/s0092-8674 (03)00314-3 50. Honbou K, Minakami R, Yuzawa S, Takeya R, Suzuki NN, Kamakura S et al (2007) Full-length p40phox structure suggests a basis for regulation mechanism of its membrane binding. EMBO J 26(4):1176–1186. https://doi.org/10.1038/sj.emboj.7601561 51. Grizot S, Fieschi F, Dagher MC, Pebay-Peyroula E (2001) The active N-terminal region of p67phox. Structure at 1.8 a resolution and biochemical characterizations of the A128V mutant implicated in chronic granulomatous disease. J Biol Chem 276(24): 21627–21631. https://doi.org/10.1074/jbc.M100893200 52. Kami K, Takeya R, Sumimoto H, Kohda D (2002) Diverse recognition of non-PxxP peptide ligands by the SH3 domains from p67 (phox), Grb2 and Pex13p. EMBO J 21(16):4268–4276. https://doi. org/10.1093/emboj/cdf428 53. Lapouge K, Smith SJ, Walker PA, Gamblin SJ, Smerdon SJ, Rittinger K (2000) Structure of the TPR domain of p67phox in complex with Rac.GTP. Mol Cell 6(4):899–907. https://doi.org/10. 1016/s1097-2765(05)00091-2 54. Ogura K, Nobuhisa I, Yuzawa S, Takeya R, Torikai S, Saikawa K et al (2006) NMR solution structure of the tandem Src homology 3 domains of p47phox complexed with a p22phox-derived prolinerich peptide. J Biol Chem 281(6):3660–3668. https://doi.org/10. 1074/jbc.M505193200 55. Maghzal GJ, Krause K-H, Stocker R, Jaquet V (2012) Detection of reactive oxygen species derived from the family of NOX NADPH oxidases. Free Radic Biol Med 53(10):1903–1918. https://doi.org/ 10.1016/j.freeradbiomed.2012.09.002 56. Zielonka J, Zielonka M, Cheng G, Hardy M, Kalyanaraman B (2019) High-throughput screening of NOX inhibitors. In: Knaus UG, Leto TL (eds) NADPH oxidases: methods and protocols. Springer, New York, NY, pp 429–446 57. Keeney MT, Hoffman EK, Farmer K, Bodle CR, Fazzari M, Zharikov A, Castro SL, Hu X, Mortimer A, Kofler JK, CifuentesPagano E, Pagano PJ, Burton EA, Hastings TG, Greenamyre JT, Di Maio R (2022) NADPH oxidase 2 activity in Parkinson’s disease. Neurobiol Dis 170:105754. https://doi.org/10.1016/j.nbd.2022. 105754 58. Brown TD, Whitehead KA, Mitragotri S (2020) Materials for oral delivery of proteins and peptides. Nat Rev Mater 5(2):127–148. https://doi.org/10.1038/s41578-019-0156-6 59. Bruno BJ, Miller GD, Lim CS (2013) Basics and recent advances in peptide and protein drug delivery. Ther Deliv 4(11):1443–1467. https://doi.org/10.4155/tde.13.104 60. Rotrosen D, Yeung CL, Leto TL, Malech HL, Kwong CH (1992) Cytochrome b558: the flavin-binding component of the phagocyte NADPH oxidase. Science 256(5062):1459–1462. https://doi.org/ 10.1126/science.1318579 61. Doussière J, Vignais PV (1992) Diphenylene iodonium as an inhibitor of the NADPH oxidase complex of bovine neutrophils. Factors controlling the inhibitory potency of diphenylene iodonium in a cell-free system of oxidase activation. Eur J Biochem 208(1): 61–71. https://doi.org/10.1111/j.1432-1033.1992.tb17159.x 62. Tew DG (1993) Inhibition of cytochrome P450 reductase by the diphenyliodonium cation. Kinetic analysis and covalent

369 modifications. Biochemistry 32(38):10209–10215. https://doi.org/ 10.1021/bi00089a042 63. Stuehr DJ, Fasehun OA, Kwon NS, Gross SS, Gonzalez JA, Levi R et al (1991) Inhibition of macrophage and endothelial cell nitric oxide synthase by diphenyleneiodonium and its analogs. FASEB J 5(1):98–103. https://doi.org/10.1096/fasebj.5.1.1703974 64. Leitsch D, Kolarich D, Duchêne M (2010) The flavin inhibitor diphenyleneiodonium renders trichomonas vaginalis resistant to metronidazole, inhibits thioredoxin reductase and flavin reductase, and shuts off hydrogenosomal enzymatic pathways. Mol Biochem Parasitol 171(1):17–24. https://doi.org/10.1016/j.molbiopara.2010. 01.001 65. Ranayhossaini DJ, Rodriguez AI, Sahoo S, Chen BB, Mallampalli RK, Kelley EE et al (2013) Selective recapitulation of conserved and nonconserved regions of putative NOXA1 protein activation domain confers isoform-specific inhibition of Nox1 oxidase and attenuation of endothelial cell migration. J Biol Chem 288(51): 36437–36450. https://doi.org/10.1074/jbc.M113.521344 66. Rey FE, Cifuentes ME, Kiarash A, Quinn MT, Pagano PJ (2001) Novel competitive inhibitor of NAD(P)H oxidase assembly attenuates vascular O(2)(-) and systolic blood pressure in mice. Circ Res 89(5):408–414. https://doi.org/10.1161/hh1701.096037 67. Li Y, Cifuentes-Pagano E, DeVallance ER, de Jesus DS, Sahoo S, Meijles DN et al (2019) NADPH oxidase 2 inhibitors CPP11G and CPP11H attenuate endothelial cell inflammation & vessel dysfunction and restore mouse hind-limb flow. Redox Biol 22:101143. https://doi.org/10.1016/j.redox.2019.101143 68. Smith SME, Min J, Ganesh T, Diebold B, Kawahara T, Zhu Y et al (2012) Ebselen and congeners inhibit NADPH oxidase 2-dependent superoxide generation by interrupting the binding of regulatory subunits. Chem Biol 19(6):752–763. https://doi.org/10. 1016/j.chembiol.2012.04.015 69. Bosco EE, Kumar S, Marchioni F, Biesiada J, Kordos M, Szczur K et al (2012) Rational design of small molecule inhibitors targeting the Rac GTPase-p67(phox) signaling axis in inflammation. Chem Biol 19(2):228–242. https://doi.org/10.1016/j.chembiol.2011. 12.017 70. Bosco EE, Mulloy JC, Zheng Y (2008) Rac1 GTPase: a “Rac” of all trades. Cell Mol Life Sci 66(3):370. https://doi.org/10.1007/ s00018-008-8552-x 71. Fontayne A, Dang PM-C, Gougerot-Pocidalo M-A, El Benna J (2002) Phosphorylation of p47phox sites by PKC α, βII, δ, and ζ: effect on binding to p22phox and on NADPH oxidase activation. Biochemistry 41(24):7743–7750. https://doi.org/10.1021/ bi011953s 72. Streeter J, Schickling BM, Jiang S, Stanic B, Thiel WH, Gakhar L et al (2014) Phosphorylation of Nox1 regulates association with NoxA1 activation domain. Circul Res 115(11):911–918. https:// doi.org/10.1161/CIRCRESAHA.115.304267 73. Kim JS, Diebold BA, Babior BM, Knaus UG, Bokoch GM (2007) Regulation of Nox1 activity via protein kinase A-mediated phosphorylation of NoxA1 and 14-3-3 binding. J Biol Chem 282(48): 34787–34800. https://doi.org/10.1074/jbc.M704754200 74. Lambeth JD, Kawahara T, Diebold B (2007) Regulation of Nox and Duox enzymatic activity and expression. Free Radic Biol Med 43(3):319–331. https://doi.org/10.1016/j.freeradbiomed.2007. 03.028 75. Putney JW, Tomita T (2012) Phospholipase C signaling and calcium influx. Adv Biol Regul 52(1):152–164. https://doi.org/10. 1016/j.advenzreg.2011.09.005 76. Gatto GJ, Ao Z, Kearse MG, Zhou M, Morales CR, Daniels E et al (2013) NADPH oxidase-dependent and -independent mechanisms of reported inhibitors of reactive oxygen generation. J Enzyme Inhib Med Chem 28(1):95–104. https://doi.org/10.3109/ 14756366.2011.636360

370 77. O’Donnell BV, Tew DG, Jones OT, England PJ (1993) Studies on the inhibitory mechanism of iodonium compounds with special reference to neutrophil NADPH oxidase. Biochem J 290(Pt 1): 41–49. https://doi.org/10.1042/bj2900041 78. Cross AR, Jones OT (1986) The effect of the inhibitor diphenylene iodonium on the superoxide-generating system of neutrophils. Specific labelling of a component polypeptide of the oxidase. Biochem J 237(1):111–116. https://doi.org/10.1042/bj2370111 79. Lu J, Risbood P, Kane CT, Hossain MT, Anderson L, Hill K et al (2017) Characterization of potent and selective iodonium-class inhibitors of NADPH oxidases. Biochem Pharmacol 143:25–38. https://doi.org/10.1016/j.bcp.2017.07.007 80. Diatchuk V, Lotan O, Koshkin V, Wikstroem P, Pick E (1997) Inhibition of NADPH oxidase activation by 4-(2-aminoethyl)benzenesulfonyl fluoride and related compounds. J Biol Chem 272(20):13292–13301. https://doi.org/10.1074/jbc.272.20.13292 81. Simons JM, t Hart BA, Ip Vai Ching TRAM, Van Dijk H, Labadie RP. (1990) Metabolic activation of natural phenols into selective oxidative burst agonists by activated human neutrophils. Free Radic Biol Med 8(3):251–258. https://doi.org/10.1016/0891-5849 (90)90070-Y 82. Ximenes VF, Kanegae MPP, Rissato SR, Galhiane MS (2007) The oxidation of apocynin catalyzed by myeloperoxidase: proposal for NADPH oxidase inhibition. Arch Biochem Biophys 457(2): 134–141. https://doi.org/10.1016/j.abb.2006.11.010 83. Cheng G, Salerno JC, Cao Z, Pagano PJ, Lambeth JD (2008) Identification and characterization of VPO1, a new animal hemecontaining peroxidase. Free Radic Biol Med 45(12):1682–1694. https://doi.org/10.1016/j.freeradbiomed.2008.09.009 84. Aldieri E, Riganti C, Polimeni M, Gazzano E, Lussiana C, Campia I et al (2008) Classical inhibitors of NOX NAD(P)H oxidases are not specific. Curr Drug Metab 9(8):686–696. https://doi.org/10. 2174/138920008786049285 85. Cayatte AJ, Rupin A, Oliver-Krasinski J, Maitland K, SansilvestriMorel P, Boussard MF et al (2001) S17834, a new inhibitor of cell adhesion and atherosclerosis that targets nadph oxidase. Arterioscler Thromb Vasc Biol 21(10):1577–1584. https://doi. org/10.1161/hq1001.096723 86. Li Y, Xu S, Mihaylova MM, Zheng B, Hou X, Jiang B et al (2011) AMPK phosphorylates and inhibits SREBP activity to attenuate hepatic steatosis and atherosclerosis in diet-induced insulin-resistant mice. Cell Metab 13(4):376–388. https://doi.org/10.1016/j. cmet.2011.03.009 87. Zang M, Xu S, Maitland-Toolan KA, Zuccollo A, Hou X, Jiang B et al (2006) Polyphenols stimulate AMP-activated protein kinase, lower lipids, and inhibit accelerated atherosclerosis in diabetic LDL receptor–deficient mice. Diabetes 55(8):2180–2191. https:// doi.org/10.2337/db05-1188 88. Altenhöfer S, Kleikers PWM, Radermacher KA, Scheurer P, Rob Hermans JJ, Schiffers P et al (2012) The NOX toolbox: validating the role of NADPH oxidases in physiology and disease. Cellular and molecular life sciences: CMLS 69(14):2327–2343. https://doi. org/10.1007/s00018-012-1010-9 89. ten Freyhaus H, Huntgeburth M, Wingler K, Schnitker J, Bäumer AT, Vantler M et al (2006) Novel Nox inhibitor VAS2870 attenuates PDGF-dependent smooth muscle cell chemotaxis, but not proliferation. Cardiovasc Res 71(2):331–341. https://doi.org/ 10.1016/j.cardiores.2006.01.022 90. Tegtmeier F (2005, 24 Nov) Compounds containing a n-HETEROARYL moiety linked to fused ring moieties for the inhibition of NAD(p)H oxidases and platelet activation. WO2005111041 91. Wind S, Beuerlein K, Eucker T, Müller H, Scheurer P, Armitage ME et al (2010) Comparative pharmacology of chemically distinct NADPH oxidase inhibitors. Br J Pharmacol 161(4):885–898. https://doi.org/10.1111/j.1476-5381.2010.00920.x

C. M. Dustin et al. 92. El Dor M, Dakik H, Polomski M, Haudebourg E, Brachet M, Gouilleux F et al (2020) VAS3947 induces UPR-mediated apoptosis through cysteine thiol alkylation in AML cell lines. Int J Mol Sci 21(15):5470. https://doi.org/10.3390/ijms21155470 93. Sun Q-A, Hess DT, Wang B, Miyagi M, Stamler JS (2012) Off-target thiol alkylation by the NADPH oxidase inhibitor 3-benzyl-7-(2-benzoxazolyl)thio-1,2,3-triazolo[4,5-d]pyrimidine (VAS2870). Free Radic Biol Med 52(9):1897–1902. https://doi. org/10.1016/j.freeradbiomed.2012.02.046 94. Joo JH, Huh JE, Lee JH, Park DR, Lee Y, Lee SG et al (2016) A novel pyrazole derivative protects from ovariectomy-induced osteoporosis through the inhibition of NADPH oxidase. Sci Rep 6: 22389. https://doi.org/10.1038/srep22389 95. Lee ES, Kim HM, Lee SH, Ha KB, Bae YS, Lee SJ et al (2020) APX-115, a pan-NADPH oxidase inhibitor, protects development of diabetic nephropathy in podocyte specific NOX5 transgenic mice. Free Radic Biol Med 161:92–101. https://doi.org/10.1016/j. freeradbiomed.2020.09.024 96. Jaquet V, Marcoux J, Forest E, Leidal KG, McCormick S, Westermaier Y et al (2011) NADPH oxidase (NOX) isoforms are inhibited by celastrol with a dual mode of action. Br J Pharmacol 164(2b):507–520. https://doi.org/10.1111/j.1476-5381.2011. 01439.x 97. Weaver CJ, Terzi A, Roeder H, Gurol T, Deng Q, Leung YF et al (2018) nox2/cybb deficiency affects zebrafish Retinotectal connectivity. J Neurosci 38(26):5854–5871. https://doi.org/10.1523/ jneurosci.1483-16.2018 98. Liu H, Wang L, Pan Y, Wang X, Ding Y, Zhou C et al (2020) Celastrol alleviates aortic valve calcification via inhibition of NADPH oxidase 2 in Valvular interstitial cells. JACC: basic to translational. Science 5(1):35–49. https://doi.org/10.1016/j.jacbts. 2019.10.004 99. Altenhofer S, Radermacher KA, Kleikers PW, Wingler K, Schmidt HH (2015) Evolution of NADPH oxidase inhibitors: selectivity and mechanisms for target engagement. Antioxid Redox Signal 23(5):406–427. https://doi.org/10.1089/ars.2013.5814 100. Xu Q, Choksi S, Qu J, Jang J, Choe M, Banfi B et al (2016) NADPH oxidases are essential for macrophage differentiation. J Biol Chem 291(38):20030–20041. https://doi.org/10.1074/jbc. M116.731216 101. Maitra U, Singh N, Gan L, Ringwood L, Li L (2009) IRAK-1 contributes to lipopolysaccharide-induced reactive oxygen species generation in macrophages by inducing NOX-1 transcription and Rac1 activation and suppressing the expression of antioxidative enzymes. J Biol Chem 284(51):35403–35411. https://doi.org/10. 1074/jbc.M109.059501 102. Kikuchi H, Hikage M, Miyashita H, Fukumoto M (2000) NADPH oxidase subunit, gp91(phox) homologue, preferentially expressed in human colon epithelial cells. Gene 254(1–2):237–243. https:// doi.org/10.1016/s0378-1119(00)00258-4 103. Laurent E, McCoy JW 3rd, Macina RA, Liu W, Cheng G, Robine S et al (2008) Nox1 is over-expressed in human colon cancers and correlates with activating mutations in K-Ras. Int J Cancer 123(1): 100–107. https://doi.org/10.1002/ijc.23423 104. Szanto I, Rubbia-Brandt L, Kiss P, Steger K, Banfi B, Kovari E et al (2005) Expression of NOX1, a superoxide-generating NADPH oxidase, in colon cancer and inflammatory bowel disease. J Pathol 207(2):164–176. https://doi.org/10.1002/path.1824 105. Hayes P, Dhillon S, O’Neill K, Thoeni C, Hui KY, Elkadri A et al (2015) Defects in nicotinamide-adenine dinucleotide phosphate oxidase genes NOX1 and DUOX2 in very early onset inflammatory bowel disease. Cell Mol Gastroenterol Hepatol 1(5):489–502. https://doi.org/10.1016/j.jcmgh.2015.06.005 106. Tréton X, Pedruzzi E, Guichard C, Ladeiro Y, Sedghi S, Vallée M et al (2014) Combined NADPH oxidase 1 and interleukin 10 deficiency induces chronic endoplasmic reticulum stress and causes

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

ulcerative colitis-like disease in mice. PLoS One 9(7):e101669. https://doi.org/10.1371/journal.pone.0101669 107. Lassègue B, Sorescu D, Szöcs K, Yin Q, Akers M, Zhang Y et al (2001) Novel gp91(phox) homologues in vascular smooth muscle cells: nox1 mediates angiotensin II-induced superoxide formation and redox-sensitive signaling pathways. Circ Res 88(9):888–894. https://doi.org/10.1161/hh0901.090299 108. Hilenski LL, Clempus RE, Quinn MT, Lambeth JD, Griendling KK (2004) Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 24(4):677–683. https://doi.org/10.1161/01.ATV.0000112024. 13727.2c 109. Dikalova AE, Góngora MC, Harrison DG, Lambeth JD, Dikalov S, Griendling KK (2010) Upregulation of Nox1 in vascular smooth muscle leads to impaired endothelium-dependent relaxation via eNOS uncoupling. Am J Physiol Heart Circ Physiol 299(3): H673–H6H9. https://doi.org/10.1152/ajpheart.00242.2010 110. Dikalova A, Clempus R, Lassègue B, Cheng G, McCoy J, Dikalov S et al (2005) Nox1 overexpression potentiates angiotensin II-induced hypertension and vascular smooth muscle hypertrophy in transgenic mice. Circulation 112(17):2668–2676. https://doi. org/10.1161/circulationaha.105.538934 111. Matsuno K, Yamada H, Iwata K, Jin D, Katsuyama M, Matsuki M et al (2005) Nox1 is involved in angiotensin II-mediated hypertension: a study in Nox1-deficient mice. Circulation 112(17): 2677–2685. https://doi.org/10.1161/circulationaha.105.573709 112. de Jesus DS, DeVallance E, Li Y, Falabella M, Guimaraes D, Shiva S et al (2019) Nox1/Ref-1-mediated activation of CREB promotes Gremlin1-driven endothelial cell proliferation and migration. Redox Biol 22:101138. https://doi.org/10.1016/j.redox.2019. 101138 113. Veit F, Pak O, Egemnazarov B, Roth M, Kosanovic D, Seimetz M et al (2013) Function of NADPH oxidase 1 in pulmonary arterial smooth muscle cells after monocrotaline-induced pulmonary vascular remodeling. Antioxid Redox Signal 19(18):2213–2231. https://doi.org/10.1089/ars.2012.4904 114. Sheehan AL, Carrell S, Johnson B, Stanic B, Banfi B, Miller FJ Jr (2011) Role for Nox1 NADPH oxidase in atherosclerosis. Atherosclerosis 216(2):321–326. https://doi.org/10.1016/j.atherosclero sis.2011.02.028 115. Szöcs K, Lassègue B, Sorescu D, Hilenski LL, Valppu L, Couse TL et al (2002) Upregulation of Nox-based NAD(P)H oxidases in restenosis after carotid injury. Arterioscler Thromb Vasc Biol 22(1):21–27. https://doi.org/10.1161/hq0102.102189 116. Garrido-Urbani S, Jemelin S, Deffert C, Carnesecchi S, Basset O, Szyndralewiez C et al (2011) Targeting vascular NADPH oxidase 1 blocks tumor angiogenesis through a PPARα mediated mechanism. PLoS One 6(2):e14665-e. https://doi.org/10.1371/journal. pone.0014665 117. Wilkinson-Berka JL, Deliyanti D, Rana I, Miller AG, Agrotis A, Armani R et al (2014) NADPH oxidase, NOX1, mediates vascular injury in ischemic retinopathy. Antioxid Redox Signal 20(17): 2726–2740. https://doi.org/10.1089/ars.2013.5357 118. Liang S, Ma H-Y, Zhong Z, Dhar D, Liu X, Xu J et al (2019) NADPH oxidase 1 in liver macrophages promotes inflammation and tumor development in mice. Gastroenterology 156(4): 1156–72.e6. https://doi.org/10.1053/j.gastro.2018.11.019 119. Lan T, Kisseleva T, Brenner DA (2015) Deficiency of NOX1 or NOX4 prevents liver inflammation and fibrosis in mice through inhibition of hepatic stellate cell activation. PLoS One 10(7): e0129743-e. https://doi.org/10.1371/journal.pone.0129743 120. Carnesecchi S, Dunand-Sauthier I, Zanetti F, Singovski G, Deffert C, Donati Y et al (2014) NOX1 is responsible for cell death through STAT3 activation in hyperoxia and is associated with the pathogenesis of acute respiratory distress syndrome. Int J Clin Exp Pathol 7(2):537–551

371

121. Carnesecchi S, Deffert C, Pagano A, Garrido-Urbani S, MétraillerRuchonnet I, Schäppi M et al (2009) NADPH oxidase-1 plays a crucial role in hyperoxia-induced acute lung injury in mice. Am J Respir Crit Care Med 180(10):972–981. https://doi.org/10.1164/ rccm.200902-0296OC 122. Gray SP, Di Marco E, Okabe J, Szyndralewiez C, Heitz F, Montezano AC et al (2013) NADPH oxidase 1 plays a key role in diabetes mellitus-accelerated atherosclerosis. Circulation 127(18):1888–1902. https://doi.org/10.1161/circulationaha.112. 132159 123. Choi D-H, Cristóvão AC, Guhathakurta S, Lee J, Joh TH, Beal MF et al (2012) NADPH oxidase 1-mediated oxidative stress leads to dopamine neuron death in Parkinson’s disease. Antioxid Redox Signal 16(10):1033–1045. https://doi.org/10.1089/ars.2011.3960 124. Cristóvão AC, Guhathakurta S, Bok E, Je G, Yoo SD, Choi DH et al (2012) NADPH oxidase 1 mediates α-synucleinopathy in Parkinson’s disease. J Neurosci 32(42):14465–14477. https://doi. org/10.1523/jneurosci.2246-12.2012 125. Bt L, Gaggini F, Orchard M, Fioraso-Cartier L, Cagnon L, Houngninou-Molango S et al (2010) First in class, potent, and orally bioavailable NADPH oxidase isoform 4 (Nox4) inhibitors for the treatment of idiopathic pulmonary fibrosis. J Med Chem 53(21):7715–7730. https://doi.org/10.1021/jm100773e 126. Aoyama T, Paik Y-H, Watanabe S, Laleu B, Gaggini F, FiorasoCartier L et al (2012) Nicotinamide adenine dinucleotide phosphate oxidase in experimental liver fibrosis: GKT137831 as a novel potential therapeutic agent. Hepatology (Baltimore, Md) 56(6): 2316–2327. https://doi.org/10.1002/hep.25938 127. Schildknecht S, Weber A, Gerding HR, Pape R, Robotta M, Drescher M et al (2014) The NOX1/4 inhibitor GKT136901 as selective and direct scavenger of peroxynitrite. Curr Med Chem 21(3):365–376. https://doi.org/10.2174/09298673113209990179 128. Musset B, Clark RA, DeCoursey TE, Petheo GL, Geiszt M, Chen Y et al (2012) NOX5 in human spermatozoa: expression, function, and regulation. J Biol Chem 287(12):9376–9388. https://doi.org/ 10.1074/jbc.M111.314955 129. De Livera AM, Reutens A, Cooper M, Thomas M, JandeleitDahm K, Shaw JE et al (2020) Evaluating the efficacy and safety of GKT137831 in adults with type 1 diabetes and persistently elevated urinary albumin excretion: a statistical analysis plan. Trials 21(1):459. https://doi.org/10.1186/s13063-020-04404-0 130. Vandierendonck A, Degroote H, Vanderborght B, Verhelst X, Geerts A, Devisscher L et al (2021) NOX1 inhibition attenuates the development of a pro-tumorigenic environment in experimental hepatocellular carcinoma. J Exp Clin Cancer Res 40(1):40. https:// doi.org/10.1186/s13046-021-01837-6 131. Kovacs L, Bruder-Nascimento T, Greene L, Kennard S, Belin de Chantemèle EJ (2021) Chronic exposure to HIV-derived protein tat impairs endothelial function via indirect alteration in fat mass and Nox1-mediated mechanisms in mice. Int J Mol Sci 22(20):10977. https://doi.org/10.3390/ijms222010977 132. Gianni D, Taulet N, Zhang H, DerMardirossian C, Kister J, Martinez L et al (2010) A novel and specific NADPH oxidase-1 (Nox1) small-molecule inhibitor blocks the formation of functional invadopodia in human colon cancer cells. ACS Chem Biol 5(10): 981–993. https://doi.org/10.1021/cb100219n 133. Walsh TG, Berndt MC, Carrim N, Cowman J, Kenny D, Metharom P (2014) The role of Nox1 and Nox2 in GPVI-dependent platelet activation and thrombus formation. Redox Biol 2:178–186. https:// doi.org/10.1016/j.redox.2013.12.023 134. Vara D, Campanella M, Pula G (2013) The novel NOX inhibitor 2-acetylphenothiazine impairs collagen-dependent thrombus formation in a GPVI-dependent manner. Br J Pharmacol 168(1): 212–224. https://doi.org/10.1111/j.1476-5381.2012.02130.x 135. Gaspar RS, Sage T, Little G, Kriek N, Pula G, Gibbins JM (2021) Protein Disulphide isomerase and NADPH oxidase 1 cooperate to

372 control platelet function and are associated with Cardiometabolic disease risk factors. Antioxidants (Basel) 10(3):497. https://doi. org/10.3390/antiox10030497 136. Gaspar RS, Ferreira PM, Mitchell JL, Pula G, Gibbins JM (2021) Platelet-derived extracellular vesicles express NADPH oxidase-1 (Nox-1), generate superoxide and modulate platelet function. Free Radic Biol Med 165:395–400. https://doi.org/10.1016/j. freeradbiomed.2021.01.051 137. Martínez-Revelles S, Avendaño MS, García-Redondo AB, Alvarez Y, Aguado A, Pérez-Girón JV et al (2013) Reciprocal relationship between reactive oxygen species and cyclooxygenase-2 and vascular dysfunction in hypertension. Antioxid Redox Signal 18(1):51–65. https://doi.org/10.1089/ars. 2011.4335 138. Seredenina T, Chiriano G, Filippova A, Nayernia Z, Mahiout Z, Fioraso-Cartier L et al (2015) A subset of N-substituted phenothiazines inhibits NADPH oxidases. Free Radic Biol Med 86:239–249. https://doi.org/10.1016/j.freeradbiomed.2015.05.023 139. Brown SG, Gianni D, Bokoch G, Mercer BA, Hodder P, Rosen HR (2009) Probe report for NOX1 inhibitors. https://www.ncbi.nlm. nih.gov/books/NBK47342/. Access 140. Smith RM, Kruzliak P, Adamcikova Z, Zulli A (2015) Role of Nox inhibitors plumbagin, ML090 and gp91ds-tat peptide on homocysteine thiolactone induced blood vessel dysfunction. Clin Exp Pharmacol Physiol 42(8):860–864. https://doi.org/10.1111/ 1440-1681.12427 141. Yamamoto T, Nakano H, Shiomi K, Wanibuchi K, Masui H, Takahashi T et al (2018) Identification and characterization of a novel NADPH oxidase 1 (Nox1) inhibitor that suppresses proliferation of colon and Stomach cancer cells. Biol Pharm Bull 41(3): 419–426. https://doi.org/10.1248/bpb.b17-00804 142. Dahan I, Molshanski-Mor S, Pick E (2012) Inhibition of NADPH oxidase activation by peptides mapping within the dehydrogenase region of Nox2-a "peptide walking" study. J Leukoc Biol 91(3): 501–515. https://doi.org/10.1189/jlb.1011507 143. Dahan I, Issaeva I, Gorzalczany Y, Sigal N, Hirshberg M, Pick E (2002) Mapping of functional domains in the p22(phox) subunit of flavocytochrome b(559) participating in the assembly of the NADPH oxidase complex by “peptide walking”. J Biol Chem 277(10):8421–8432. https://doi.org/10.1074/jbc.M109778200 144. DeLeo FR, Yu L, Burritt JB, Loetterle LR, Bond CW, Jesaitis AJ et al (1995) Mapping sites of interaction of p47-phox and flavocytochrome b with random-sequence peptide phage display libraries. Proc Natl Acad Sci 92(15):7110. https://doi.org/10.1073/ pnas.92.15.7110 145. Meijles DN, Sahoo S, Al Ghouleh I, Amaral JH, Bienes-MartinezR, Knupp HE et al (2017) The matricellular protein TSP1 promotes human and mouse endothelial cell senescence through CD47 and Nox1. Sci Signal 10(501):eaaj1784. https://doi.org/10.1126/ scisignal.aaj1784 146. Li Y, Kračun D, Dustin CM, El Massry M, Yuan S, Goossen CJ et al (2021) Forestalling age-impaired angiogenesis and blood flow by targeting NOX: interplay of NOX1, IL-6, and SASP in propagating cell senescence. Proc Natl Acad Sci U S A 118(42): e2015666118. https://doi.org/10.1073/pnas.2015666118 147. Rodríguez AI, Csányi G, Ranayhossaini DJ, Feck DM, Blose KJ, Assatourian L et al (2015) MEF2B-Nox1 signaling is critical for stretch-induced phenotypic modulation of vascular smooth muscle cells. Atertio Thromb Vasc Biol 35(2):430–438. https://doi.org/10. 1161/ATVBAHA.114.304936 148. Camargo LL, Harvey AP, Rios FJ, Tsiropoulou S, Da Silva RNO, Cao Z et al (2018) Vascular Nox (NADPH oxidase) compartmentalization, protein Hyperoxidation, and endoplasmic reticulum stress response in hypertension. Hypertension 72(1):235–246. https://doi.org/10.1161/hypertensionaha.118.10824

C. M. Dustin et al. 149. Neves KB, Rios FJ, van der Mey L, Alves-Lopes R, Cameron AC, Volpe M et al (2018) VEGFR (vascular endothelial growth factor receptor) inhibition induces cardiovascular damage via redoxsensitive processes. Hypertension 71(4):638–647. https://doi.org/ 10.1161/hypertensionaha.117.10490 150. Fernandes DC, Wosniak J Jr, Gonçalves RC, Tanaka LY, Fernandes CG, Zanatta DB et al (2021) PDIA1 acts as master organizer of NOX1/NOX4 balance and phenotype response in vascular smooth muscle. Free Radic Biol Med 162:603–614. https://doi.org/10.1016/j.freeradbiomed.2020.11.020 151. Daneva Z, Marziano C, Ottolini M, Chen Y-L, Baker TM, Kuppusamy M et al (2021) Caveolar peroxynitrite formation impairs endothelial TRPV4 channels and elevates pulmonary arterial pressure in pulmonary hypertension. Proc Natl Acad Sci U S A 118(17):e2023130118. https://doi.org/10.1073/pnas.2023130118 152. Muñoz M, López-Oliva ME, Rodríguez C, Martínez MP, SáenzMedina J, Sánchez A et al (2020) Differential contribution of Nox1, Nox2 and Nox4 to kidney vascular oxidative stress and endothelial dysfunction in obesity. Redox Biol 28:101330. https://doi.org/10.1016/j.redox.2019.101330 153. Ottolini M, Hong K, Cope EL, Daneva Z, DeLalio LJ, Sokolowski JD et al (2020) Local Peroxynitrite impairs endothelial transient receptor potential Vanilloid 4 channels and elevates blood pressure in obesity. Circulation 141(16):1318–1333. https://doi.org/10. 1161/circulationaha.119.043385 154. Chu F-F, Esworthy RS, Shen B, Gao Q, Doroshow JH (2019) Dexamethasone and Tofacitinib suppress NADPH oxidase expression and alleviate very-early-onset ileocolitis in mice deficient in GSH peroxidase 1 and 2. Life Sci 239:116884. https://doi.org/10. 1016/j.lfs.2019.116884 155. De Bessa TC, Pagano A, Moretti AIS, Oliveira PVS, Mendonça SA, Kovacic H et al (2019) Subverted regulation of Nox1 NADPH oxidase-dependent oxidant generation by protein disulfide isomerase A1 in colon carcinoma cells with overactivated KRas. Cell Death Dis 10(2):143. https://doi.org/10.1038/s41419-019-1402-y 156. Mousslim M, Pagano A, Andreotti N, Garrouste F, Thuault S, Peyrot V et al (2017) Peptide screen identifies a new NADPH oxidase inhibitor: impact on cell migration and invasion. Eur J Pharmacol 794:162–172. https://doi.org/10.1016/j.ejphar.2016. 10.011 157. Chocry M, Leloup L (2020) The NADPH oxidase family and its inhibitors. Antioxid Redox Signal 33(5):332–353. https://doi.org/ 10.1089/ars.2019.7915 158. Kuhns DB, Alvord WG, Heller T, Feld JJ, Pike KM, Marciano BE et al (2010) Residual NADPH oxidase and survival in chronic granulomatous disease. New England J Med 363(27):2600–2610. https://doi.org/10.1056/NEJMoa1007097 159. Konior A, Schramm A, Czesnikiewicz-Guzik M, Guzik TJ (2014) NADPH oxidases in vascular pathology. Antioxid Redox Signal 20(17):2794–2814. https://doi.org/10.1089/ars.2013.5607 160. Murdoch CE, Alom-Ruiz SP, Wang M, Zhang M, Walker S, Yu B et al (2011) Role of endothelial Nox2 NADPH oxidase in angiotensin II-induced hypertension and vasomotor dysfunction. Basic Res Cardiol 106(4):527–538. https://doi.org/10.1007/s00395-0110179-7 161. Liu J, Yang F, Yang XP, Jankowski M, Pagano PJ (2003) NAD (P)H oxidase mediates angiotensin II-induced vascular macrophage infiltration and medial hypertrophy. Arterioscler Thromb Vasc Biol 23(5):776–782. https://doi.org/10.1161/01.Atv. 0000066684.37829.16 162. Liu J, Ormsby A, Oja-Tebbe N, Pagano PJ (2004) Gene transfer of NAD(P)H oxidase inhibitor to the vascular adventitia attenuates medial smooth muscle hypertrophy. Circ Res 95(6):587–594. https://doi.org/10.1161/01.RES.0000142317.88591.e6 163. Jung O, Schreiber JG, Geiger H, Pedrazzini T, Busse R, Brandes RP (2004) gp91phox-containing NADPH oxidase mediates

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

endothelial dysfunction in renovascular hypertension. Circulation 109(14):1795–1801. https://doi.org/10.1161/01.Cir.0000124223. 00113.A4 164. Sorescu D, Weiss D, Lassègue B, Clempus RE, Szöcs K, Sorescu GP et al (2002) Superoxide production and expression of nox family proteins in human atherosclerosis. Circulation 105(12): 1429–1435. https://doi.org/10.1161/01.cir.0000012917.74432.66 165. Quesada IM, Lucero A, Amaya C, Meijles DN, Cifuentes ME, Pagano PJ et al (2015) Selective inactivation of NADPH oxidase 2 causes regression of vascularization and the size and stability of atherosclerotic plaques. Atherosclerosis 242(2):469–475. https:// doi.org/10.1016/j.atherosclerosis.2015.08.011 166. Judkins CP, Diep H, Broughton BRS, Mast AE, Hooker EU, Miller AA et al (2010) Direct evidence of a role for Nox2 in superoxide production, reduced nitric oxide bioavailability, and early atherosclerotic plaque formation in ApoE-/- mice. Am J Physiol Heart Circ Physiol 298(1):H24–H32. https://doi.org/10.1152/ajpheart. 00799.2009 167. Douglas G, Bendall JK, Crabtree MJ, Tatham AL, Carter EE, Hale AB et al (2012) Endothelial-specific Nox2 overexpression increases vascular superoxide and macrophage recruitment in ApoE-/- mice. Cardiovasc Res 94(1):20–29. https://doi.org/10. 1093/cvr/cvs026 168. Braunersreuther V, Montecucco F, Ashri M, Pelli G, Galan K, Frias M et al (2013) Role of NADPH oxidase isoforms NOX1, NOX2 and NOX4 in myocardial ischemia/reperfusion injury. J Mol Cell Cardiol 64:99–107. https://doi.org/10.1016/j.yjmcc. 2013.09.007 169. Looi YH, Grieve DJ, Siva A, Walker SJ, Anilkumar N, Cave AC et al (2008) Involvement of Nox2 NADPH oxidase in adverse cardiac remodeling after myocardial infarction. Hypertension 51(2):319–325. https://doi.org/10.1161/hypertensionaha.107. 101980 170. Chen H, Song YS, Chan PH (2009) Inhibition of NADPH oxidase is neuroprotective after ischemia-reperfusion. Journal of cerebral blood flow and metabolism: official journal of the International Society of Cerebral Blood Flow and Metabolism. 29(7): 1262–1272. https://doi.org/10.1038/jcbfm.2009.47 171. Brait VH, Jackman KA, Walduck AK, Selemidis S, Diep H, Mast AE et al (2010) Mechanisms contributing to cerebral infarct size after stroke: gender, reperfusion, T lymphocytes, and Nox2derived superoxide. J Cereb Blood Flow Metab: Official Journal of the International Society of Cerebral Blood Flow and Metabolism 30(7):1306–1317. https://doi.org/10.1038/jcbfm.2010.14 172. De Silva TM, Brait VH, Drummond GR, Sobey CG, Miller AA (2011) Nox2 oxidase activity accounts for the oxidative stress and vasomotor dysfunction in mouse cerebral arteries following ischemic stroke. PLoS One 6(12):e28393. https://doi.org/10.1371/ journal.pone.0028393 173. Ansari MA, Scheff SW (2011) NADPH-oxidase activation and cognition in Alzheimer disease progression. Free Radic Biol Med 51(1):171–178. https://doi.org/10.1016/j.freeradbiomed.2011. 03.025 174. Park L, Anrather J, Zhou P, Frys K, Pitstick R, Younkin S et al (2005) NADPH-oxidase-derived reactive oxygen species mediate the cerebrovascular dysfunction induced by the amyloid beta peptide. J Neurosci 25(7):1769–1777. https://doi.org/10.1523/ JNEUROSCI.5207-04.2005 175. Park L, Anrather J, Girouard H, Zhou P, Iadecola C (2007) Nox2derived reactive oxygen species mediate neurovascular dysregulation in the aging mouse brain. J Cereb Blood Flow Metab 27(12):1908–1918. https://doi.org/10.1038/sj.jcbfm. 9600491 176. Di Maio R, Hoffman EK, Rocha EM, Keeney MT, Sanders LH, De Miranda BR et al (2018) LRRK2 activation in idiopathic

373

Parkinson’s disease. Sci Transl Med 10(451):eaar5429. https:// doi.org/10.1126/scitranslmed.aar5429 177. Levesque S, Wilson B, Gregoria V, Thorpe LB, Dallas S, Polikov VS et al (2010) Reactive microgliosis: extracellular micro-calpain and microglia-mediated dopaminergic neurotoxicity. Brain: A J Neurol 133(Pt 3):808–821. https://doi.org/10.1093/brain/awp333 178. Trocme C, Deffert C, Cachat J, Donati Y, Tissot C, Papacatzis S et al (2015) Macrophage-specific NOX2 contributes to the development of lung emphysema through modulation of SIRT1/MMP-9 pathways. J Pathol 235(1):65–78. https://doi.org/10.1002/path. 4423 179. Liu JQ, Zelko IN, Erbynn EM, Sham JSK, Folz RJ (2006) Hypoxic pulmonary hypertension: role of superoxide and NADPH oxidase (gp91phox). Am J Phys Lung Cell Mol Phys 290(1):L2–L10. https://doi.org/10.1152/ajplung.00135.2005 180. Adesina SE, Kang B-Y, Bijli KM, Ma J, Cheng J, Murphy TC et al (2015) Targeting mitochondrial reactive oxygen species to modulate hypoxia-induced pulmonary hypertension. Free Radic Biol Med 87:36–47. https://doi.org/10.1016/j.freeradbiomed.2015. 05.042 181. Fresquet F, Pourageaud F, Leblais V, Brandes RP, Savineau JP, Marthan R et al (2006) Role of reactive oxygen species and gp91phox in endothelial dysfunction of pulmonary arteries induced by chronic hypoxia. Br J Pharmacol 148(5):714–723. https://doi. org/10.1038/sj.bjp.0706779 182. Mittal M, Roth M, König P, Hofmann S, Dony E, Goyal P et al (2007) Hypoxia-dependent regulation of nonphagocytic NADPH oxidase subunit NOX4 in the pulmonary vasculature. Circ Res 101(3):258–267. https://doi.org/10.1161/circresaha.107.148015 183. Csányi G, Cifuentes-Pagano E, Al Ghouleh I, Ranayhossaini DJ, Egaña L, Lopes LR et al (2011) Nox2 B-loop peptide, Nox2ds, specifically inhibits the NADPH oxidase Nox2. Free Radic Biol Med 51(6):1116–1125. https://doi.org/10.1016/j.freeradbiomed. 2011.04.025 184. Jacobson GM, Dourron HM, Liu J, Carretero OA, Reddy DJ, Andrzejewski T et al (2003) Novel NAD (P) H oxidase inhibitor suppresses angioplasty-induced superoxide and neointimal hyperplasia of rat carotid artery. Circ Res 92(6):637–643 185. Sukumar P, Viswambharan H, Imrie H, Cubbon RM, Yuldasheva N, Gage M et al (2013) Nox2 NADPH oxidase has a critical role in insulin resistance-related endothelial cell dysfunction. Diabetes 62(6):2130–2134. https://doi.org/10.2337/ db12-1294 186. Krötz F, Sohn HY, Gloe T, Zahler S, Riexinger T, Schiele TM et al (2002) NAD(P)H oxidase–dependent platelet superoxide anion release increases platelet recruitment. Blood 100(3):917–924. https://doi.org/10.1182/blood.V100.3.917 187. Greig JA, Shirley R, Graham D, Denby L, Dominiczak AF, Work LM et al (2010) Vascular-targeting antioxidant therapy in a model of hypertension and stroke. J Cardiovasc Pharmacol 56(6): 642–650. https://doi.org/10.1097/FJC.0b013e3181f8f19f 188. Wang J, Liu Y, Shen H, Li H, Wang Z, Chen G (2020) Nox2 and Nox4 participate in ROS-induced neuronal apoptosis and brain injury during ischemia-reperfusion in rats. Acta Neurochir Suppl 127:47–54. https://doi.org/10.1007/978-3-030-04615-6_8 189. Hirano K, Chen WS, Chueng AL, Dunne AA, Seredenina T, Filippova A et al (2015) Discovery of GSK2795039, a novel small molecule NADPH oxidase 2 inhibitor. Antioxid Redox Signal 23(5):358–374. https://doi.org/10.1089/ars.2014.6202 190. Padilha EC, Shah P, Rai G, Xu X (2021) NOX2 inhibitor GSK2795039 metabolite identification towards drug optimization. J Pharm Biomed Anal 201:114102. https://doi.org/10.1016/j.jpba. 2021.114102 191. Yauger YJ, Bermudez S, Moritz KE, Glaser E, Stoica B, Byrnes KR (2019) Iron accentuated reactive oxygen species release by NADPH oxidase in activated microglia contributes to oxidative

374 stress in vitro. J Neuroinflammation 16(1):41. https://doi.org/10. 1186/s12974-019-1430-7 192. Wu N, Zheng F, Li N, Han Y, Xiong X-Q, Wang J-J et al (2021) RND3 attenuates oxidative stress and vascular remodeling in spontaneously hypertensive rat via inhibiting ROCK1 signaling. Redox Biol 48:102204. https://doi.org/10.1016/j.redox.2021.102204 193. Wang M, Luo L (2020) An effective NADPH oxidase 2 inhibitor provides neuroprotection and improves functional outcomes in animal model of traumatic brain injury. Neurochem Res 45(5): 1097–1106. https://doi.org/10.1007/s11064-020-02987-3 194. Kuntic M, Oelze M, Steven S, Kröller-Schön S, Stamm P, Kalinovic S et al (2019) Short-term e-cigarette vapour exposure causes vascular oxidative stress and dysfunction: evidence for a close connection to brain damage and a key role of the phagocytic NADPH oxidase (NOX-2). Eur Heart J 41(26):2472–2483. https:// doi.org/10.1093/eurheartj/ehz772 195. Malkov A, Popova I, Ivanov A, Jang S-S, Yoon SY, Osypov A et al (2021) Aβ initiates brain hypometabolism, network dysfunction and behavioral abnormalities via NOX2-induced oxidative stress in mice. Commun Biol 4(1):1054. https://doi.org/10.1038/ s42003-021-02551-x 196. Leung HHL, Perdomo J, Ahmadi Z, Yan F, McKenzie SE, Chong BH (2021) Inhibition of NADPH oxidase blocks NETosis and reduces thrombosis in heparin-induced thrombocytopenia. Blood Adv 5(23):5439–5451. https://doi.org/10.1182/bloodadvances. 2020003093 197. Liu F-C, Yu H-P, Chen P-J, Yang H-W, Chang S-H, Tzeng C-C et al (2019) A novel NOX2 inhibitor attenuates human neutrophil oxidative stress and ameliorates inflammatory arthritis in mice. Redox Biol 26:101273. https://doi.org/10.1016/j.redox.2019. 101273 198. Schuett J, Schuett H, Oberoi R, Koch AK, Pretzer S, Luchtefeld M et al (2017) NADPH oxidase NOX2 mediates TLR2/6-dependent release of GM-CSF from endothelial cells. FASEB J 31(6): 2612–2624. https://doi.org/10.1096/fj.201600729R 199. Ma Y, Silveri L, LaCava J, Dokudovskaya S (2017) Tumor suppressor NPRL2 induces ROS production and DNA damage response. Sci Rep 7(1):15311. https://doi.org/10.1038/s41598017-15497-0 200. Wang X, Elksnis A, Wikström P, Walum E, Welsh N, Carlsson P-O (2018) The novel NADPH oxidase 4 selective inhibitor GLX7013114 counteracts human islet cell death in vitro. PLoS One 13(9):e0204271-e. https://doi.org/10.1371/journal.pone. 0204271 201. Hagenbuchner J, Scholl-Buergi S, Karall D, Ausserlechner MJ (2018) Very long-/ and long Chain-3-Hydroxy acyl CoA dehydrogenase deficiency correlates with deregulation of the mitochondrial fusion/fission machinery. Sci Rep 8(1):3254. https://doi.org/ 10.1038/s41598-018-21519-2 202. Kennedy JA, Beck-Oldach K, McFadden-Lewis K, Murphy GA, Wong YW, Zhang Y et al (2006) Effect of the anti-anginal agent, perhexiline, on neutrophil, valvular and vascular superoxide formation. Eur J Pharmacol 531(1–3):13–19. https://doi.org/10.1016/ j.ejphar.2005.11.058 203. Ashrafian H, Horowitz JD, Frenneaux MP (2007) Perhexiline 25(1):76–97. https://doi.org/10.1111/j.1527-3466.2007.00006.x 204. Sallustio BC, Westley IS, Morris RG (2002) Pharmacokinetics of the antianginal agent perhexiline: relationship between metabolic ratio and steady-state dose. Br J Clin Pharmacol 54(2):107–114. https://doi.org/10.1046/j.1365-2125.2002.01618.x 205. Wang Q, Zhou H, Gao H, Chen S-H, Chu C-H, Wilson B et al (2012) Naloxone inhibits immune cell function by suppressing superoxide production through a direct interaction with gp91phox subunit of NADPH oxidase. J Neuroinflammation 9:32. https://doi. org/10.1186/1742-2094-9-32

C. M. Dustin et al. 206. Cifuentes-Pagano E, Saha J, Csányi G, Ghouleh IA, Sahoo S, Rodríguez A et al (2013) Bridged tetrahydroisoquinolines as selective NADPH oxidase 2 (Nox2) inhibitors. Medchemcomm 4(7): 1085–1092. https://doi.org/10.1039/C3MD00061C 207. Solbak SMØ, Zang J, Narayanan D, Høj LJ, Bucciarelli S, Softley C et al (2020) Developing Inhibitors of the p47phox–p22phox protein–protein interaction by fragment-based drug discovery. J Med Chem 63(3):1156–1177. https://doi.org/10.1021/acs. jmedchem.9b01492 208. Lyle AN, Deshpande NN, Taniyama Y, Seidel-Rogol B, Pounkova L, Du P et al (2009) Poldip2, a novel regulator of Nox4 and cytoskeletal integrity in vascular smooth muscle cells. Circul Res. 105(3):249–259. https://doi.org/10.1161/ CIRCRESAHA.109.193722 209. Guo S, Chen X (2015) The human Nox4: gene, structure, physiological function and pathological significance. J Drug Target 23(10):888–896. https://doi.org/10.3109/1061186x.2015.1036276 210. Schröder K, Zhang M, Benkhoff S, Mieth A, Pliquett R, Kosowski J et al (2012) Nox4 is a protective reactive oxygen species generating vascular NADPH oxidase. Circ Res 110(9): 1217–1225. https://doi.org/10.1161/circresaha.112.267054 211. Geiszt M, Kopp JB, Várnai P, Leto TL (2000) Identification of renox, an NAD(P)H oxidase in kidney. Proc Natl Acad Sci U S A 97(14):8010–8014. https://doi.org/10.1073/pnas.130135897 212. Gorin Y, Block K (2013) Nox4 and diabetic nephropathy: with a friend like this, who needs enemies? Free Radic Biol Med 61:130– 142. https://doi.org/10.1016/j.freeradbiomed.2013.03.014 213. Nlandu Khodo S, Dizin E, Sossauer G, Szanto I, Martin P-Y, Feraille E et al (2012) NADPH-Oxidase 4 protects against kidney fibrosis during chronic renal injury. J Am Soc Nephrol 23(12): 1967–1976. https://doi.org/10.1681/ASN.2012040373 214. Cowley AW Jr, Yang C, Zheleznova NN, Staruschenko A, Kurth T, Rein L et al (2016) Evidence of the importance of Nox4 in production of hypertension in dahl salt-sensitive rats. Hypertension (Dallas, Tex: 1979) 67(2):440–450. https://doi.org/10.1161/ HYPERTENSIONAHA.115.06280 215. Zhang M, Brewer AC, Schröder K, Santos CXC, Grieve DJ, Wang M et al (2010) NADPH oxidase-4 mediates protection against chronic load-induced stress in mouse hearts by enhancing angiogenesis. Proc Natl Acad Sci U S A 107(42):18121–18126. https:// doi.org/10.1073/pnas.1009700107 216. Kuroda J, Ago T, Matsushima S, Zhai P, Schneider MD, Sadoshima J (2010) NADPH oxidase 4 (Nox4) is a major source of oxidative stress in the failing heart. Proc Natl Acad Sci 107(35): 15565. https://doi.org/10.1073/pnas.1002178107 217. Kleinschnitz C, Grund H, Wingler K, Armitage ME, Jones E, Mittal M et al (2010) Post-stroke inhibition of induced NADPH oxidase type 4 prevents oxidative stress and neurodegeneration. PLoS Biol 8(9):e1000479. https://doi.org/10.1371/journal.pbio. 1000479 218. Frazziano G, Ghouleh IA, Baust J, Shiva S, Champion HC, Pagano PJ (2014) Nox-derived ROS are acutely activated in pressure overload pulmonary hypertension: indications for a seminal role for mitochondrial Nox4. Am J Physiol - Heart Circ 306(2):H197– H205. https://doi.org/10.1152/ajpheart.00977.2012 219. Green DE, Murphy TC, Kang BY, Kleinhenz JM, Szyndralewiez C, Page P et al (2012) The Nox4 inhibitor GKT137831 attenuates hypoxia-induced pulmonary vascular cell proliferation. Am J Respir Cell Mol Biol 47(5):718–726. https:// doi.org/10.1165/rcmb.2011-0418OC 220. Veith C, Kraut S, Wilhelm J, Sommer N, Quanz K, Seeger W et al (2016) NADPH oxidase 4 is not involved in hypoxia-induced pulmonary hypertension. Pulm Circ 6(3):397–400. https://doi.org/ 10.1086/687756 221. Schürmann C, Rezende F, Kruse C, Yasar Y, Löwe O, Fork C et al (2015) The NADPH oxidase Nox4 has anti-atherosclerotic

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

functions. Eur Heart J 36(48):3447–3456. https://doi.org/10.1093/ eurheartj/ehv460 222. Amara N, Goven D, Prost F, Muloway R, Crestani B, Boczkowski J (2010) NOX4/NADPH oxidase expression is increased in pulmonary fibroblasts from patients with idiopathic pulmonary fibrosis and mediates TGFbeta1-induced fibroblast differentiation into myofibroblasts. Thorax 65(8):733–738. https://doi.org/10.1136/ thx.2009.113456 223. Hecker L, Logsdon NJ, Kurundkar D, Kurundkar A, Bernard K, Hock T et al (2014) Reversal of persistent fibrosis in aging by targeting Nox4-Nrf2 redox imbalance. Sci Transl Med 6(231): 231ra47–231ra47. https://doi.org/10.1126/scitranslmed.3008182 224. Zhao QD, Viswanadhapalli S, Williams P, Shi Q, Tan C, Yi X et al (2015) NADPH oxidase 4 induces cardiac fibrosis and hypertrophy through activating Akt/mTOR and NFκB signaling pathways. Circulation 131(7):643–655. https://doi.org/10.1161/ CIRCULATIONAHA.114.011079 225. Hecker L, Vittal R, Jones T, Jagirdar R, Luckhardt TR, Horowitz JC et al (2009) NADPH oxidase-4 mediates myofibroblast activation and fibrogenic responses to lung injury. Nat Med 15(9): 1077–1081. https://doi.org/10.1038/nm.2005 226. Csányi G, Pagano PJ (2013) Strategies aimed at Nox4 oxidase inhibition employing peptides from Nox4 B-loop and C-terminus and p22 (phox) N-terminus: an elusive target. Int J Hypertens 2013: 842827. https://doi.org/10.1155/2013/842827 227. von Löhneysen K, Noack D, Hayes P, Friedman JS, Knaus UG (2012) Constitutive NADPH oxidase 4 activity resides in the composition of the B-loop and the penultimate C terminus. J Biol Chem 287(12):8737–8745. https://doi.org/10.1074/jbc.M111. 332494 228. Ghatak S, Hascall VC, Markwald RR, Feghali-Bostwick C, Artlett CM, Gooz M et al (2017) Transforming growth factor β1 (TGFβ1)induced CD44V6-NOX4 signaling in pathogenesis of idiopathic pulmonary fibrosis. J Biol Chem 292(25):10490–10519. https:// doi.org/10.1074/jbc.M116.752469 229. Carnesecchi S, Deffert C, Donati Y, Basset O, Hinz B, PreynatSeauve O et al (2011) A key role for NOX4 in epithelial cell death during development of lung fibrosis. Antioxid Redox Signal 15(3): 607–619. https://doi.org/10.1089/ars.2010.3829 230. Jarman ER, Khambata VS, Cope C, Jones P, Roger J, Ye LY et al (2014) An inhibitor of NADPH oxidase-4 attenuates established pulmonary fibrosis in a rodent disease model. Am J Respir Cell Mol Biol 50(1):158–169. https://doi.org/10.1165/rcmb.20130174OC 231. Cao Z, Ye T, Sun Y, Ji G, Shido K, Chen Y et al (2017) Targeting the vascular and perivascular niches as a regenerative therapy for lung and liver fibrosis. Sci Transl Med 9(405):eaai8710. https://doi. org/10.1126/scitranslmed.aai8710 232. Liu T, De Los Santos FG, Phan SH (2017) The bleomycin model of pulmonary fibrosis. Methods Mol Biol 1627:27–42. https://doi.org/ 10.1007/978-1-4939-7113-8_2 233. University of Alabama at B: GKT137831 in IPF patients with idiopathic pulmonary fibrosis. https://ClinicalTrials.gov/show/ NCT03865927 (2023). Accessed 234. Jiang JX, Chen X, Serizawa N, Szyndralewiez C, Page P, Schröder K et al (2012) Liver fibrosis and hepatocyte apoptosis are attenuated by GKT137831, a novel NOX4/NOX1 inhibitor in vivo. Free Radic Biol Med 53(2):289–296. https://doi.org/10. 1016/j.freeradbiomed.2012.05.007 235. Bettaieb A, Jiang JX, Sasaki Y, Chao T-I, Kiss Z, Chen X et al (2015) Hepatocyte nicotinamide adenine dinucleotide phosphate reduced oxidase 4 regulates stress signaling, fibrosis, and insulin sensitivity during development of steatohepatitis in mice. Gastroenterology 149(2):468–80.e10. https://doi.org/10.1053/j.gastro. 2015.04.009

375

236. Gorin Y, Cavaglieri RC, Khazim K, Lee D-Y, Bruno F, Thakur S et al (2015) Targeting NADPH oxidase with a novel dual Nox1/ Nox4 inhibitor attenuates renal pathology in type 1 diabetes. Am J Physiol Renal Physiol 308(11):F1276–F1F87. https://doi.org/10. 1152/ajprenal.00396.2014 237. Jha JC, Gray SP, Barit D, Okabe J, El-Osta A, Namikoshi T et al (2014) Genetic targeting or pharmacologic inhibition of NADPH oxidase nox4 provides renoprotection in long-term diabetic nephropathy. J Am Soc Nephrol: JASN 25(6):1237–1254. https:// doi.org/10.1681/ASN.2013070810 238. Somanna NK, Valente AJ, Krenz M, Fay WP, Delafontaine P, Chandrasekar B (2016) The Nox1/4 dual inhibitor GKT137831 or Nox4 knockdown inhibits angiotensin-II-induced adult mouse cardiac fibroblast proliferation and migration. AT1 physically associates with Nox4. J Cell Physiol 231(5):1130–1141. https:// doi.org/10.1002/jcp.25210 239. Anvari E, Wikström P, Walum E, Welsh N (2015) The novel NADPH oxidase 4 inhibitor GLX351322 counteracts glucose intolerance in high-fat diet-treated C57BL/6 mice. Free Radic Res 49(11):1308–1318. https://doi.org/10.3109/10715762.2015. 1067697 240. Szekeres FLM, Walum E, Wikström P, Arner A (2021) A small molecule inhibitor of Nox2 and Nox4 improves contractile function after ischemia–reperfusion in the mouse heart. Sci Rep 11(1): 11970. https://doi.org/10.1038/s41598-021-91575-8 241. Kofler PA, Pircher H, von Grafenstein S, Diener T, Höll M, Liedl KR et al (2013) Characterisation of Nox4 inhibitors from edible plants. Planta Med 79(3–4):244–252. https://doi.org/10.1055/s0032-1328129 242. Ugusman A, Zakaria Z, Hui CK, Nordin NAMM (2011) Piper sarmentosum inhibits ICAM-1 and Nox4 gene expression in oxidative stress-induced human umbilical vein endothelial cells. BMC Complement Altern Med 11:31. https://doi.org/10.1186/14726882-11-31 243. Bhandarkar SS, Jaconi M, Fried LE, Bonner MY, Lefkove B, Govindarajan B et al (2009) Fulvene-5 potently inhibits NADPH oxidase 4 and blocks the growth of endothelial tumors in mice. J Clin Invest 119(8):2359–2365. https://doi.org/10.1172/JCI33877 244. Murley JS, Arbiser JL, Weichselbaum RR, Grdina DJ (2018) ROS modifiers and NOX4 affect the expression of the survivinassociated radio-adaptive response. Free Radic Biol Med 123:39– 52. https://doi.org/10.1016/j.freeradbiomed.2018.04.547 245. Zhang Y, Shimizu H, Siu KL, Mahajan A, Chen J-N, Cai H (2014) NADPH oxidase 4 induces cardiac arrhythmic phenotype in zebrafish*. J Biol Chem 289(33):23200–23208. https://doi.org/ 10.1074/jbc.M114.587196 246. Weyemi U, Redon CE, Aziz T, Choudhuri R, Maeda D, Parekh PR et al (2015) NADPH oxidase 4 is a critical mediator in ataxia telangiectasia disease. Proc Natl Acad Sci U S A 112(7): 2121–2126. https://doi.org/10.1073/pnas.1418139112 247. Xu Q, Kulkarni AA, Sajith AM, Hussein D, Brown D, Güner OF et al (2018) Design, synthesis, and biological evaluation of inhibitors of the NADPH oxidase, Nox4. Biorg Med Chem 26(5): 989–998. https://doi.org/10.1016/j.bmc.2017.12.023 248. Hecker L (2020, May 19) Indoline derivatives and method for using and producing the same. US Patent 10654802 249. Bánfi B, Molnár G, Maturana A, Steger K, Hegedûs B, Demaurex N et al (2001) A ca(2+)-activated NADPH oxidase in testis, spleen, and lymph nodes. J Biol Chem 276(40):37594–37601. https://doi. org/10.1074/jbc.M103034200 250. Tirone F, Cox JA (2007) NADPH oxidase 5 (NOX5) interacts with and is regulated by calmodulin. FEBS Lett 581(6):1202–1208. https://doi.org/10.1016/j.febslet.2007.02.047 251. Chen F, Haigh S, Yu Y, Benson T, Wang Y, Li X et al (2015) Nox5 stability and superoxide production is regulated by C-terminal

376 binding of Hsp90 and CO-chaperones. Free Radic Biol Med 89: 793–805. https://doi.org/10.1016/j.freeradbiomed.2015.09.019 252. Jagnandan D, Church JE, Banfi B, Stuehr DJ, Marrero MB, Fulton DJ (2007) Novel mechanism of activation of NADPH oxidase 5. Calcium sensitization via phosphorylation. J Biol Chem 282(9):6494–6507. https://doi.org/10.1074/jbc.M608966200 253. Pandey D, Gratton J-P, Rafikov R, Black SM, Fulton DJR (2011) Calcium/calmodulin-dependent kinase II mediates the phosphorylation and activation of NADPH oxidase 5. Mol Pharmacol 80(3): 407–415. https://doi.org/10.1124/mol.110.070193 254. Montezano AC, Burger D, Paravicini TM, Chignalia AZ, Yusuf H, Almasri M et al (2010) Nicotinamide adenine dinucleotide phosphate reduced oxidase 5 (Nox5) regulation by angiotensin II and Endothelin-1 is mediated via calcium/calmodulin-dependent. Rac-1-Independent Pathways in Human Endothelial Cells 106(8): 1363–1373. https://doi.org/10.1161/CIRCRESAHA.109.216036 255. Pandey D, Patel A, Patel V, Chen F, Qian J, Wang Y et al (2012) Expression and functional significance of NADPH oxidase 5 (Nox5) and its splice variants in human blood vessels. Am J Physiol Heart Circ Physiol 302(10):H1919–H1928. https://doi.org/ 10.1152/ajpheart.00910.2011 256. Zhao G-J, Zhao C-L, Ouyang S, Deng K-Q, Zhu L, Montezano AC et al (2020) Ca2+-dependent NOX5 (NADPH oxidase 5) exaggerates cardiac hypertrophy through reactive oxygen species production. Hypertension 76(3):827–838. https://doi.org/10.1161/ HYPERTENSIONAHA.120.15558 257. Guzik TJ, Chen W, Gongora MC, Guzik B, Lob HE, Mangalat D et al (2008) Calcium-dependent NOX5 nicotinamide adenine dinucleotide phosphate oxidase contributes to vascular oxidative stress in human coronary artery disease. J Am Coll Cardiol 52(22): 1803–1809. https://doi.org/10.1016/j.jacc.2008.07.063 258. Guzik B, Sagan A, Ludew D, Mrowiecki W, Chwała M, BujakGizycka B et al (2013) Mechanisms of oxidative stress in human aortic aneurysms--association with clinical risk factors for atherosclerosis and disease severity. Int J Cardiol 168(3):2389–2396. https://doi.org/10.1016/j.ijcard.2013.01.278 259. Camargo LL, Montezano AC, Hussain M, Wang Y, Zou Z, Rios FJ et al (2021) Central role of c-Src in NOX5- mediated redox signalling in vascular smooth muscle cells in human hypertension. Cardiovasc Res. https://doi.org/10.1093/cvr/cvab171 260. Elbatreek MH, Sadegh S, Anastasi E, Guney E, Nogales C, Kacprowski T et al (2020) NOX5-induced uncoupling of endothelial NO synthase is a causal mechanism and theragnostic target of an age-related hypertension endotype. PLoS Biol 18(11): e3000885-e. https://doi.org/10.1371/journal.pbio.3000885 261. Holterman CE, Thibodeau J-F, Towaij C, Gutsol A, Montezano AC, Parks RJ et al (2014) Nephropathy and elevated BP in mice with podocyte-specific NADPH oxidase 5 expression. Journal of the American Society of Nephrology: JASN 25(4):784–797. https://doi.org/10.1681/ASN.2013040371 262. Jha JC, Banal C, Okabe J, Gray SP, Hettige T, Chow BSM et al (2017) NADPH oxidase Nox5 accelerates renal injury in diabetic nephropathy. Diabetes 66(10):2691–2703. https://doi.org/10.2337/ db16-1585 263. Brar SS, Corbin Z, Kennedy TP, Hemendinger R, Thornton L, Bommarius B et al (2003) NOX5 NAD(P)H oxidase regulates growth and apoptosis in DU 145 prostate cancer cells. Am J Physiol Cell Physiol 285(2):C353–CC69. https://doi.org/10.1152/ ajpcell.00525.2002 264. Fu X, Beer DG, Behar J, Wands J, Lambeth D, Cao W (2006) cAMP-response element-binding protein mediates acid-induced NADPH oxidase NOX5-S expression in Barrett esophageal adenocarcinoma cells*. J Biol Chem 281(29):20368–20382. https://doi. org/10.1074/jbc.M603353200 265. Chen J, Wang Y, Zhang W, Zhao D, Zhang L, Fan J et al (2020) Membranous NOX5-derived ROS oxidizes and activates local Src

C. M. Dustin et al. to promote malignancy of tumor cells. Signal Transduct Target Ther 5(1):139. https://doi.org/10.1038/s41392-020-0193-z 266. Juhasz A, Ge Y, Markel S, Chiu A, Matsumoto L, van Balgooy J et al (2009) Expression of NADPH oxidase homologues and accessory genes in human cancer cell lines, tumours and adjacent normal tissues. Free Radic Res 43(6):523–532. https://doi.org/10. 1080/10715760902918683 267. Tirone F, Radu L, Craescu CT, Cox JA (2010) Identification of the binding site for the regulatory calcium-binding domain in the catalytic domain of NOX5. Biochemistry 49(4):761–771. https:// doi.org/10.1021/bi901846y 268. Bánfi B, Tirone F, Durussel I, Knisz J, Moskwa P, Molnár GZ et al (2004) Mechanism of Ca2+ activation of the NADPH oxidase 5 (NOX5). J Biol Chem 279(18):18583–18591. https://doi.org/ 10.1074/jbc.M310268200 269. Casas AI, Kleikers PW, Geuss E, Langhauser F, Adler T, Busch DH et al (2019) Calcium-dependent blood-brain barrier breakdown by NOX5 limits postreperfusion benefit in stroke. J Clin Invest 129(4):1772–1778. https://doi.org/10.1172/JCI124283 270. Marqués J, Cortés A, Pejenaute Á, Ansorena E, Abizanda G, Prósper F et al (2020) Induction of Cyclooxygenase-2 by overexpression of the human NADPH oxidase 5 (NOX5) gene in aortic endothelial cells. Cell 9(3):637. https://doi.org/10.3390/ cells9030637 271. Cortés A, Pejenaute Á, Marqués J, Izal Í, Cenoz S, Ansorena E et al (2021) NADPH oxidase 5 induces changes in the unfolded protein response in human aortic endothelial cells and in endothelialspecific Knock-in mice. Antioxidants (Basel) 10(2):194. https:// doi.org/10.3390/antiox10020194 272. Mazumdar S, Marar T, Devarajan S, Patki J (2021) Functional relevance of Gedunin as a bona fide ligand of NADPH oxidase 5 and ROS scavenger: an in silico and in vitro assessment in a hyperglycemic RBC model. Biochemistry and biophysics reports 25:100904. https://doi.org/10.1016/j.bbrep.2020.100904 273. van der Vliet A, Danyal K, Heppner DE (2018) Dual oxidase: a novel therapeutic target in allergic disease. Br J Pharmacol 175(9): 1401–1418. https://doi.org/10.1111/bph.14158 274. De Deken X, Corvilain B, Dumont JE, Miot F (2014) Roles of DUOX-mediated hydrogen peroxide in metabolism, host defense, and signaling. Antioxid Redox Signal 20(17):2776–2793. https:// doi.org/10.1089/ars.2013.5602 275. Sham D, Wesley UV, Hristova M, van der Vliet A (2013) ATP-mediated transactivation of the epidermal growth factor receptor in airway epithelial cells involves DUOX1-dependent oxidation of Src and ADAM17. PLoS One 8(1):e54391. https:// doi.org/10.1371/journal.pone.0054391 276. Little AC, Sham D, Hristova M, Danyal K, Heppner DE, Bauer RA et al (2016) DUOX1 silencing in lung cancer promotes EMT, cancer stem cell characteristics and invasive properties. Oncogenesis 5(10):e261. https://doi.org/10.1038/oncsis.2016.61 277. Luxen S, Belinsky SA, Knaus UG (2008) Silencing of DUOX NADPH oxidases by promoter hypermethylation in lung cancer. Cancer Res 68(4):1037–1045. https://doi.org/10.1158/0008-5472. Can-07-5782 278. Little AC, Sulovari A, Danyal K, Heppner DE, Seward DJ, van der Vliet A (2017) Paradoxical roles of dual oxidases in cancer biology. Free Radic Biol Med 110:117–132. https://doi.org/10.1016/j. freeradbiomed.2017.05.024 279. Habibovic A, Hristova M, Heppner DE, Danyal K, Ather JL, Janssen-Heininger YM et al (2016) DUOX1 mediates persistent epithelial EGFR activation, mucous cell metaplasia, and airway remodeling during allergic asthma. JCI Insight 1(18):e88811. https://doi.org/10.1172/jci.insight.88811 280. Hristova M, Habibovic A, Veith C, Janssen-Heininger YM, Dixon AE, Geiszt M et al (2016) Airway epithelial dual oxidase 1 mediates allergen-induced IL-33 secretion and activation of

21

Isoform-Selective Nox Inhibitors: Advances and Future Perspectives

type 2 immune responses. J Allergy Clin Immunol 137(5):1545–56 e11. https://doi.org/10.1016/j.jaci.2015.10.003 281. Dustin CM, Habibovic A, Hristova M, Schiffers C, Morris CR, Lin M-CJ et al (2021) Oxidation–dependent activation of src kinase mediates epithelial IL-33 production and signaling during acute airway allergen challenge. J Immunol 206(12):2989–2999. https:// doi.org/10.4049/jimmunol.2000995 282. Schiffers C, Hristova M, Habibovic A, Dustin CM, Danyal K, Reynaert NL et al (2020) The transient receptor potential channel Vanilloid 1 is critical in innate airway epithelial responses to protease allergens. Am J Respir Cell Mol Biol 63(2):198–208. https://doi.org/10.1165/rcmb.2019-0170OC 283. Dickinson JD, Sweeter JM, Warren KJ, Ahmad IM, De Deken X, Zimmerman MC et al (2018) Autophagy regulates DUOX1 localization and superoxide production in airway epithelial cells during chronic IL-13 stimulation. Redox Biol 14:272–284. https://doi.org/ 10.1016/j.redox.2017.09.013 284. Cephus J-Y, Gandhi VD, Shah R, Brooke Davis J, Fuseini H, Yung JA et al (2021) Estrogen receptor-α signaling increases allergeninduced IL-33 release and airway inflammation. Allergy 76(1): 255–268. https://doi.org/10.1111/all.14491 285. Harper RW, Xu C, Eiserich JP, Chen Y, Kao C-Y, Thai P et al (2005) Differential regulation of dual NADPH oxidases/ peroxidases, Duox1 and Duox2, by Th1 and Th2 cytokines in respiratory tract epithelium. FEBS Lett 579(21):4911–4917. https://doi.org/10.1016/j.febslet.2005.08.002 286. Schiffers C, van de Wetering C, Bauer RA, Habibovic A, Hristova M, Dustin CM et al (2021) Downregulation of epithelial DUOX1 in chronic obstructive pulmonary disease. JCI Insight 6(2):e142189. https://doi.org/10.1172/jci.insight.142189 287. Gattas MV, Forteza R, Fragoso MA, Fregien N, Salas P, Salathe M et al (2009) Oxidative epithelial host defense is regulated by infectious and inflammatory stimuli. Free Radic Biol Med 47(10): 1450–1458. https://doi.org/10.1016/j.freeradbiomed.2009.08.017 288. Joo JH, Ryu JH, Kim CH, Kim HJ, Suh MS, Kim JO et al (2012) Dual oxidase 2 is essential for the toll-like receptor 5-mediated inflammatory response in airway mucosa. Antioxid Redox Signal 16(1):57–70. https://doi.org/10.1089/ars.2011.3898 289. Voraphani N, Gladwin MT, Contreras AU, Kaminski N, Tedrow JR, Milosevic J et al (2014) An airway epithelial iNOS-DUOX2thyroid peroxidase metabolome drives Th1/Th2 nitrative stress in human severe asthma. Mucosal Immunol 7(5):1175–1185. https:// doi.org/10.1038/mi.2014.6 290. Cha JJ, Min HS, Kim KT, Kim JE, Ghee JY, Kim HW et al (2017) APX-115, a first-in-class pan-NADPH oxidase (Nox) inhibitor,

377

protects db/db mice from renal injury. Lab Investig 97(4): 419–431. https://doi.org/10.1038/labinvest.2017.2 291. Hayes P, Dhillon S, O’Neill K, Thoeni C, Hui KY, Elkadri A et al (2015) Defects in NADPH oxidase genes NOX1 and DUOX2 in very early onset inflammatory bowel disease. Cell Mol Gastroenterol Hepatol 1(5):489–502. https://doi.org/10.1016/j. jcmgh.2015.06.005 292. Giusti N, Gillotay P, Trubiroha A, Opitz R, Dumont J-E, Costagliola S et al (2020) Inhibition of the thyroid hormonogenic H2O2 production by Duox/DuoxA in zebrafish reveals VAS2870 as a new goitrogenic compound. Mol Cell Endocrinol 500:110635. https://doi.org/10.1016/j.mce.2019.110635 293. Niethammer P, Grabher C, Look AT, Mitchison TJ (2009) A tissue-scale gradient of hydrogen peroxide mediates rapid wound detection in zebrafish. Nature 459(7249):996–999. https://doi.org/ 10.1038/nature08119 294. Danyal K, de Jong W, O’Brien E, Bauer RA, Heppner DE, Little AC et al (2016) Acrolein and thiol-reactive electrophiles suppress allergen-induced innate airway epithelial responses by inhibition of DUOX1 and EGFR. Am J Physiol Lung Cell Mol Physiol 311(5): L913–LL23. https://doi.org/10.1152/ajplung.00276.2016 295. Meitzler JL, Hinde S, Bánfi B, Nauseef WM, Ortiz de Montellano PR (2013) Conserved cysteine residues provide a protein-protein interaction surface in dual oxidase (DUOX) proteins. J Biol Chem 288(10):7147–7157. https://doi.org/10.1074/jbc.M112.414797 296. van der Vliet A (2017, May 11) Covalent inhibitors of dual oxidase 1 (DUOX 1). US Patent 20170128517 297. Helfinger V, Freiherr von Gall F, Henke N, Kunze MM, Schmid T, Rezende F et al (2021) Genetic deletion of Nox4 enhances cancerogen-induced formation of solid tumors. Proc Natl Acad Sci 118(11):e2020152118. https://doi.org/10.1073/pnas. 2020152118 298. Crosas-Molist E, Bertran E, Sancho P, López-Luque J, Fernando J, Sánchez A et al (2014) The NADPH oxidase NOX4 inhibits hepatocyte proliferation and liver cancer progression. Free Radic Biol Med 69:338–347. https://doi.org/10.1016/j.freeradbiomed. 2014.01.040 298. Keeney MT, Hoffman EK, Farmer K, Bodle CR, Fazzari M, Zharikov A, Castro SL, Hu X, Mortimer A, Kofler JK, CifuentesPagano E, Pagano PJ, Burton EA, Hastings TG, Greenamyre JT, Di Maio R (2022) NADPH oxidase 2 activity in Parkinson’s disease . Neurobiol Dis 170:105754. https://doi.org/10.1016/j.nbd.2022. 105754. Epub 2022 May 13.PMID: 35577065

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes

22

Tiphany Coralie de Bessa and Francisco R. M. Laurindo

Abstract

Keywords

Nox NADPH Oxidases exhibit a basic organization comprising a catalytic transmembrane subunit closely regulated by canonical regulatory subunits, discussed in other chapters of this book. However, many additional proteins regulate the expression, assembly, structure, activity and subcellular traffic of Nox subunits. As such, they gravitate around Nox complexes and physically associate with at least one among the regulatory or catalytic subunits. Given that such associated proteins, in turn, exert canonical effects distinct from Nox regulation, they connect Nox function to physiological cell programs, mediating cross-talk to and from Noxes. This chapter provides a systematic overview of proteins for which the physical interaction with Noxes has been validated by “wet-lab” experiments. Such proteins support both stimulatory or inhibitory effects towards several aspects of Nox regulation and can be roughly classified as: (a) kinaserelated organizers; (b) general organizers; (c) chaperonelike organizers; (d) RhoGTPase and/or cytoskeletonrelated organizers; (e) scaffold proteins. In addition, we provide an overview of the Nox interactome “in silico”, indicating that Noxes cross-talk with their environment preferentially via interactive protein hubs associated with their regulatory, rather than catalytic subunits. Characterizing the roles of Nox-associated proteins is essential to provide an integrative understanding of Noxes within multiple cellular physiological contexts.

NADPH oxidases · Nox-interacting proteins · NADPH oxidase organizers · Protein-disulfide isomerase · Poldip2 · RhoGDI · P47phox · Rac1 · Nox interactome

T. C. de Bessa · F. R. M. Laurindo (✉) Laboratório de Biologia Vascular, LIM-64 (Biologia Cardiovascular Translacional), Instituto do Coracao (InCor), Hospital das Clinicas HCFMUSP, Faculdade de Medicina, Universidade de Sao Paulo, Sao Paulo, SP, Brazil e-mail: [email protected]

The Nox family NADPH oxidases are central enzyme complexes dedicated to the generation of reactive oxygen intermediates with purposes including signal transduction, homeostasis regulation and microbial defense. The complex is composed of catalytic subunits (Nox 1–5, Duox 1–2) and a host of canonical regulatory subunits which associate, in variable combinations, with each type of catalytic subunit. Such regulatory subunits include: p22phox, Rac1 or Rac2, p47phox, p67phox, p40phox, NoxO1 and NoxA1, as well as DuoxA1 and DuoxA2. Many other proteins, however, gravitate around Nox complexes and may eventually associate with one or more catalytic or regulatory subunits, while not being part of the central Nox complex. These proteins uncover important pathways whereby Noxes integrate physiological cell programs, undergo subcellular traffic, modulate complex activation, undergo stabilization or degradation and associate with other protein complexes or microenvironments, among several possibilities. In other words, such a host of associated proteins represent crosstalk pathways to and from Noxes. This chapter aims to systematically discuss, on a non-exhaustive fashion, the most conspicuous of such associated proteins. For the purposes of this chapter, we define Noxassociated proteins as those that are distinct from canonical regulatory subunits (which are well covered in other chapters of this book) and for which there is evidence for all of: 1. Known cellular functions distinct from direct Nox regulation

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_22

379

380

2. Physical interaction, validated by at least two “wet-lab” experimental methods, with one or more catalytic or canonical regulatory subunit of a Nox complex 3. Loss or gain of function experiments suggesting their Nox-related functional roles We start with a basic hypothesis-generating exercise aimed to roughly assess the hierarchic organization of the Nox interactome, using a general protein interaction tool such as BioGRID. A high degree of interactivity indicates that the protein in focus may be a physiologically relevant node, but has multiple and likely less specific targets. Such results (Fig. 22.1) indicate that catalytic Nox subunits per se are not highly populated hubs and interact with few proteins other than their canonical regulatory subunits. Meanwhile, the canonical regulatory subunits, while being better interactors, still tend to exhibit (except for Rac1) an interactome network one order of magnitude lower than that of most Nox-associated proteins fulfilling the above definition adopted for this chapter. This scenario suggests, not unexpectedly, that Nox complexes cross-talk with their environment preferentially through highly interactive but less specific proteins associated to their canonical regulatory subunits. Figure 22.2 and Table 22.1 summarize the Nox-associated proteins discussed in this chapter, tentatively

Fig. 22.1 Nox interactome. Diagram representing the interactome of Nox NADPH Oxidase catalytic (at the center) or canonical regulatory subunits, according to BioGRID database. The outer circles depict Nox-

T. C. de Bessa and F. R. M. Laurindo

classified according to their main effect regarding Nox regulation. It is likely that such proteins are only some conspicuous representatives of large protein complex networks. Also, one must consider that the impact of several of these proteins on Nox function is complex and caution must be exercised with respect to generalizations.

1

Associated Proteins with a Main Stimulatory Effect on Noxes

1.1

Kinase-Related Nox Organizers

Serine/Threonine Kinases Correct Nox subunit translocation and complex assembly are a major step for Nox regulation. For this task, p47phox and other organizer subunits of Nox1, 2 and 3 complexes are crucial. In its inactive form, p47phox auto-inhibitory region masks two Src-homology 3 (SH3) domains and the phosphoinositide-binding (PX) domain. Activation of p47phox requires phosphorylation of this autoinhibitory region at Ser 303, 304 and 328 [1], which then exposes SH3 and PX domains and allows p47phox to translocate to the membrane, where it binds to p22phox [2–4]. For that, p47phox PX domain binds to phosphatidylinositol 3,4-bisphosphate and

associated proteins discussed in the present chapter. The number of interactors for each protein is indicated in red

22

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes

381

Fig. 22.2 Scheme depicting NADPH oxidase complexes and their associated proteins. Proteins within green frames display a main activator effect on NOX complexes, whereas proteins within red frames have

mainly an inhibitory effect. Proteins without frames can display either stimulatory or inhibitory effects

phosphatidic acid (PA) at the membrane or to moesin, an actin cytoskeleton-related Nox regulator tethering to membranes [3, 5, 6]. Several kinases have been implicated in p47phox phosphorylation [3, 4], in particular protein kinase-C (PKC) isoforms (δ, β, α, ζ) [7, 8], which may variably synergize, in a cell type and stimulus-dependent way, with other PKCs, p21(Rac1) activated kinase (PAK)1 [9], mitogen-activated protein kinases (MAPK) [10], extracellular signal-regulated kinase (ERK)1/2 [11], protein

kinase B-α (Akt) [12, 13], p38MAPKinase [14], protein kinase cAMP-activated catalytic subunit-α (PKA) [15, 16], interleukin-1 receptor associated kinase-4 (IRAK-4) [17] and SRC proto-oncogene, non-receptor tyrosine kinase (Src) [18]. Other Nox subunits, e.g., p22phox, may also be phosphorylated by PKC [19]. Together, those kinases can be considered as Noxassociated proteins, themselves subjected to varied regulatory stimuli and able to interact with other Nox regulators.

382

T. C. de Bessa and F. R. M. Laurindo

Table 22.1 Nox-associated proteins and their most relevant known target interactors. Proteins are classified according to main proposed Nox-related function, as discussed in the text Stimulatory effect

Kinase-related Nox Organizers

General Nox Organizers

Chaperone-like Organizers RhoGTPase & Cytoskeleton Nox Organizer

Scaffold proteins supportive of Nox

Inhibitory effect

Kinases Chaperone-like

RhoGTPase

A specific example is the Rac1 effector PAK1, which was shown to inhibit Nox2 activation by competition of its p21-binding domain with p67phox [20]. Tks4 and Tks5 (Tyrosine Kinase Substrate Proteins 4 and 5) Tyrosine kinase substrate (Tks) proteins, initially described as substrates of Src tyrosine-kinase, are scaffold proteins involved as regulators of cell signaling associated with cell migration, proliferation and differentiation. Tks(s) bind to and anchor key signaling proteins to the cytosolic face of plasma membrane [21] and play crucial roles in the recruitment of proteins involved in the epidermal growth factor receptor (EGFR) pathway. Tks4 and TKs5 expressions are

Protein AKT/PKB ERK1/2 IRAK-4 p38MAPKinase PAK PKA PKC isoforms (δ, β, α, ζ) Src kinase TKs4 TKs5 RFK PDI/PDIA1 TXNDC11/EFP1 Prdx6 SH3YL1 CYB5R3 SYTL1 Rab27A/B CYBC1/EROS Hsp90 RhoGDI1 β-Pix P-Rex1 Tiam1 Vav1-3 EBP50 IQGAP1 Poldip2 PCNA Cyclophilin A CKII HSP70 NRROS HACE1 Cdc42

Main target interactors p47phox p47phox p47phox p47phox Rac1/2; p22phox; p47phox p47phox p47phox; p22phox p47phox NoxA1; p22phox NoxA1; p22phox p22phox Rac1; p47phox Duox1/2 NoxA1; p67phox Nox4; p22phox Nox4 p67phox Rac1 Nox2; p22phox Nox5 Rac1, Rac2 Rac1 Rac1 Rac1 Rac1 p47phox Rac1 p22phox; Nox4 p47phox p47phox p47phox; Rac1 Nox2; Nox5 Nox2 Rac1 Nox2

References [12, 13] [10, 11] [17] [15] [9] [15, 16] [7, 8] [18] [26–28] [26–28] [30] [49, 55, 59] [70] [73–76] [80] [82] [84] [90] [93–96] [99, 100] [103–105] [119, 120] [115] [117, 118] [116] [122] [124] [127] [140] [147] [152] [99] [157] [158, 159] [160]

particularly enriched in invadopodia and podosomes [22– 24]. The involvement of Tks(s) in Nox activation was proposed on the basis of their functional roles as Nox organizers similar to p47phox [25, 26]. Indeed, Tks4 and Tks5 display an overall structure analogous to that of p47phox organizer superfamily [27], containing N-terminal tandem SH3 domains plus a PX domain, multiple proline-rich regions (PRRs) and several Src phosphorylation sites. Tks4 can bind to NoxA1, promoting Nox1 localization at invadopodia [26]. Through loss- and gain-of-function assays in distinct cell types, Tks4 and Tks5 were shown to support Nox1, Nox2, Nox3 or Nox4-dependent ROS generation [25, 26, 28] (NB: in this chapter, we adopt the term “ROS” to

22

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes

designate unspecified reactive oxygen intermediates; however, it must be stressed that each type of ROS has distinct and specific signaling effects [29]). The mechanisms of Tks effects associate with their capacity to bind NoxA1 PRR through their SH3 domains [26]. In addition, Tks5 can support Nox-ROS production also through p22phox binding, while inducing invadopodia formation [28]. Altogether, these data indicate that Tks4 and Tks5 act as Nox organizers able to recruit Nox complexes to specific subcellular compartments. RFK (Riboflavin Kinase; RIFK) RFK is an essential enzyme that catalyzes the phosphorylation of riboflavin (vitamin B2) to form flavin mononucleotide (FMN), which, via flavin adenine dinucleotide (FAD) synthetase, yields FAD. Both FMN and FAD are essential cofactors of dehydrogenases, reductases and oxidases. The requirement of RFK for Nox NADPH oxidases was proposed [30] following yeast-two hybrid experiments depicting RFK binding to tumor necrosis factor receptor type 1 (TNFR1) and its coupling proteins. RFK loss-of-function markedly impairs superoxide generation triggered by tumor necrosis factor (TNF)-α or other Toll-like receptor (TLR) ligands. TNF-α was able to recruit p22phox, Nox1 or Nox2 and coimmunoprecipitated with these subunits in cells bearing RFK but not in those depleted of this enzyme. Importantly, Noxes are normally not fully saturated with FAD at rest and sustained FAD synthesis is required for their full activation, indicating a mechanism whereby RFK supports Nox activation. Accordingly, exogenous FMN and FAD fully substituted for the absence of RFK in TNF-α primed cells [30] and even a short-term deficiency of riboflavin can impair phagocytic respiratory burst [31]. An analogous requirement for RFK has been shown regarding Nox1mediated tumor cell death [32].

1.2

General Nox Organizers

Protein Disulfide Isomerases and Other ER Chaperones Protein disulfide isomerases (PDIs) are thiol redox chaperones from the thioredoxin superfamily, primarily located at the endoplasmic reticulum (ER) lumen. The canonical function of most PDIs is the introduction and/or isomerization of disulfide bonds during protein folding of nascent proteins. In addition, PDIs regulate the ER-cytosol traffic of un/misfolded proteins for degradation by the ubiquitinproteasome system (a process called ERAD or ER-assisted degradation). The PDI family prototype PDIA1, also known simply as PDI, is a 55-kDa protein comprising 4 thioredoxin domains in tandem, denominated a-b-b′-a′. Domains a and a ′ are catalytic, displaying Cys-Gly-His-Cys dithiol motifs,

383

while domains b and b′ are non-redox and non-catalytic, displaying thioredoxin folds enriched in hydrophobic residues involved in substrate recognition and binding [33, 34]. Domains a′ and b′ are connected by an unstructured segment conferring substantial mobility to PDI [33–36]. In addition, PDI also displays a C-terminal domain containing the KDEL (Lys-Asp-Glu-Leu) sequence, responsible for its ER retrieval from the secretory system. All four domains are required for PDI isomerase activity, while substrate oxidation or reduction may be achieved by truncated protein constructs. In addition, PDIA1 and PDIs in general exhibit a chaperone activity, per se independent of the thiol motifs, which is important for binding to unfolded protein substrates and is accounted mainly by the b domains [37]. PDIA1 oxidation enhances its chaperone activity by exposing the b domains. Moreover, PDI activities relate to their high dynamic mobility. Crystal structures predict a relatively closed form of reduced PDI, while oxidized PDI displays an open form [38]. Furthermore, in solution PDI exhibits fast dynamic motion at atomic force microscopy, with oxidized PDI populating more often the open configurations and able to form dimers and tetramers in proportion to substrate complexity [39]. These data indicate that PDI can essentially behave as if an intrinsically unfolded protein, consistent with functions in protein folding and regulation of many possible interactions. Typical clients of PDI are ER-processed trans-membrane or secreted proteins [33, 40]. However, in the course of signaling events, PDIs associate with many additional proteins, possibly involving an interplay of regulatory and client-type interactions. In addition to the canonical ER pool, there is evidence for additional pools of PDI at extra-ER locations. The peri/ epicellular pool has been well described [41]. Although such extracellular PDIA1 levels are typically 820 million years. Such functional ties are yet unclear, but co-expression and protein-protein interactions were identified between PDIA1 and RhoGDI1 [58]. GEFs (P-Rex1, β-Pix,Vav1, Vav2, Vav3 and Tiam1) Like all G-proteins, Rac1 and Rac2, which behave as activator subunits of Nox1–3 complexes, are regulated by GEFs and guanine-activating proteins (GAPs). GEFs induce GTPase activation and stimulate signal transduction by catalyzing the exchange of GDP by GTP. GAPs catalyze the return to inactive form, facilitating the hydrolysis of GTP to GDP. Different GEFs and GAPs may act specifically for each RhoGTPase [112]. Some GEFs and GAPs were specifically described to regulate Nox activity in the context of Rac regulation, such as phosphatidylinositol-3,4,5-trisphosphate dependent Rac exchange factor (P-Rex)1, Rho guanine nucleotide exchange factor-7 (β-Pix), VAV guanine nucleotide exchange factor (Vav)1, Vav2, Vav3, T-lymphoma invasion and metastasisinducing protein (Tiam)1 and Trio Rho guanine nucleotide exchange factor (Trio) [113–116]. Endogenous Rac1 activation in COS-phox cells by Vav1, Vav2 or Tiam1 transfection induces increased Nox-dependent ROS production, with Vav1 having the strongest effect [113]. In phagocytes, both Vav1 and Vav3 are important for oxidative burst, with Vav1 related to its initiation and Vav3 to persistence [114]. In endothelial cells subjected to shear stress, both Vav2 and Tiam1 are involved in Rac1-dependent Nox activation [115]. Tiam1 polarizes Rac1 activation by binding VE-cadherin and then p67phox, while Tiam1 downregulation prevents ROS production by endothelial cells or aortic tissue [115]. In Caco2 and HEK293 cells, growth factor-induced ROS production dependent on Rac1 and Nox1 requires β-Pix, which associates with Nox1 C-terminal region, and β-Pix overexpression increases ROS production [117]. Also β-Pix phosphorylation at Ser-340 supports Nox1 through Rac1 activation [118]. EBP50 (Ezrin-Radixin-Moesin Binding Phosphoprotein 50; NHERF1; SLC9A3 Regulator1) EBP50 binds ERM (Ezrin-Radixin-Moesin) proteins when phosphorylated, e.g., by ROCK1, and acts as a scaffold protein at sites connecting actin cytoskeleton and plasma membrane [119]. EBP50 supports VSMC proliferation and neointimal hyperplasia, a process in which Nox1 is known to play a significant role. EBP50 was reported, in a series of experiments, to bind p47phox and to support non-canonical (that is, NoxO1- independent) Nox1 activation [120]. EBP50 is essentially required for angiotensin-II mediated superoxide

22

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes

generation. Nox2, as well as the canonical NoxO1/Nox1 system, were ruled out for this effect, while several results pointed to the p47phox/Nox1 system. Evidences for direct p47phox/EBP50 interaction involve confocal colocalization experiments, coimmunoprecipitation, FRET assays and proximity ligation assays in COS-7 cells, VSMCs or aortic rings. Deletion experiments mapped specific p47phox residues binding to EBP50, indicating that the specific residues are absent in NoxO1. Functionally, VSMCs devoid of EBP50 do not exhibit angiotensin-II mediated, Nox1-dependent hypertrophy and aortas from EBP50 knock-out mice do not vasoconstrict in response to angiotensin-II. The functional EBP50/p47phox interaction indicates specific pathways for the so-called “hybrid Nox1 system” (i.e., p47phoxdependent, as opposed to NoxO1-dependent) [120] and further extends roles for cytoskeleton/plasma membrane tethering proteins such as ERM as Nox scaffolds. Indeed, p47phox binding to moesin has been previously described [5]. IQGAP1 (IQ Domain GTPase-Activating Protein 1) IQGAP1 is a scaffold protein involved as mediator/regulator of diverse cell processes, mostly related to cytoskeleton regulation, including cell adhesion and migration. The structure of IQGAP1 exhibits four calcium-independent calmodulin binding domains named IQ, and a Ras-GAP-related domain devoid of a traditional GAP activity [121]. Concerning Nox regulation, IQGAP1 is an important effector of Rac1, which directly binds to this RhoGTPase [122]. IQGAP1 mediates vascular endothelial growth factor-A (VEGF)-induced activation of Nox2 and ROS-dependent migration in endothelial cells in the wound healing response. In this case, oxidant production, as well as enrichment of Nox2, actin and IQGAP1 occur at the wound edge [123]. IQGAP1 downregulation impairs Nox2 translocation and prevents H2O2 production, Nox2-actin interaction and cell migration [123]. Thus, IQGAP1 may function as a scaffold to link Nox2 and actin at the leading edge [123].

1.5

Scaffold Proteins Supportive of Nox Activation

Poldip2 (DNA Polymerase Delta Interacting Protein 2; PDIP38, Polymerase Delta Interacting Protein 38) Poldip2 was originally described as a protein associated with DNA-polymerase Delta and proliferating cell nuclear antigen (PCNA) and as such putatively implicated in gene expression regulation and DNA repair [124]. Increasing evidence points to a multifunctional role of Poldip2 through its interaction with distinct proteins, possibly through its dynamic flexible N-terminal domain, which also bears a

389

putative cleavable mitochondrial-addressing sequence. In line with its multiple functions, Poldip2 exhibits several subcellular locations, including the nucleus, cytosol, focal adhesions, plasma membrane and mitochondria. Such mitochondrial location seems variable according to cell type and context [124]. Multiple functions of Poldip2 have been reviewed elsewhere [124]. Here we focus on how the interaction of Poldip2 with Noxes may be involved in several of these effects, particularly in vascular cells, which have been better studied. The role of Poldip2 in Nox regulation was discovered following yeast-two hybrid experiments searching for Noxassociated proteins, using the C-terminal tail of p22phox as bait in a VSMC library [125]. The Poldip2-p22phox association was confirmed through pull-down and coimmunoprecipitation assays in transfected or non-transfected cells. While both Nox1 and Nox4 also associate with Poldip2, only Nox4 transfection stimulates ROS generation in a way requiring coexpression of p22phox, confirming that the latter is likely the direct Poldip2 interactor. Poldip2 overexpression stimulated, while siPoldip2 decreased ROS production by Nox4. Poldip2 down-regulation promoted a phenotype not unlike that of Nox4 underexpression and delocalized Nox4 and p22phox from focal adhesions. An important phenotype of Poldip2 overexpression was the increase in stress fiber formation, stabilization of focal adhesion and marked RhoA activation. Both over and underexpression of Poldip2 impaired VSMC migration, probably due to disrupted coordination of front vs. rear focal adhesion turnovers [125]. This was shown in additional studies depicting loss of force polarization due to Poldip2-mediated impairment of focal adhesion dissolution [126], while Poldip2/Nox4-dependent actin oxidation drives focal adhesion assembly and maturation [127]. However, the interplay between Nox4 and RhoAdependent Poldip2 effects may be complex. While Nox4 in general leads to RhoA activation, e.g., in VSMCs [125], Poldip2-mediated RhoA activation, while redox-dependent [125], seems unrelated to Nox4, since the RhoGEF epithelial cell transforming 2 (Ect2) was shown to mediate Poldip2dependent RhoA activation in a Nox4-independent manner [128]. Importantly, Poldip2 affects vascular structure, since aortas from Poldip2-deficient heterozygous mice display increased collagen accumulation and elastic fiber breaks, together with impaired contractile force and enhanced stiffness [129]. Poldip2+/– mice exhibit protection against aortic aneurysm development [129]. Such mice also exhibit decreased VSMC proliferation and neointimal size after injury [130]. However, effects of Poldip2 on VSMC phenotype also depend on other determinants. While Poldip2 deficiency was shown to promote a VSMC phenotype locked on

390

cell differentiation, this effect was related to impaired mitochondrial function [131]. Indeed, Poldip2 deficiency is known to reprogram the metabolism, with increased glycolysis and depressed mitochondrial respiration [132]. In addition to VSMC effects, Poldip2 enhances blood-brain barrier permeability and inflammation after induced stroke [133] and sustains lung edema in respiratory distress models [134]. However, the extent to which Nox4 regulates such endothelial permeability effects has not been investigated. Together, this implicates Poldip2 as a redox signaling regulator able to integrate Nox4-related processes with mechanobiological, mitochondrial, metabolic and structural responses at several levels, while also linking these processes to additional redox/Nox-independent pathways. PCNA (Proliferating Cell Nuclear Antigen; Cyclin) Proliferating Cell Nuclear Antigen (PCNA) in most cell types is located at the nucleus, as a cell cycle regulator and as cofactor in DNA replication and damage response [135]. PCNA displays flexible interdomain connecting loops prone to interaction with other proteins [136], thus conferring PCNA the characteristics of a versatile protein scaffold. At least in terminally differentiated neutrophils, PCNA becomes mostly located at the cytosol, where it regulates survival by sequestering apoptotic procaspases [137] and supporting glycolysis [138]. However, the roles of PCNA in Nox2 activation remained obscure until recently, when a search for PCNA-associated proteins using PCNA immunoprecipitation and mass spectrometry yielded p47phox as a hit, among additional proteins related to NADPH oxidase, GTP-binding, redox processes, cytoskeletal regulation and glycolysis [139]. Further studies using confocal colocalization, coimmunoprecipitation, proximity ligation assays and surface plasmon resonance indicated that p47phox, but not p67phox, p40phox or Rac2 depicts stable robust association with PCNA. Additional experiments indicated association of PCNA interdomain-connecting loop with p47phox residues at the lipid-binding PX domain, a model further supported by crystallographic evidence. Importantly, such PCNA-p47phox interaction seems functionally relevant. First, Nox2 activation in neutrophil-like differentiated PLB985 cells was prevented by forcefully redirecting PCNA to the nucleus or by an siRNA against PCNA. Second, in a cell-free NADPH oxidase system, pharmacological PCNA inhibition decreased superoxide production [154]. Third, western analyses, confocal immunofluorescence and proximity ligation assays indicated that p47phox and PCNA are associated in neutrophil cytosol at rest and co-translocate to membranes upon activation, where they subsequently loose their association. The proposed model is that PCNA links p47phox in the cytosol to

T. C. de Bessa and F. R. M. Laurindo

restrain its assembly and activation, but supports NADPH oxidase assembly after activating stimuli. Additional evidence supported roles for PCNA-p47phox interaction in vivo, with pharmacological PCNA inhibition significantly preventing redox-dependent mucosal damage due to induced colitis in mice [139]. These results, together with previous literature data, establish PCNA as a novel co-organizer of Nox NADPH oxidase and neutrophil activation, in addition to being an apoptosis regulator. Whether this also occurs in other cell types and contexts remains to be investigated. Peptidyl-Prolyl Cis-Trans-Isomerases: Cyclophilin-A and Pin-1 Cyclophilin A (CyPA), a canonical ligand of cyclosporin-A, is a peptidyl-prolyl-cis-trans isomerase, and acts also as a scaffold protein independent of its enzymatic activity [140], depicting multiple functions in protein folding, traffic and gene transcription [141]. CyPA is a secreted factor triggered by oxidative stress in VSMCs [142], acting as a proinflammatory cytokine with roles in diseases such as aortic aneurysm [143] and cardiac hypertrophy [144]. In vascular cells, CyPA loss-of-function significantly impairs ROS generation induced by angiotensin-II [143], while gain-of-function enhances it [145]. In VSMCs stimulated with angiotensin-II, p47phox and CyPA, together with actin filaments, colocalized in caveolae-containing fractions, while no p47phox was detected in caveolae fractions of CyPA-/- VSMCs. Importantly, physical interaction between CyPA and p47phox was confirmed via confocal localization and coimmunoprecipitation assays. Moreover, inhibiting p47phox phosphorylation or deleting its phosphoinositide-binding (PX) domain significantly disrupted CyPA-p47phox interaction. This suggests that p47phox phosphorylation is required to expose its domains for CyPA interaction. In summary, CyPA functionally interacts with Nox by physical interaction with p47phox, together with its co-translocation with actin filaments to caveolae [145]. Interestingly, a cyclophilin (Cyp19) from Trypanosoma cruzi, a protozoan known to induce ROS generation for its survival, co-localizes with p47phox-enriched membrane domains in macrophages [146]. In addition to CyPA, the peptidylprolyl cis/trans isomerase, NIMA-interacting 1(Pin-1) was shown bind to p47phox, inducing its conformational change to facilitate PKC-induced phosphorylation. Such effect supports TNFα-induced neutrophil priming of Nox2 [147] and Toll-like receptor [TLR]7/8mediated neutrophil effects [148].

22

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes

2

Associated Proteins with a Main Inhibitory/Mixed Effect on Noxes

2.1

Kinase-Related

CKII (Casein Kinase II) While several kinases play a stimulatory role on p47phox, some negatively regulate p47phox and Nox activation, as is the case of casein kinase 2 (CKII) and p38MAPKinase. CKII is a ubiquitous Ser/Thr-kinase with more than 300 substrates implicated in cellular processes including DNA repair, cell cycle regulation, proliferation, survival and ER stress responses [149]. CKII catalyzes the in vitro phosphorylation of p47phox at both SH3 domains [150], while binding to p47phox in its C-terminal portion [151], negatively regulating Nox. In neutrophils, CKII inhibition prior to N-formyl-methionylleucyl-phenylalanine (fMLF) stimulation enhances p47phox membrane translocation, while augmenting and prolonging the respiratory burst [150]. The CKII-mediated deactivating p47phox phosphorylation is enhanced by p47phox Cys196 S-nitrosylation, which induces a conformational change in p47phox and promotes its binding to CKII [151]. Distinct mechanisms of CKII-mediated Nox negative regulation were also identified [152]. In cerebral ischemia, a decrease of CKII protein expression was observed, concomitant to ROS production and neuronal cell death. In parallel, pharmacological CKII inhibition before cerebral ischemia-reperfusion promotes enhanced gp91phox, p67phox and Rac1 expressions, but not p47phox membrane translocation. CKII and Rac1 co-immunoprecipitation in the mouse brain was shown at baseline, while after cerebral ischemia such CKII/Rac1 interaction is lost, together with increased Rac1 activity. In primary neurons, Rac1 silencing prevents cell death induced by CKII inhibition. Together, these results suggest that Rac1 inactivation is a mechanism underlying CKII-induced negative regulation of Nox2 in the brain [152]. Therefore, CKII seems to negatively regulate Nox by targeting distinct Nox subunits (p47phox and/or Rac1) according to cell type or context.

2.2

Chaperone-Like

Hsp70 (Heat-Shock Protein-70) Concerning Nox regulation, Hsp70 has an effect opposed to that of Hsp90 (discussed above). Hsp70 binding to Nox2 and Nox5 mediates their degradation promoted by Hsp90 inhibition, while the Hsp70-linked ubiquitin ligase carboxyl-

391

terminus of Hsc70 interacting protein (CHIP) was shown to mediate such degradation. Forced Hsp70 overexpression decreases Nox5 activity [98]. Overall, the molecular chaperones Hsp70 and Hsp90 bind to and contribute to govern in opposite ways the stability and degradation of catalytic Nox subunits other than Nox4 (since Nox4 is unaffected by Hsp90, as discussed above). NRROS (Negative Regulator of Reactive Oxygen Species; LRRC33, Leucine-Rich Repeat-Containing Protein 33) NRROS is a transmembrane protein mainly localized at the ER, containing several leucine-rich domains and a short cytosolic sequence, identified to inhibit TLR-associated inflammation [153]. The NRROS function associated with its name relates to down-regulation of the respiratory burst in phagocytes [154], while other important functions have also been uncovered. NRROS was identified in bone marrow-derived macrophages through a microarray analysis of genes downregulated by combined exposure to interferon-gamma and LPS. NRROS genetic deletion significantly enhanced macrophage ROS response to each of these agonists alone, while upon incubation with both—a condition in which NRROS is strongly down-regulated in wild-type cells—there was no difference in the already maximal response. Several types of assays indicated that NRROS-deficient phagocytes promote exacerbated bacterial killing dependent on Nox2derived ROS. Interestingly, while NRROS KO mice display Nox2-dependent enhanced survival when infected with such bacteria, mice knock-out for NRROS in bone marrow cells or phagocytes were significantly more sensitive to autoimmune encephalomyelitis through an oxidative mechanism. Importantly, in NRROS-knockout myeloid cells, there was enhanced protein, though not mRNA expression levels of gp91phox and p22phox. Meanwhile, the expression and phosphorylation levels/activity of p47phox, p67phox, p40phox and Rac1, as well as mitochondrial-derived ROS were unaffected. Moreover, in NRROS knockout cells, the endoglycosidase H-sensitive (i.e., of ER origin) 58kDa Nox2 accumulated to a larger extent vs. wild-type cells. This indicates that normally NRROS, which is ER-resident, down-regulates Nox2 protein via ERAD/proteasomedependent degradation. Interestingly, NRROS was shown to interact with nascent Nox2 but not with p22phox; however, the more stable Nox2-p22phox dimer outcompetes NRROS-mediated Nox2 degradation. Thus, NRROS downregulation during some inflammatory stimuli ensures enhanced Nox2 activation, while normal NRROS levels

392

T. C. de Bessa and F. R. M. Laurindo

balance such pathway to prevent excessive oxidantdependent [auto]immune processes [154]. Somewhat unexpectedly, mice with genetic NRROS ablation display a prominent neurologic phenotype, with lethal motor deficits associated with microglial defects rather than neuronal loss [155]. A major player in these effects may be Transforming Growth Factor [TGF]-β1 loss of function, since NRROS was identified to associate with latent TGF-β1 and anchor it to the cell surface of myeloid-type cells [156]. HACE1 (HECT Domain and Ankyrin Repeat Containing E3 Ubiquitin Protein Ligase 1) HACE1 is a ubiquitous E3 ubiquitin ligase mainly localized at the ER, Golgi and cytosol. HACE1 structure depicts a homologous to the E6-AP carboxyl terminus (HECT) domain accounting for ubiquitin ligase activity plus C-terminal ankyrin repeats (known to mediate cytoskeletal attachment of integral membrane proteins) involved in substrate recognition [157]. HACE1 was described as a tumor suppressor [157], the first one known to regulate ROS production and Nox complexes by ubiquitin-mediated proteasomal degradation of active Rac-1 (Rac1-GTP) [158]. HACE1 was shown to bind Rac1 in vitro. K-48 ubiquitination at lysine-147 of Rac1 was shown to be a HACE1-binding motif, which culminates in Rac1 proteasomal degradation [159]. HACE1 loss of function in mice and/or cell models leads to increase in Nox-dependent ROS production [158], in a way reversible by wild-type HACE1 expression, by Rac1 silencing with si-RNA or ML171 Nox1 inhibitor treatment [158]. Also, HACE1 knock-out in zebrafish leads to ROS production reverted by Nox inhibitors. Interestingly, HACE1 preferentially targets active Rac1 bound to NoxA1 and, in turn, NoxA1 silencing impairs Rac1 ubiquitination and proteasomal degradation. Overall, the proposed model is that HACE1 negatively regulates Nox1 activity (and potentially Nox2/Nox3) through proteasomal degradation of Rac1 bound to Nox complexes [158].

2.3

RHOGTPase and/or Cytoskeleton-Related

Cdc42 Cdc42, like Rac1 and RhoA, is a Rho GTPase and as such is involved mainly in cytoskeleton organization (particularly in the formation of filopodia), but also in other processes such as transcription, cell cycle progression, vesicle trafficking and cell survival. In a cell-free system and COS7-phox cells (engineered cells with gp91phox, p22phox, p47phox, and p67phox stable expression), prenylated Cdc42 is able to bind to Nox2, however in competition with Rac2 and Rac1.

Since Cdc42 is unable to activate the NADPH oxidase because it does not bind to p67phox but competes with Rac for binding to Nox2, it exhibits an inhibitory effect on Rac-mediated NADPH oxidase activation [160]. In fact, in neutrophils exposed to fMLF, endogenous Cdc42 inhibition induces an increase of Nox2 activity [160].

3

Concluding Remarks

Nox NADPH oxidases are the main enzyme complexes dedicated, in their most characteristic format, to microbial killing and host defense, while in their basic configuration they assume several distinct organizations to account for the production of distinct ROS with signaling purposes. These features translate into a sophisticated multilevel regulation. At the basic level, Nox catalytic subunits are strictly controlled by a host of canonical regulatory subunits. At an extended level, such regulatory subunits and, to a lesser extent, some catalytic subunits as well, interact with several associated proteins. Such proteins connect Nox NADPH oxidases to physiological cell programs, while being co-responsible for Nox complex organization, subunit localization, degradation, traffic, cytoskeletal interactions, among several other functions. The discussion of these proteins provides an extended view of how these ancestral protein complexes deal with the potentially damaging species generated by them. Moreover, they offer an idea of the hierarchical organization of Nox complexes within a cellcontext based perspective. Finally, they allow us a glimpse of the social lives of these fascinating proteins.

References 1. Shiose A, Sumimoto H (2000) Arachidonic acid and phosphorylation synergistically induce a conformational change of p47phox to activate the phagocyte NADPH oxidase. J Biol Chem 275(18): 13793–13801 2. de Mendez I, Homayounpour N, Leto TL (1997) Specificity of p47phox SH3 domain interactions in NADPH oxidase assembly and activation. Mol Cell Biol 17(4):2177–2185 3. El-Benna J et al (2009) p47phox, the phagocyte NADPH oxidase/ NOX2 organizer: structure, phosphorylation and implication in diseases. Exp Mol Med 41(4):217–225 4. Meijles DN et al (2014) Molecular insights of p47phox phosphorylation dynamics in the regulation of NADPH oxidase activation and superoxide production. J Biol Chem 289(33):22759–22770 5. Wientjes FB et al (2001) The NADPH oxidase components p47 (phox) and p40(phox) bind to moesin through their PX domain. Biochem Biophys Res Commun 289(2):382–388 6. Zhan Y et al (2004) p47(phox) PX domain of NADPH oxidase targets cell membrane via moesin-mediated association with the actin cytoskeleton. J Cell Biochem 92(4):795–809 7. Fontayne A et al (2002) Phosphorylation of p47phox sites by PKC alpha, beta II, delta, and zeta: effect on binding to p22phox and on NADPH oxidase activation. Biochemistry 41(24):7743–7750

22

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes

8. Kilpatrick LE et al (2010) Regulation of TNF-induced oxygen radical production in human neutrophils: role of delta-PKC. J Leukoc Biol 87(1):153–164 9. Martyn KD et al (2005) p21-activated kinase (Pak) regulates NADPH oxidase activation in human neutrophils. Blood 106(12): 3962–3969 10. Dang PM et al (2006) A specific p47phox -serine phosphorylated by convergent MAPKs mediates neutrophil NADPH oxidase priming at inflammatory sites. J Clin Invest 116(7):2033–2043 11. Dewas C et al (2000) The mitogen-activated protein kinase extracellular signal-regulated kinase 1/2 pathway is involved in formylmethionyl-leucyl-phenylalanine-induced p47phox phosphorylation in human neutrophils. J Immunol 165(9):5238–5244 12. Chen Q et al (2003) Akt phosphorylates p47phox and mediates respiratory burst activity in human neutrophils. J Immunol 170(10):5302–5308 13. Hoyal CR et al (2003) Modulation of p47PHOX activity by sitespecific phosphorylation: Akt-dependent activation of the NADPH oxidase. Proc Natl Acad Sci U S A 100(9):5130–5135 14. El Benna J et al (1996) Activation of p38 in stimulated human neutrophils: phosphorylation of the oxidase component p47phox by p38 and ERK but not by JNK. Arch Biochem Biophys 334(2): 395–400 15. El Benna J et al (1996) Phosphorylation of the respiratory burst oxidase subunit p47phox as determined by two-dimensional phosphopeptide mapping. Phosphorylation by protein kinase C, protein kinase A, and a mitogen-activated protein kinase. J Biol Chem 271(11):6374–6378 16. Kramer IM et al (1988) The 47-kDa protein involved in the NADPH:O2 oxidoreductase activity of human neutrophils is phosphorylated by cyclic AMP-dependent protein kinase without induction of a respiratory burst. Biochim Biophys Acta 971(2): 189–196 17. Pacquelet S et al (2007) Cross-talk between IRAK-4 and the NADPH oxidase. Biochem J 403(3):451–461 18. Chowdhury AK et al (2005) Src-mediated tyrosine phosphorylation of p47phox in hyperoxia-induced activation of NADPH oxidase and generation of reactive oxygen species in lung endothelial cells. J Biol Chem 280(21):20700–20711 19. Lewis EM et al (2010) Phosphorylation of p22phox on threonine 147 enhances NADPH oxidase activity by promoting p47phox binding. J Biol Chem 285(5):2959–2967 20. Sigal N, Gorzalczany Y, Pick E (2003) Two pathways of activation of the superoxide-generating NADPH oxidase of phagocytes in vitro—distinctive effects of inhibitors. Inflammation 27(3): 147–159 21. Kudlik G et al (2020) Advances in understanding TKS4 and TKS5: molecular scaffolds regulating cellular processes from podosome and invadopodium formation to differentiation and tissue homeostasis. Int J Mol Sci 21(21) 22. Buschman MD et al (2009) The novel adaptor protein Tks4 (SH3PXD2B) is required for functional podosome formation. Mol Biol Cell 20(5):1302–1311 23. Seals DF et al (2005) The adaptor protein Tks5/Fish is required for podosome formation and function, and for the protease-driven invasion of cancer cells. Cancer Cell 7(2):155–165 24. Abram CL et al (2003) The adaptor protein fish associates with members of the ADAMs family and localizes to podosomes of Src-transformed cells. J Biol Chem 278(19):16844–16851 25. Gianni D et al (2009) Novel p47(phox)-related organizers regulate localized NADPH oxidase 1 (Nox1) activity. Sci Signal 2(88):ra54 26. Gianni D, DerMardirossian C, Bokoch GM (2011) Direct interaction between Tks proteins and the N-terminal proline-rich region (PRR) of NoxA1 mediates Nox1-dependent ROS generation. Eur J Cell Biol 90(2-3):164–171

393

27. Kawahara T, Quinn MT, Lambeth JD (2007) Molecular evolution of the reactive oxygen-generating NADPH oxidase (Nox/Duox) family of enzymes. BMC Evol Biol 7:109 28. Diaz B et al (2009) Tks5-dependent, nox-mediated generation of reactive oxygen species is necessary for invadopodia formation. Sci Signal 2(88):ra53 29. Sies H et al (2022) Defining roles of specific reactive oxygen species (ROS) in cell biology and physiology. Nat Rev Mol Cell Biol 23(7):499–515 30. Yazdanpanah B et al (2009) Riboflavin kinase couples TNF receptor 1 to NADPH oxidase. Nature 460(7259):1159–1163 31. Schramm M et al (2014) Riboflavin (vitamin B2 ) deficiency impairs NADPH oxidase 2 (Nox2) priming and defense against Listeria monocytogenes. Eur J Immunol 44(3):728–741 32. Park KJ et al (2012) Death receptors 4 and 5 activate Nox1 NADPH oxidase through riboflavin kinase to induce reactive oxygen species-mediated apoptotic cell death. J Biol Chem 287(5): 3313–3325 33. Hatahet F, Ruddock LW (2009) Protein disulfide isomerase: a critical evaluation of its function in disulfide bond formation. Antioxid Redox Signal 11(11):2807–2850 34. Kozlov G et al (2010) A structural overview of the PDI family of proteins. FEBS J 277(19):3924–3936 35. Pirneskoski A et al (2004) Molecular characterization of the principal substrate binding site of the ubiquitous folding catalyst protein disulfide isomerase. J Biol Chem 279(11):10374–10381 36. Tian G et al (2006) The crystal structure of yeast protein disulfide isomerase suggests cooperativity between its active sites. Cell 124(1):61–73 37. Laurindo FR, Pescatore LA, Fernandes DEC (2012) Protein disulfide isomerase in redox cell signaling and homeostasis. Free Radic Biol Med 52(9):1954–1969 38. Römer RA et al (2016) The flexibility and dynamics of protein disulfide isomerase. Proteins 84(12):1776–1785 39. Okumura M et al (2019) Dynamic assembly of protein disulfide isomerase in catalysis of oxidative folding. Nat Chem Biol 15(5): 499–509 40. Hudson DA, Gannon SA, Thorpe C (2015) Oxidative protein folding: from thiol-disulfide exchange reactions to the redox poise of the endoplasmic reticulum. Free Radic Biol Med 80: 171–182 41. Tanaka LY, Oliveira PVS, Laurindo FRM (2020) Peri/epicellular thiol oxidoreductases as mediators of extracellular redox signaling. Antioxid Redox Signal 33(4):280–307 42. Araujo TLS et al (2017) Protein disulfide isomerase externalization in endothelial cells follows classical and unconventional routes. Free Radic Biol Med 103:199–208 43. Flaumenhaft R, Furie B (2016) Vascular thiol isomerases. Blood 128(7):893–901 44. Wang L, Yu J, Wang CC (2021) Protein disulfide isomerase is regulated in multiple ways: Consequences for conformation, activities, and pathophysiological functions. Bioessays 43(3): e2000147 45. Xu S, Sankar S, Neamati N (2014) Protein disulfide isomerase: a promising target for cancer therapy. Drug Discov Today 19(3): 222–240 46. Xiong B et al (2020) Protein disulfide isomerase in cardiovascular disease. Exp Mol Med 52(3):390–399 47. Laurindo FR et al (2008) Novel role of protein disulfide isomerase in the regulation of NADPH oxidase activity: pathophysiological implications in vascular diseases. Antioxid Redox Signal 10(6): 1101–1113 48. Janiszewski M et al (2005) Regulation of NAD(P)H oxidase by associated protein disulfide isomerase in vascular smooth muscle cells. J Biol Chem 280(49):40813–40819

394 49. Pescatore LA et al (2012) Protein disulfide isomerase is required for platelet-derived growth factor-induced vascular smooth muscle cell migration, Nox1 NADPH oxidase expression, and RhoGTPase activation. J Biol Chem 287(35):29290–29300 50. Santos CX et al (2009) Protein disulfide isomerase (PDI) associates with NADPH oxidase and is required for phagocytosis of Leishmania chagasi promastigotes by macrophages. J Leukoc Biol 86(4):989–998 51. de A Paes AM et al (2011) Protein disulfide isomerase redoxdependent association with p47(phox): evidence for an organizer role in leukocyte NADPH oxidase activation. J Leukoc Biol 90(4): 799–810 52. Cho J (2013) Protein disulfide isomerase in thrombosis and vascular inflammation. J Thromb Haemost 11(12):2084–2091 53. Fernandes DC et al (2009) Protein disulfide isomerase overexpression in vascular smooth muscle cells induces spontaneous preemptive NADPH oxidase activation and Nox1 mRNA expression: effects of nitrosothiol exposure. Arch Biochem Biophys 484(2):197–204 54. Fernandes DC et al (2021) PDIA1 acts as master organizer of NOX1/NOX4 balance and phenotype response in vascular smooth muscle. Free Radic Biol Med 162:603–614 55. Gimenez M et al (2019) Redox activation of Nox1 (NADPH oxidase 1) involves an intermolecular disulfide bond between protein disulfide isomerase and p47. Arterioscler Thromb Vasc Biol 39(2):224–236 56. Tanaka LY et al (2019) Peri/epicellular protein disulfide isomeraseA1 acts as an upstream organizer of cytoskeletal mechanoadaptation in vascular smooth muscle cells. Am J Physiol Heart Circ Physiol 316(3):H566–H579 57. Soares Moretti AI, Martins Laurindo FR (2017) Protein disulfide isomerases: Redox connections in and out of the endoplasmic reticulum. Arch Biochem Biophys 617:106–119 58. Moretti AIS et al (2017) Conserved gene microsynteny unveils functional interaction between protein disulfide isomerase and Rho guanine-dissociation inhibitor families. Sci Rep 7(1):17262 59. De Bessa TC et al (2019) Subverted regulation of Nox1 NADPH oxidase-dependent oxidant generation by protein disulfide isomerase A1 in colon carcinoma cells with overactivated KRas. Cell Death Dis 10(2):143 60. Bechor E et al (2015) The dehydrogenase region of the NADPH oxidase component Nox2 acts as a protein disulfide isomerase (PDI) resembling PDIA3 with a role in the binding of the activator protein p67 (phox.). Front Chem 3:3 61. Tanaka LY et al (2016) Peri/epicellular protein disulfide isomerase sustains vascular lumen caliber through an anticonstrictive remodeling effect. Hypertension 67(3):613–622 62. Brandes RP, Weissmann N, Schröder K (2014) Nox family NADPH oxidases in mechano-transduction: mechanisms and consequences. Antioxid Redox Signal 20(6):887–898 63. Sobierajska K et al (2014) Protein disulfide isomerase directly interacts with β-actin Cys374 and regulates cytoskeleton reorganization. J Biol Chem 289(9):5758–5773 64. Hsiai TK et al (2007) Hemodynamics influences vascular peroxynitrite formation: Implication for low-density lipoprotein apo-B-100 nitration. Free Radic Biol Med 42(4):519–529 65. Marschall R, Tudzynski P (2017) The protein disulfide isomerase of. Front Microbiol 8:960 66. Laurindo FR, Araujo TL, Abrahão TB (2014) Nox NADPH oxidases and the endoplasmic reticulum. Antioxid Redox Signal 20(17):2755–2775 67. Prior KK et al (2016) The endoplasmic reticulum chaperone calnexin is a NADPH oxidase NOX4 interacting protein. J Biol Chem 291(13):7045–7059 68. Zhang X et al (2019) Redox signals at the ER-mitochondria interface control melanoma progression. EMBO J 38(15):e100871

T. C. de Bessa and F. R. M. Laurindo 69. George G et al (2020) EDEM2 stably disulfide-bonded to TXNDC11 catalyzes the first mannose trimming step in mammalian glycoprotein ERAD. Elife 9 70. Wang D et al (2005) Identification of a novel partner of duox: EFP1, a thioredoxin-related protein. J Biol Chem 280(4): 3096–3103 71. De Deken X et al (2002) Characterization of ThOX proteins as components of the thyroid H(2)O(2)-generating system. Exp Cell Res 273(2):187–196 72. Arevalo JA, Vázquez-Medina JP (2018) The role of peroxiredoxin 6 in cell signaling. Antioxidants (Basel) 7(12) 73. Kwon J et al (2016) Peroxiredoxin 6 (Prdx6) supports NADPH oxidase1 (Nox1)-based superoxide generation and cell migration. Free Radic Biol Med 96:99–115 74. Krishnaiah SY et al (2013) p67(phox) terminates the phospholipase A(2)-derived signal for activation of NADPH oxidase (NOX2). FASEB J 27(5):2066–2073 75. Chatterjee S et al (2011) Peroxiredoxin 6 phosphorylation and subsequent phospholipase A2 activity are required for agonistmediated activation of NADPH oxidase in mouse pulmonary microvascular endothelium and alveolar macrophages. J Biol Chem 286(13):11696–11706 76. Hasegawa J et al (2011) SH3YL1 regulates dorsal ruffle formation by a novel phosphoinositide-binding domain. J Cell Biol 193(5): 901–916 77. Kobayashi M et al (2014) Dock4 forms a complex with SH3YL1 and regulates cancer cell migration. Cell Signal 26(5):1082–1088 78. Yoo JY et al (2020) LPS-induced acute kidney injury is mediated by Nox4-SH3YL1. Cell Rep 33(3):108245 79. Nguyen MV et al (2013) Quinone compounds regulate the level of ROS production by the NADPH oxidase Nox4. Biochem Pharmacol 85(11):1644–1654 80. Yuan S et al (2021) Cooperation between CYB5R3 and NOX4 via coenzyme Q mitigates endothelial inflammation. Redox Biol 47: 102166 81. Ramadass M, Catz SD (2016) Molecular mechanisms regulating secretory organelles and endosomes in neutrophils and their implications for inflammation. Immunol Rev 273(1):249–265 82. McAdara Berkowitz JK et al (2001) JFC1, a novel tandem C2 domain-containing protein associated with the leukocyte NADPH oxidase. J Biol Chem 276(22):18855–18862 83. Munafó DB et al (2007) Rab27a is a key component of the secretory machinery of azurophilic granules in granulocytes. Biochem J 402(2):229–239 84. Ramadass M et al (2019) The trafficking protein JFC1 regulates Rac1-GTP localization at the uropod controlling neutrophil chemotaxis and in vivo migration. J Leukoc Biol 105(6):1209–1224 85. Catz SD (2008) Characterization of Rab27a and JFC1 as constituents of the secretory machinery of prostate-specific antigen in prostate carcinoma cells. Methods Enzymol 438:25–40 86. Johnson JL et al (2010) Rab27a and Rab27b regulate neutrophil azurophilic granule exocytosis and NADPH oxidase activity by independent mechanisms. Traffic 11(4):533–547 87. Ejlerskov P et al (2012) NADPH oxidase is internalized by clathrin-coated pits and localizes to a Rab27A/B GTPase-regulated secretory compartment in activated macrophages. J Biol Chem 287(7):4835–4852 88. Johnson JL et al (2016) Identification of neutrophil exocytosis inhibitors (nexinhibs), small molecule inhibitors of neutrophil exocytosis and inflammation: druggability of the small GTPase Rab27a. J Biol Chem 291(50):25965–25982 89. Johnson JL et al (2012) Vesicular trafficking through cortical actin during exocytosis is regulated by the Rab27a effector JFC1/Slp1 and the RhoA-GTPase-activating protein Gem-interacting protein. Mol Biol Cell 23(10):1902–1916

22

Proteins Cross-talking with Nox Complexes: The Social Life of Noxes

90. Thomas DC et al (2017) Eros is a novel transmembrane protein that controls the phagocyte respiratory burst and is essential for innate immunity. J Exp Med 214(4):1111–1128 91. Thomas DC et al (2019) EROS/CYBC1 mutations: decreased NADPH oxidase function and chronic granulomatous disease. J Allergy Clin Immunol 143(2):782–785.e1 92. Roos D et al (2021) Hematologically important mutations: X-linked chronic granulomatous disease (fourth update). Blood Cells Mol Dis 90:102587 93. Arnadottir GA et al (2018) A homozygous loss-of-function mutation leading to CYBC1 deficiency causes chronic granulomatous disease. Nat Commun 9(1):4447 94. Perez-Heras I et al (2021) HSCT in two brothers with CGD arising from mutations in CYBC1 corrects the defect in neutrophil function. Clin Immunol 229:108799 95. Ryoden Y et al (2020) Functional expression of the P2X7 ATP receptor requires Eros. J Immunol 204(3):559–568 96. Boudreau E et al (1997) The chloroplast ycf3 and ycf4 open reading frames of Chlamydomonas reinhardtii are required for the accumulation of the photosystem I complex. EMBO J 16(20): 6095–6104 97. Chen F et al (2011) Hsp90 regulates NADPH oxidase activity and is necessary for superoxide but not hydrogen peroxide production. Antioxid Redox Signal 14(11):2107–2119 98. Chen F et al (2015) Nox5 stability and superoxide production is regulated by C-terminal binding of Hsp90 and CO-chaperones. Free Radic Biol Med 89:793–805 99. Chen F et al (2012) Opposing actions of heat shock protein 90 and 70 regulate nicotinamide adenine dinucleotide phosphate oxidase stability and reactive oxygen species production. Arterioscler Thromb Vasc Biol 32(12):2989–2999 100. Ahmad Mokhtar AMB et al (2021) A complete survey of RhoGDI targets reveals novel interactions with atypical small GTPases. Biochemistry 60(19):1533–1551 101. Pick E (2014) Role of the Rho GTPase Rac in the activation of the phagocyte NADPH oxidase: outsourcing a key task. Small GTPases 5:e27952 102. Hordijk PL (2006) Regulation of NADPH oxidases: the role of Rac proteins. Circ Res 98(4):453–462 103. Garcia-Mata R, Boulter E, Burridge K (2011) The ‘invisible hand’: regulation of RHO GTPases by RHOGDIs. Nat Rev Mol Cell Biol 12(8):493–504 104. DerMardirossian C, Schnelzer A, Bokoch GM (2004) Phosphorylation of RhoGDI by Pak1 mediates dissociation of Rac GTPase. Mol Cell 15(1):117–127 105. Zhang B et al (2001) Oligomerization of Rac1 gtpase mediated by the carboxyl-terminal polybasic domain. J Biol Chem 276(12): 8958–8967 106. Carol RJ et al (2005) A RhoGDP dissociation inhibitor spatially regulates growth in root hair cells. Nature 438(7070):1013–1016 107. Kreck ML et al (1996) Membrane association of Rac is required for high activity of the respiratory burst oxidase. Biochemistry 35(49): 15683–15692 108. Gorzalczany Y et al (2000) Targeting of Rac1 to the phagocyte membrane is sufficient for the induction of NADPH oxidase assembly. J Biol Chem 275(51):40073–40081 109. Ugolev Y et al (2006) Liposomes comprising anionic but not neutral phospholipids cause dissociation of Rac(1 or 2) x RhoGDI complexes and support amphiphile-independent NADPH oxidase activation by such complexes. J Biol Chem 281(28):19204–19219 110. Ugolev Y et al (2008) Dissociation of Rac1(GDP).RhoGDI complexes by the cooperative action of anionic liposomes containing phosphatidylinositol 3,4,5-trisphosphate, Rac guanine nucleotide exchange factor, and GTP. J Biol Chem 283(32): 22257–22271

395

111. Grizot S et al (2001) Crystal structure of the Rac1-RhoGDI complex involved in nadph oxidase activation. Biochemistry 40(34): 10007–10013 112. Schaefer A, Reinhard NR, Hordijk PL (2014) Toward understanding RhoGTPase specificity: structure, function and local activation. Small GTPases 5(2):6 113. Price MO et al (2002) Rac activation induces NADPH oxidase activity in transgenic COSphox cells, and the level of superoxide production is exchange factor-dependent. J Biol Chem 277(21): 19220–19228 114. Utomo A et al (2006) Vav proteins in neutrophils are required for FcgammaR-mediated signaling to Rac GTPases and nicotinamide adenine dinucleotide phosphate oxidase component p40(phox). J Immunol 177(9):6388–6397 115. Liu Y et al (2013) A novel pathway spatiotemporally activates Rac1 and redox signaling in response to fluid shear stress. J Cell Biol 201(6):863–873 116. Mizrahi A et al (2005) Activation of the phagocyte NADPH oxidase by Rac Guanine nucleotide exchange factors in conjunction with ATP and nucleoside diphosphate kinase. J Biol Chem 280(5):3802–3811 117. Park HS et al (2004) Sequential activation of phosphatidylinositol 3-kinase, beta Pix, Rac1, and Nox1 in growth factor-induced production of H2O2. Mol Cell Biol 24(10):4384–4394 118. Kaito Y et al (2014) Nox1 activation by βPix and the role of Ser-340 phosphorylation. FEBS Lett 588(11):1997–2002 119. Bretscher A et al (2000) ERM-Merlin and EBP50 protein families in plasma membrane organization and function. Annu Rev Cell Dev Biol 16:113–143 120. Al Ghouleh I et al (2016) Binding of EBP50 to Nox organizing subunit p47phox is pivotal to cellular reactive species generation and altered vascular phenotype. Proc Natl Acad Sci U S A 113(36): E5308–E5317 121. Weissbach L et al (1994) Identification of a human rasGAP-related protein containing calmodulin-binding motifs. J Biol Chem 269(32):20517–20521 122. Bashour AM et al (1997) IQGAP1, a Rac- and Cdc42-binding protein, directly binds and cross-links microfilaments. J Cell Biol 137(7):1555–1566 123. Ikeda S et al (2005) IQGAP1 regulates reactive oxygen speciesdependent endothelial cell migration through interacting with Nox2. Arterioscler Thromb Vasc Biol 25(11):2295–2300 124. Hernandes MS, Lassègue B, Griendling KK (2017) Polymerase δ-interacting protein 2: a multifunctional protein. J Cardiovasc Pharmacol 69(6):335–342 125. Lyle AN et al (2009) Poldip2, a novel regulator of Nox4 and cytoskeletal integrity in vascular smooth muscle cells. Circ Res 105(3):249–259 126. Datla SR et al (2014) Poldip2 controls vascular smooth muscle cell migration by regulating focal adhesion turnover and force polarization. Am J Physiol Heart Circ Physiol 307(7):H945–H957 127. Vukelic S et al (2018) NOX4 (NADPH oxidase 4) and Poldip2 (polymerase δ-interacting protein 2) induce filamentous actin oxidation and promote its interaction with vinculin during integrinmediated cell adhesion. Arterioscler Thromb Vasc Biol 38(10): 2423–2434 128. Huff LP et al (2019) Polymerase-δ-interacting protein 2 activates the RhoGEF epithelial cell transforming sequence 2 in vascular smooth muscle cells. Am J Physiol Cell Physiol 316(5):C621– C631 129. Sutliff RL et al (2013) Polymerase delta interacting protein 2 sustains vascular structure and function. Arterioscler Thromb Vasc Biol 33(9):2154–2161 130. Datla SR et al (2019) Poldip2 knockdown inhibits vascular smooth muscle proliferation and neointima formation by regulating the expression of PCNA and p21. Lab Invest 99(3):387–398

396 131. Paredes F et al (2020) Mitochondrial protein Poldip2 (polymerase delta interacting protein 2) controls vascular smooth muscle differentiated phenotype by O-linked GlcNAc (N-acetylglucosamine) transferase-dependent inhibition of a ubiquitin proteasome system. Circ Res 126(1):41–56 132. Paredes F et al (2018) Poldip2 is an oxygen-sensitive protein that controls PDH and αKGDH lipoylation and activation to support metabolic adaptation in hypoxia and cancer. Proc Natl Acad Sci U S A 115(8):1789–1794 133. Hernandes MS et al (2018) Polymerase delta-interacting protein 2 deficiency protects against blood-brain barrier permeability in the ischemic brain. J Neuroinflamm 15(1):45 134. Forrester SJ et al (2019) Poldip2 deficiency protects against lung edema and vascular inflammation in a model of acute respiratory distress syndrome. Clin Sci (Lond) 133(2):321–334 135. Nicolae CM et al (2014) The ADP-ribosyltransferase PARP10/ ARTD10 interacts with proliferating cell nuclear antigen (PCNA) and is required for DNA damage tolerance. J Biol Chem 289(19): 13627–13637 136. Warbrick E (1998) PCNA binding through a conserved motif. Bioessays 20(3):195–199 137. Witko-Sarsat V et al (2010) Proliferating cell nuclear antigen acts as a cytoplasmic platform controlling human neutrophil survival. J Exp Med 207(12):2631–2645 138. Ohayon D et al (2016) Cytoplasmic proliferating cell nuclear antigen connects glycolysis and cell survival in acute myeloid leukemia. Sci Rep 6:35561 139. Ohayon D et al (2019) Cytosolic PCNA interacts with p47phox and controls NADPH oxidase NOX2 activation in neutrophils. J Exp Med 216(11):2669–2687 140. Krummrei U et al (1995) Cyclophilin-A is a zinc-dependent DNA binding protein in macrophages. FEBS Lett 371(1):47–51 141. Nigro P, Pompilio G, Capogrossi MC (2013) Cyclophilin A: a key player for human disease. Cell Death Dis 4:e888 142. Jin ZG et al (2000) Cyclophilin A is a secreted growth factor induced by oxidative stress. Circ Res 87(9):789–796 143. Satoh K et al (2009) Cyclophilin A enhances vascular oxidative stress and the development of angiotensin II-induced aortic aneurysms. Nat Med 15(6):649–656 144. Satoh K et al (2011) Cyclophilin A promotes cardiac hypertrophy in apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol 31(5):1116–1123 145. Soe NN et al (2013) Cyclophilin A is required for angiotensin II-induced p47phox translocation to caveolae in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 33(9):2147–2153

T. C. de Bessa and F. R. M. Laurindo 146. Dos Santos GP et al (2021) Cyclophilin 19 secreted in the host cell cytosol by Trypanosoma cruzi promotes ROS production required for parasite growth. Cell Microbiol 23(4):e13295 147. Boussetta T et al (2010) The prolyl isomerase Pin1 acts as a novel molecular switch for TNF-alpha-induced priming of the NADPH oxidase in human neutrophils. Blood 116(26):5795–5802 148. Makni-Maalej K et al (2012) The TLR7/8 agonist CL097 primes N-formyl-methionyl-leucyl-phenylalanine-stimulated NADPH oxidase activation in human neutrophils: critical role of p47phox phosphorylation and the proline isomerase Pin1. J Immunol 189(9):4657–4665 149. Litchfield DW (2003) Protein kinase CK2: structure, regulation and role in cellular decisions of life and death. Biochem J 369 (Pt 1):1–15 150. Park HS et al (2001) Phosphorylation of the leucocyte NADPH oxidase subunit p47(phox) by casein kinase 2: conformationdependent phosphorylation and modulation of oxidase activity. Biochem J 358(Pt 3):783–790 151. Kil IS et al (2015) S-Nitrosylation of p47(phox) enhances phosphorylation by casein kinase 2. Redox Rep 20(5):228–233 152. Kim GS et al (2009) CK2 is a novel negative regulator of NADPH oxidase and a neuroprotectant in mice after cerebral ischemia. J Neurosci 29(47):14779–14789 153. Liu J et al (2013) Identification and characterization of a unique leucine-rich repeat protein (LRRC33) that inhibits Toll-like receptor-mediated NF-κB activation. Biochem Biophys Res Commun 434(1):28–34 154. Noubade R et al (2014) NRROS negatively regulates reactive oxygen species during host defence and autoimmunity. Nature 509(7499):235–239 155. Wong K et al (2017) Mice deficient in NRROS show abnormal microglial development and neurological disorders. Nat Immunol 18(6):633–641 156. Ma W et al (2019) LRRC33 is a novel binding and potential regulating protein of TGF-β1 function in human acute myeloid leukemia cells. PLoS One 14(10):e0213482 157. Anglesio MS et al (2004) Differential expression of a novel ankyrin containing E3 ubiquitin-protein ligase, Hace1, in sporadic Wilms’ tumor versus normal kidney. Hum Mol Genet 13(18):2061–2074 158. Daugaard M et al (2013) Hace1 controls ROS generation of vertebrate Rac1-dependent NADPH oxidase complexes. Nat Commun 4:2180 159. Torrino S et al (2011) The E3 ubiquitin-ligase HACE1 catalyzes the ubiquitylation of active Rac1. Dev Cell 21(5):959–965 160. Diebold BA et al (2004) Antagonistic cross-talk between Rac and Cdc42 GTPases regulates generation of reactive oxygen species. J Biol Chem 279(27):28136–28142

Part V Non-Mammalian NADPH Oxidases

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum

23

Laurence Aubry and Bernard Lardy

Abstract

Dictyostelium discoideum, also known as social amoeba, is at the boundary between unicellular and multicellular life. This protist is a professional phagocyte in its vegetative unicellular stage, feeding on bacterial preys, and it enters a multicellular phase in response to starvation. Dictyostelium and human phagocytic cells share several unique functions and this similarity is supported by a high degree of conservation between their proteomes. Three homologs of the large subunit of NADPH oxidase (NoxA, B, C), a homolog of the small subunit p22phox and a p67phox-like factor have been identified. In this chapter, roles involving Dictyostelium NADPH oxidase activities are described: development and cell differentiation, intraphagosomal bacterial killing and formation of DNA-based extracellular traps by sentinel cells, a subtype of amoebal cells that support ancestral innate immunity. The exploration of NOX functions and regulation is still ongoing, and their study in alternative models is needed for a comprehensive and integrated view of the contribution of NADPH oxidases to key biological processes. In this context, the amoeba Dictyostelium is a particularly attractive model to enrich our current understanding of this family of enzymes. Keywords

Dictyostelium · NADPH oxidase · Development · Phagocytosis · NETs

L. Aubry University Grenoble Alpes, CEA, INSERM, IRIG, BGE, Grenoble, France B. Lardy (✉) University Grenoble Alpes, TIMC, UMR 5525 CNRS, GREPI, Grenoble, France e-mail: [email protected]

1

Introduction

NADPH oxidases are membrane enzymes specialized in the production of reactive oxygen species (ROS) that catalyse electron transfer across the membrane from cytosolic NADPH to molecular oxygen. This function was long thought to be restricted to the phagocyte NADPH oxidase NOX2. In the last decades, the discovery and study of novel NOX isoforms in vertebrates and homologs in a wide variety of organisms, including plants, fungi and cellular slime molds has largely broadened the landscape of processes modulated by the activity of NOXs. Despite their shared property of ROS production, members of the NADPH oxidase family differ in their level of activity, modes of activation/regulation, type of ROS generated (superoxide (O2•–) vs. hydrogen peroxide (H2O2)), expression pattern in multicellular organisms and physiological functions. Their characterization is still ongoing, and their study in alternative models contributes directly to the building of a complete and integrated knowledge of the role of NADPH oxidases in the biology of eukaryotes, with the addition of the evolutionary component. In this context, the amoeba Dictyostelium discoideum whose genome encodes three NOX proteins, NoxA, B and C, and homologs of the p22phox and p67phox proteins, is a particularly attractive model to enrich our current understanding of this family of enzymes.

2

The Model Organism Dictyostelium discoideum

The cellular slime mold Dictyostelium discoideum (called thereafter Dictyostelium) is a haploid, eukaryotic organism that belongs to the Dictyostelides, in the Amoebozoa phylum. This phylum diverged from the main branch leading to the animals shortly after the plant-animal split and before the separation of the fungal lineage [1, 2]. A striking feature of Dictyostelium is its capacity to transition from the unicellular

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_23

399

400

L. Aubry and B. Lardy

state to a multicellular state depending on nutrient availability. Indeed, under conditions of abundant nutrients, Dictyostelium exists as unicellular amoebae, dividing by mitosis as long as nutritive needs are fulfilled. In its natural habitat, this organism feeds on soil microorganisms that are engulfed by phagocytosis. When food becomes scarce, this growth phase, also called the vegetative phase, gives way to a multicellular phase that starts with the aggregation of about 100,000 individual starving cells. Nutrient depletion induces a tightly coordinated developmental program leading to the formation, within 24 h of the onset of starvation, of a multicellular structure consisting of a long stalk supporting a mass of spores, the fruiting body (Fig. 23.1) [3, 4]. Spores ensure the survival of the species to the absence of nutrients. After their dispersal, they will germinate, provided that environmental conditions are favourable, releasing each an individual amoeba and starting again the unicellular phase of the cycle. In contrast, although essential for the formation of the fruiting body and differentiation of spores, stalk cells will all Chemotactic aggregation

8h

die in the late phase of development through a highlyregulated cell-death process. This collective behaviour and altruism of stalk cells earned Dictyostelium its name of social amoeba. Since the first studies in the mid-1930s, Dictyostelium has emerged as an attractive and valuable model organism for exploring the molecular mechanisms underlying fundamental eukaryotic biological processes. Several features have contributed to the success of Dictyostelium as a model system. The amoeba optimally grows at temperature around 21°C and is harmless to humans. While wild-type strains feed on bacteria, axenic strains have been isolated that are able to grow in the absence of bacteria in complex or defined liquid medium, thereby simplifying their culture conditions in suspension. With a doubling time between 2 and 12 h depending on the strains and source of nutrients, large quantities of cells can be readily obtained, facilitating biochemical approaches requiring large amounts of material. Cell division is mostly restricted to the vegetative phase.

10h Emergence of cell-types Prestalk cells

Cell sorting Mound

Starvation

(around 100 000 cells)

Prespore cells

14h

1st

finger

Growth

UNICELLULAR PHASE

18h

24h

(mitotic division)

Slug

16h

MULTICELLULAR PHASE

PstA

22h

PstAB

Upper cup

PstO ALC

Spore mass PstB

Lower cup

Prestalk region (anterior)

Spore germination

2nd finger Spore dispersal

Prespore region (posterior)

Stalk

Culmination Basal disc

Fruiting-body Terminal differentiation

Fig. 23.1 Dictyostelium discoideum developmental cycle. Dictyostelium grows as a single-celled amoeba, feeding by phagocytosis of soil bacteria. Starvation induces the aggregation of individual cells by chemotaxis. Resulting mounds undergo successive steps of morphogenesis leading to the formation of migrating slugs that evolve into sporecontaining fruiting bodies. In appropriate conditions, spores germinate, releasing each a single amoeba able to reinitiate the unicellular phase of the life cycle. Two distinct cell populations differentiate in the multicellular structure, emerging as early as mound stage as precursor prespore

and prestalk cells. Prespore cells will eventually form spores while prestalk cells will form all the support structures of the fruiting body: the stalk that lifts the spore mass from the ground, the basal disc that ensures the anchoring of the fruiting body to the substratum, and the upper and lower cups that hold the spore mass together. The prestalk population includes distinct subtypes among which prestalk A, B, AB, O and anterior like cells (ALC) that differentially contribute to the multicellular structure

23

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum

The transition from the unicellular to the multicellular state is easily controlled by simple transfer of cells to non-nutritive buffer. The developmental cycle is achieved in 24 h, giving rapidly access to relatively simple multicellular forms. Of major importance, Dictyostelium is amenable to molecular genetic manipulations. The Dictyostelium genome (34 Mb) was sequenced and published in 2005, revealing a repertoire of about 12,500 genes [1]. Various genetic tools can be applied to the organism—including the Cre/LoxP technique and more recently, Crispr/Cas9-mediated genome editing [5– 7]—allowing targeted gene disruption and expression of mutated or tagged versions of proteins. In this context, the haploidy of the genome is an asset, easing the generation of knock-out and knock-in mutants and analysis of resulting phenotypes. Development of the REMI (restriction enzymemediated integration) technique for random insertion mutagenesis has been pivotal in the discovery and characterization of gene functions in the amoeba [8]. Recent improvements of the REMI tool leading to the REMI-seq technology have enabled the construction of a genome-wide library of insertional mutants and the high throughput identification of insertion sites, and the possibility of parallel phenotypic analyses, opening new perspectives to Dictyostelium users [9]. Finally, Dictybase (http://dictybase.org/), the online bioinformatics database for Dictyostelium discoideum (and other Dictyostelides) and the Dicty Stock Center (https://dictycr. org/), a central repository for Dictyostelium strains, plasmids and antibodies represent an invaluable resource for the Dictyostelium community. Decades of work on Dictyostelium attest to the strength of the model to dissect fundamental biological processes such as motility, endocytosis/phagocytosis, chemotactism, cell differentiation, morphogenesis and programmed cell death [10–13]. Recognized by the NIH as a biomedical model, Dictyostelium has proven to be a valid alternative model to study the molecular bases of a variety of human pathologies while avoiding the ethical considerations associated with mammalian models [14, 15]. Among others, it has successfully contributed to deciphering the role of proteins linked to human neurological disorders such as Alzheimer’s and Huntington’s diseases and ceroid neuronal lipofuscinosis. It is also a relevant model to investigate host-pathogen interactions and microbial infection, for instance by Legionella pneumophila, Pseudomonas aeruginosa and Mycobacterium spp. [11, 13, 16].

3

The Amoebal Repertoire of NADPH Oxidases

In humans, the NOX family includes seven members, NOX 1–5 and DUOX 1–2. The first characterized member of the human NOX family was the gp91phox isoform—now called

401

NOX2—encoded by the CYBB gene. It is primarily expressed in phagocytes and responsible for the ROS-mediated killing of microbes [17]. Its major role in innate immunity and inflammation is well illustrated by the susceptibility to bacterial infection of patients suffering from chronic granulomatous disease (CGD) and carrying a defective NOX2 activity. NOX2 consists of an N-terminal ferricreductase domain that comprises six transmembrane (TM) helices and a cytoplasmic C-terminal dehydrogenase domain harbouring key binding sites for NADPH and for FAD, the initial acceptor of the electrons resulting from NADPH oxidation (Fig. 23.2). The transmembrane domain mediates the electron transport across the membrane from cytoplasmic FAD to the oxygen molecule through two heme groups coordinated by two pairs of histidine moieties within helices III and V (H101, 115, 209, and 222 in human NOX2). These domains are conserved in all the members of the NADPH oxidase family. NOX1, NOX3 and NOX4 share a similar overall structural organization with NOX2 while NOX5 differs from NOX2 by the presence of four calciumbinding EF-hand motifs in its N-terminal cytoplasmic extremity and DUOX1 and DUOX2 by two EF-hand motifs in addition to a transmembrane domain and an extracellular peroxidase domain at the very N-terminus of the proteins [18].

3.1

Dictyostelium Genome Encodes 3 Members of the NOX Family: NoxA, NoxB and NoxC

The amazing efficiency of Dictyostelium in phagocytosing and feeding on bacteria prompted us, almost 20 years ago, to explore the amoebal repertoire of NADPH oxidases [19]. We identified three homologs of NOX-encoding genes using 3′ and 5′ RACE, genomic and RT-PCR experiments, called noxA (GenBank accession number AF123275, Dictybase DDB_G0289653), noxB (AY221173, DDB_G0287101) and noxC (AY224390, DDB_G0291117) [19]. The proteins encoded by noxA, B and C genes are 517, 698 and 1142 amino acid long respectively (Fig. 23.3a). All three homologs contain a core domain harboring the canonical features of human NOX2: six predicted transmembrane α-helices, binding sites for flavin and pyridine nucleotides and the pairs of histidine residues involved in heme binding in the third and fifth transmembrane segments (Figs. 23.3a and 23.4). This core region in NoxA, B and C shares 66% (38%), 67% (34%) and 55% (25%) homologies (identities) with human NOX2, respectively (Fig. 23.4). Several missense mutations have been identified in CGD patients, resulting in normal levels but nonfunctional NOX2containing NADPH oxidase complexes [20]. Out of the 15 residues mutated in the human NOX2 protein (X-linked

402

L. Aubry and B. Lardy

.O2

Fig. 23.2 Structural organization of the core region of NOX enzymes. All NOX proteins present a ferric-reductase domain containing six transmembrane helices and a dehydrogenase domain with binding sites for FAD and NADPH. Transmembrane helices III and V both contain two histidines, involved in heme (Fe) binding. Superoxide is generated by electron transfer across the membrane from cytosolic NADPH to oxygen in intraluminal or extracellular environment

O2

Ferric-reductase domain C

A

Cytosol

Fe

II

I

e-

III IV

E

V VI

Fe D

B

eNADPH

FAD

e-

Dehydrogenase domain

B

A NoxA

1

4 5 6

2 3

NoxB

1

517 aa 2

3

4 5 6

NoxC

698 EF EF

1

2

3

4 5 6

1142

CybA 1 2 3 4 118 T T T

T

NoxA

PB1

W 604

NoxC

NoxB

noxA

Transcript abundance

NcfA

noxB

noxC

EF EF

Rac NcfA

cybA

? CybA

Ca2+

ncfA 0 2

4

6

8 10 12 16 20 24h

Development

Fig. 23.3 The NOX machinery of Dictyostelium. (a) Domain organization of the amoebal NOX proteins, the p22phox-like protein CybA and the p67phox-like protein NcfA. In contrast to NoxA and NoxB, NoxC presents two EF-hand domains in its N-terminal region that may respond to calcium. Despite the prediction of four putative transmembrane domains, the protein CybA was represented with two membrane pillars

by analogy with human p22phox whose transmembrane domain organization is still unclear. T, tetratricopeptide motif; EF, EF-hand domain; W, WW-domain; PB1, PB1 domain, numbered boxes: putative transmembrane domains. (b) Expression profile of noxA, B, C, cybA and ncfA during growth and starvation-induced development. Transcript levels were extracted from DictyExpress [38, 39]

23

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum

Fig. 23.4 Protein sequence alignment of Dictyostelium NoxA, B and C with human NOX2 and NOX5. Alignment was done using ClustalW program. Residues identical or with a conserved substitution were colored in black or grey respectively. The six transmembrane domains

403

(TM1–6) and sequences involved in FAD and NADPH binding are boxed. Orange spots indicate residues found mutated in CGD X+ patients

404

L. Aubry and B. Lardy

Fig. 23.4 (continued)

CGD variants), 11 are conserved in the three Dictyostelium NOX proteins (R54, A57, H338, T341, G389, G408, G413, P415, T481, C537 and E568 in human NOX2) (Fig. 23.4, orange spots) and 3 residues (H303, C369 and D500 in human NOX2) that are not conserved in Dictyostelium NoxA, B or/and C, are also not conserved in human NOX4 or NOX5. Sequence analysis of Dictyostelium NOXs also indicates that the E-loop joining TM 5 and TM 6 is much shorter in all amoebal NOXs compared to NOX2 (Fig. 23.4). This is also observed in human NOX5 and DUOX as well as in the Arabidopsis thaliana NOX homologs AtRbohA-J. In addition to the core domain, NoxB and NoxC both present an N-terminal extension of 166 and 560 amino acids (aa) respectively. These extensions contain homopolymeric stretches of asparagine residues, an extremely frequent feature of Dictyostelium proteins [1, 21]. In this N-terminal

region, NoxB presents no other characteristic domain. In contrast, NoxC exhibits two predicted EF-hand motifs (aa 466–494 and aa 510–538), indicating a relatedness with NOX5, also supported by the sequence of the core domain [19]. As NoxC, all NOX homologs of A. thaliana (AtRboh A-J) harbor two EF-hands N-terminally to the core domain, and not four as NOX5. The Rboh family of NOX proteins is regulated by an elevation of cytoplasmic calcium or by phosphorylation through a calcium-dependent protein kinase [22]. Along the same line, NOX5 is activated in response to calcium that causes a conformational change of the EF3-EF4 pair allowing its binding to the dehydrogenase domain and subsequent enzyme activation [23, 24]. In the case of Dictyostelium NoxC, whether the EF-hands are bona fide calcium-binding motifs has not yet been established but in view of the shared structural organisation with calcium-

23

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum

regulated NOXs, a calcium-dependent activation is expected as well for this amoebal NOX.

3.2

Dd-CybA, the Amoebal Homolog of the p22phox Subunit

Except for NOX5 and DUOXs, human NOX proteins function as heterodimers with the integral membrane protein p22phox, encoded by the CYBA gene [25]. Despite numerous studies, the structure and topology (such as the exact number of transmembrane helices) of p22phox are still a matter of debate [26, 27]. Binding of p22phox to NOX2 contributes to the stabilization of the complex by preventing proteasomal degradation of individual subunits and to enzyme activation through the recruitment of the cytosolic regulator p47phox [25, 28, 29]. Binding of p47phox by p22phox involves a proline-rich region (PPR) (151PSNPPPRPP) in p22phox C-terminal tail that behaves as a high affinity binding site for p47phox double src homology 3 (SH3) motif [30] (Fig. 23.5). We have cloned the gene encoding the Dictyostelium homolog of mammalian p22phox, called CybA (GenBank AY221170, Dictybase DDB_G0267460). Sequence alignment with human p22phox is shown in Fig. 23.5. Analysis of Dd-CybA using Phobius (https://phobius.sbc.su.se/) and PHYRE2 (http://www.sbg.bio.ic.ac.uk/phyre2/) web servers suggests the presence of four hydrophobic helices located in regions conserved with human p22phox, that could represent transmembrane anchoring pillars. A striking difference between Dd-CybA and its human counterpart resides in the

Fig. 23.5 Protein sequence alignment of human p22phox and Dictyostelium CybA. Alignment was done using ClustalW program. Residues identical or with a conserved substitution were coloured in black or grey respectively. Hydrophobic helices (H) corresponding to

405

absence of the C-terminal tail. Using truncated forms of human p22phox, Dinauer and colleagues established that the C-terminal tail of p22phox (aa 142–195) is dispensable for NOX2-p22phox heterodimer formation [31]. However, a proline-rich region (PRR) sequence, present in the p22phox C-terminal tail and involved in binding of p47phox and its homolog NOXO1, is essential for NOX1, 2 and 3 oxidase activity [25]. Despite a much shorter C-terminus in comparison to human p22phox, Dd-CybA may have retained the capacity to associate with some of the amoebal NOXs, NoxA and NoxB in particular. However, the lack of a PRR-like sequence (Fig. 23.5, blue boxes) and the apparent absence of p47phox homologs in Dictyostelium (see below), indicate a p47phox-independent mode of regulation for Dd-CybA/NOX complexes.

3.3

A Limited Repertoire of Cytosolic Factors

The molecular mechanism underlying the activation of the human NADPH oxidase NOX2 has been thoroughly studied and shown to involve several cytosolic factors, p40phox, p67phox, p47phox and the small GTPases Rac1/2. In resting cells, p47phox and p67phox are localized in the cytosol. Stimulation with chemoattractants, phorbol 11-myristate 12-acetate (PMA) or opsonized bacteria, induces a phosphorylationdependent conformational change of p47phox, and its recruitment along with the NOX2 activator p67phox to the membrane NADPH oxidase complex via binding to the p22phox PRR and to local phosphoinositides. In the meantime, cell activation leads to the release of Rac from the Rho GDP

putative transmembrane domains were coloured in yellow and the proline-rich sequence present in human p22phox C-terminal tail and known to interact with p47phox in blue

406

dissociation inhibitor RhoGDI and its anchoring to the membrane in a GTP-bound form. Rac translocation to the membrane and its interaction with p67phox and NOX2 are all required to form an active enzyme complex (reviewed in [32, 33]). The subunit p40phox rather functions as a modulator of NOX2 activity. This cytosolic subunit was identified as a p67phox partner and shown to facilitate the translocation of p47phox and p67phox in response to stimuli. NOX1 functions with a distinct set of cytosolic factors, NOXO1 and NOXA1, respective homologs of p47phox and p67phox, in addition to Rac GTPases. NOX3, like NOX1, is regulated by NOXO1 but does not require NOXA1 for activation. NOX4 does not require any of the above cytosolic factors for function [34], as NOX5 that uses calcium for activation [35]. Dictyostelium cells express a large family of Rho GTPases, among which Rac1A, -1B and -1C share highest homologies with neutrophil Rac2 [19, 36]. Search for homologs of the p40phox, p67phox, p47phox genes in Dictyostelium genome failed to identify amoebal counterparts except for the p67phox-like gene ncfA (Dictybase DDB_G0288773) (Fig. 23.3a). As human p67phox and NOXA1, NcfA contains a four tetratricopeptide repeat (TPR)-containing domain at the N-terminus of the protein. The TPR domain of p67phox provides an interaction surface for Rac [37]. In its TPR region, NcfA contains the highly conserved 20 amino acid hairpin insertion between TPR3 and TPR4, involved in Rac2-p67phox binding. Arg-102 and Asp-108 that are involved in p67phox/Rac complex formation are conserved in NcfA (Arg-101 and Asp-107) as well as Ser-37 and Asp-67 located in TPR2 that also contribute to the binding interface. On the other hand, all the residues in Rac involved in p67phox binding are present in Dictyostelium Rac1A-C proteins, suggesting that an NcfA-Rac interaction may similarly be at play in the amoeba to regulate some, if not all, NOX activities. Besides the TPRs, NcfA harbors a central PB1 (Phox and Bem1) domain as does p67phox, flanked by several prolinerich sequences and a C-terminal WW domain. WW domains are established poly-proline motif binders. This terminal domain could very well substitute for the C-terminal SH3 domain present in p67phox—but absent in NcfA—in the recruitment of poly-proline motif-carrying partners. Of note, NcfA is devoid of the p67phox activation domain (AD), found conserved in NOXA1, pointing towards a distinct mode of operation for this p67phox-like protein. The absence of a p40phox homologue raises the question of the role for NcfA PB1 domain as this domain in human p67phox mediates binding to p40phox through a PB1-PB1 interaction. A similar situation is observed in fungi like A. nidulans and M. grisea that possess a PB1 domain containing p67phox ortholog, NoxR, but no p40phox-like protein. Several proteins harboring a PB1 domain are encoded

L. Aubry and B. Lardy

by Dictyostelium genome and it is possible that one of these proteins targets NcfA and participates to NOX control.

3.4

NADPH Oxidase Components Are Differentially Expressed During Development

Several large scale transcriptomic studies have been performed in Dictyostelium to explore gene expression in various genetic backgrounds and experimental conditions. Data are available (open source) on the web-based application, DictyExpress (https://www.dictyexpress.org). Expression profiles of noxA, noxB, noxC, cybA and ncfA during Dictyostelium development have been established in the AX4 strain and are schematized in Fig. 23.3b based on DictyExpress data [38, 39]. All nox transcripts are detected in vegetative amoebae but at very different levels. NoxA transcripts being by far the most abundant, NoxA was proposed to be responsible for most—if not all—of the vegetative NADPH oxidase activity. However, studies investigating Dictyostelium transcriptional response to infection by various bacterial species evidenced that actors of the NADPH oxidase machinery could be induced depending on the bacteria and duration of infection [40, 41]. During development, noxA exhibits a bimodal expression profile with an increased expression during the first hours of starvation and a second but lower peak around 10–12 h of development (late aggregation phase) (Figs. 23.1 and 23.3b). A similar profile is observed for cybA mRNA although with weaker fold changes. The level of noxB transcripts rises later on, with a peak overlapping with the late peak of noxA. Expression of noxC is induced at even later time points with a maximum during the terminal phase of fruiting body formation (Figs. 23.1 and 23.3b). These data are in general agreement with our published results in the thymidine auxotroph JH10 strain, derived from the AX4 axenic strain [19]. Regarding ncfA, transcripts are also present in vegetative cells and during the early part of Dictyostelium development program (Fig. 23.3b). Past 6 h, the level of ncfA mRNA decreases progressively and remains low thereafter. The expression profile of the three NOX-encoding genes during development and the structural characteristics of the encoded proteins suggest the involvement of successive ROS-dependent events implicating distinct and probably differentially regulated NADPH oxidase complexes. Analysis, at the protein level, of the expression of the different players of the NOX machinery has been impossible so far because of the lack of good quality antibodies. Recombinant antibodies against Dictyostelium proteins under development at the Geneva Antibody Facility (www.unige.ch/medecine/ antibodies/) should allow to enrich the current RNA-seq

23

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum

data with direct information on the corresponding proteins [42].

3.5

ROS Production and Dictyostelium NOXs

Several probes have been used to successfully evidence and measure the production of ROS by Dictyostelium amoebae. Dihydroethidium (DHE) is a cell permeant compound that emits a blue fluorescence. Upon oxidation by superoxide, the probe intercalates into nuclear and mitochondrial DNA and emits a red fluorescence. Using a DHE-based assay with fluorescence measurement on cell population, Zhang and Soldati [43] demonstrated that vegetative Dictyostelium cells stimulated with E. coli LPS, a lipopolysaccharide shown to increase the bactericidal activity of the amoeba towards phagocytosed bacteria, produce a significantly higher amount of intracellular O2•–, compared to unstimulated cells [43, 44]. This O2•– is rapidly converted into H2O2 as shown by the membrane impermeant reagent Amplex Ultrared (AUR). DHE was also used to visualize by fluorescence microscopy imaging, the presence of ROS-producing cells at the slug stage of Dictyostelium developmental cycle and the impairment of ROS production by these so-called sentinel cells (see below) when knocked-out for all three nox genes [45]. Within cells, production of ROS was observed in Dictyostelium phagosomes by live fluorescence microscopy Fig. 23.6 Roles of NOX proteins in Dictyostelium. NADPH oxidases are involved in vegetative amoebae and during the multicellular phase where they participate to intraphagosomal bacterial killing, formation of extracellular DNA traps and fruiting body formation. Other biological processes relying on the NOX-dependent production of ROS in Dictyostelium will certainly emerge from coming studies (see text)

407

using silica beads coated with the ROS-sensitive dye OxyBurst Green (OBG) fluorescein. This method evidenced the accumulation of ROS inside phagosomes within the first minutes after beads uptake, with a plateau within the next 5–10 min [43]. This ROS reporter confirmed the production of intraphagosomal ROS in sentinel cells [45]. Quantitative data reporting the amount of O2•– produced by the amoebal NOX machinery are currently missing in the literature. Assays based on the reduction of Tetrazolium salt (XTT) evidenced a difference in basal activity between wildtype and noxA/B null cells of 4.6 nmol O2•–min/107 cells (B. Lardy and G. Klein, personal communication). NOX4 activity present in human embryonic kidney (HEK293) cells generates a basal level of H2O2 of about 0.22 nmol/min/107 cells, as measured by Amplex Red assay [46]. A basal NOX activity is thus present in Dictyostelium in vegetative conditions of a similar order as that of the constitutively active human NOX4.

4

Roles for NADPH Oxidases in Dictyostelium Organism

Several aspects of Dictyostelium biology have been shown to rely on the activity of the NADPH oxidase machinery that are illustrated in Fig. 23.6.

Intraphagosomal killing of internalized microorganisms Development and cell-type differentiation

Extracellular DNA-trap formation

NOXs

ROS

Other roles (motility, cytoskeleton dynamics…)

?

408

4.1

L. Aubry and B. Lardy

Involvement of NADPH Oxidases in Dictyostelium Development

Dictyostelium amoebae live on decaying leaves of forest soil, feeding primarily on bacteria. Depletion of food triggers the starving cells to engage in a developmental program allowing the survival of the species through the formation of spores (Fig. 23.1). This developmental program includes successive steps of morphogenesis and cell differentiation relying on a complex transcriptional remodelling and during which cyclic adenosine monophosphate (cAMP) functions as a key organiser [4, 47]. During the initial phase of the developmental cycle, pulses of cAMP in the nM range are released by starving free-living cells driving their chemotactic gathering and the formation of mounds containing several thousands of cells. Loose mounds get more compact as cell-cell contacts strengthen. Each of these tight mounds elongates, forming a migrating slug sensitive to light and temperature that will eventually culminate as a fruiting body. Cell differentiation initiates rapidly after mound formation with the emergence of stalk and spore precursor cells (prestalk and prespore cells respectively) with a prestalk/prespore ratio around 30/70. This ratio will remain roughly constant throughout development. Prestalk cells rapidly segregate at the tip of the mound and continue to release cAMP, which coordinate cell movement and shape the multicellular structure. Prestalk cells continuously secrete an extracellular matrix (slime sheath) composed of cellulose, polysaccharides, glycoconjugates and proteins [48]. This matrix contributes to the cohesion of the multicellular structure and provides a support facilitating cell and slug migration. Within the multicellular organism, prestalk and prespore cells adopt distinct behaviors and organize along an antero-posterior axis easily tractable at slug stage. Most of the prestalk cells occupy the anterior region leaving all prespore cells in the posterior part. The prespore cell-type population is homogenous in terms of transcriptomic signature and localization, in contrast to the prestalk cell-type population that comprises distinct subtypes among which prestalk A, O, AB and anterior-like cells (ALC) (Fig. 23.1). These cells localize differently in the slug and differentially contribute to the final structures that make up the mature fruiting body, forming the basal disc anchoring the fruiting body to the substratum, the stalk and the upper and lower cups surrounding the spore mass. Studies had provided evidence for a role of ROS in Dictyostelium chemotactic aggregation [49]. Pharmacological agents reducing ROS levels or overexpression of the Cu-Zn cofactored superoxide dismutase SodA were shown to impair the formation of aggregates and arrest development [49]. The stimulus responsible for ROS production had not been clearly identified but was shown to involve a heat-labile factor present in medium conditioned by starving cells [49]. We directly investigated a possible contribution of NADPH oxidases and

partners to the amoebal development by individual gene disruption [19]. We failed to produce an ncfA null mutant, either due to lethality or poor accessibility of the ncfA gene locus for homologous recombination. Dictyostelium JH10 cells lacking NoxA, NoxB or NoxC were viable. All three mutants displayed a similar developmental defect, despite the distinct expression profiles of the NOXs along the developmental time course. They were able to aggregate with kinetics similar to that of the parental strain, but failed to produce wild-type looking fruiting bodies, indicating a role for the NOX machinery in Dictyostelium development. Instead, tiny spore-less finger-like structures mostly composed of prestalk cells emerged at the top of the mounds. Even though massively retained in the endoplasmic reticulum (B. Lardy and G. Klein, unpublished observation), expression of a C-terminally myc-tagged version of NoxA in the noxA null strain partially restored the formation of spore-carrying fruiting bodies. Mixing of noxA/B null cells together with wild-type cells in a 1:3 ratio did not restore a wild-type behaviour for the null cells indicating that the developmental defect is at least in part cell-autonomous [19]. A developmental phenotype similar to that of nox-deleted strains was observed after disruption of the cybA gene in JH10 cells, supporting a conjoint role with NOX proteins. As mentioned earlier, in mammals, mutations affecting expression of p22phox have been shown to impair expression of associated NOX subunits [50]. Whether disrupting cybA affects the stability of the Dictyostelium NOX proteins will need to be tested. Unexpectedly, similar experiments conducted in the Ax2 strain, another Dictyostelium axenic strain, did not result in obvious developmental defects (our observations and [45]), suggesting a different sensitivity of strains to nox gene disruption, the JH-10 background being seemingly more favourable to the expression of the nox knock-out phenotypes. In the course of our work, we observed that deletion, in any of the noxA, noxB and cybA null cells, of the gene encoding the apoptosis-linked gene 2 protein Alg2B could restore a wild-type development with the formation of normal looking fruiting bodies, suggesting an inhibitory role for Alg2B on a downstream effector of NoxA/NoxB/p22phox signalling. Accordingly, re-expression of Alg2B under a constitutive promoter in the noxA/alg2b double null mutant reproduced the developmental defect observed in the noxA single null mutant [19]. In contrast, alg2B disruption did not correct the noxC null phenotype, further emphasizing functional differences between the EF-hand containing NoxC and the other two amoebal NOXs, NoxA and NoxB. Alg2B is a calcium-binding protein belonging to the penta EF-hand domain-containing protein family, whose physiological function in Dictyostelium is still unclear [51, 52]. Its mammalian counterpart ALG2 is involved in various calcium-dependent

23

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum

processes including vesicular trafficking, membrane remodelling and repair and cell death [53, 54]. These findings, in addition to the presence of EF-hands in NoxC, further support the existence of an interplay between calcium and ROS production by NOX proteins. The importance of calcium during Dictyostelium development and in particular for cell-type differentiation has been well documented [55– 57]. Its exact contribution along the NOX pathways and the cross-talk between calcium and ROS signalling now remain to be established. Calcium could modulate ROS production, not only by binding directly to NOXs (NoxC in particular) but also through calcium-dependent regulators as shown for NOX2 and NOX5 whose activities are regulated by S100A8/ A9, two members of another EF-hand protein family, and calmodulin respectively [58, 59]. ROS produced by amoebal NOXs could function upstream of calcium by acting, directly or not, on the activity of calcium channels, pumps or exchangers, thereby modulating calcium concentration and downstream effects, as shown for NOX1-generated ROS in vascular smooth muscle cells [60]. Along this line, we observed that noxA null cells display a lack of calcium sensitivity when subjected to hydrodynamic mild shear stress, in contrast to wild-type cells, to which addition of external calcium markedly stimulates cell movement along the flux (B Lardy and G Klein, unpublished data).

4.2

Involvement in Intraphagosomal Processing of Microbes

Vegetative Dictyostelium cells are professional phagocytes that perform phagocytosis for nutritive purposes as well as defence. Indeed, their natural habitat is a reservoir of microorganisms used as nutrient source but potentially pathogenic for the amoebae and that can compromise cell survival during growth or development. It has been shown that the maturation dynamics of the phagocytic pathway and signalling underlying the phagocytic activity are well conserved with human phagocytes, justifying the extensive use of Dictyostelium to investigate the molecular mechanisms controlling phagocytosis and predator/prey interactions [11– 13, 16]. Leucocytes and Dictyostelium rely on similar strategies to kill phagocytosed organisms. A large repertoire of hydrolases and antimicrobial molecules added to the acidic environment of the phagolysosomes are key elements. ROS are also part of the amoebal arsenal. Studies aimed at investigating the molecular mechanisms involved in bacterial killing established that NoxA is necessary for efficient killing of E. coli and to a lesser extent P. aeruginosa whereas it is dispensable for growth on K. pneumoniae, S. aureus and B. subtillis [61, 62]. In parallel, high throughput RNAseqbased approaches revealed that noxB that is expressed at very low level in vegetative cells grown in axenic medium is

409

upregulated after extended exposure of Dictyostelium cells to K. pneumoniae or within a few hours following infection by L. pneumophila together with other actors involved in ROS production and scavenging [40, 41]. These results suggest a differential involvement of NOXs in the amoebal response to encountered bacterial species. The signals, in addition to recognized highly-conserved pathogen-associated molecular patterns, and pathways responsible for NOX induction and/or activation remain to be identified as well as the set of actors of the host NOX machinery at play in the different infection contexts. In Dictyostelium cells infected with the human pathogenic fungus Aspergillus fumigatus, a production of ROS, sensitive to noxA/B/C triple mutation, is observed in conidiacontaining phagosomes [63]. Phagosomal enrichment in CybA-mCherry and ROS generation coincide with the neutralization of the A. fumigatus carrying phagosomes. ROS added to neutral pH would contribute to the intraphagosomal degradation of conidial melanin and a quenching of ROS by the pathogen pigment has been proposed to explain the slower neutralization and extended residence time of wildtype conidia in the phagocytic pathway compared to melanindeficient conidia. Dictyostelium has proven to be a pertinent host model to study interactions with a long list of pathogenic microorganisms [61, 62, 64, 65]. The systematic evaluation of the contribution of NADPH oxidase activity in Dictyostelium response to infection and prey killing will definitely further our knowledge of the biology of environmental pathogens, the arsenal used by host cells to protect themselves against infection, and possibly the strategies devised by pathogens to modulate and resist their hosts response.

4.3

A Role in S-Cell Mediated Innate Immunity

As Dictyostelium cells engage in the formation of the multicellular organism in response to starvation, they lose their phagocytic capacity with the exception of a subset of specialized cells called sentinel cells or S-cells [66– 68]. This population represents about 1% of the total cell number present in the multicellular structure. Sentinel cells exhibit a distribution pattern within the slug and a gene expression profile distinct from those of prespore and prestalk cells. In particular, these cells express high levels of the TollInterleukin Receptor (TIR)-domain containing protein TirA and the putative Leucine-Rich-Repeat (LRR)-domain receptor SlrA, two homologs of innate immunity signalling proteins present in plants and mammals. Sentinel cells are scattered throughout the slug. As they patrol the multicellular structure, they have been shown to sequester toxins and

410

pathogens inside large-sized intracellular vesicles, thereby removing harmful contaminants prior culmination in fruiting body. Sentinel cells—and internalized material—are discarded from the slug as small cell aggregates associated to the slime sheath left behind during slug migration. Their population is nonetheless maintained constant, ensuring the continuity of the immune surveillance activity [68]. Recent data from Zhang and colleagues established that these S-cells also use DNA-based extracellular traps (ETs) for bacteria clearance [45]. ETs were first described in neutrophils [69]. They consists in a network of DNA fibers carrying a variety of proteins including elastase, members of the S100 protein family, matrix metalloproteinase-9, lactotransferrin and cathepsin G [70]. These traps allow the capture and killing of pathogens and the degradation of released toxins via their antimicrobial protein arsenal, thereby limiting pathogen dissemination, host invasion and cell injury. The DNA present in the ETs can originate from the nucleus or from mitochondria. Upon release of nuclear DNA-based traps, phagocytes die through a cell death mechanism called ETosis, which is not occurring when mitochodrial DNA-based traps are formed. Various stimuli have been shown to induce ET formation by neutrophils including PMA, micro-organisms, and various cytokines. Several pieces of evidence suggest a role for NADPH oxidase and ROS at least in PMA-induced ETs: DNA traps are not produced by neutrophils treated with NADPH oxidase inhibitors or collected from CGD-suffering patients or from knock-out mice lacking a functional NADPH oxidase complex [71, 72]. In Dictyostelium, S-cell released DNA traps are mainly composed of mitochondrial DNA and accordingly, ET release does not affect cell viability. ET formation is stimulated by the presence of bacteria or LPS and requires NADPH oxidases and production of ROS, similarly to neutrophils [45]: slugs derived from Ax2 cells devoid of NoxA, B and C contain less ROS-producing foci detected by DHE and released less ETs when stimulated by K. pneumonia or LPS respectively. Along the same line, addition of the H2O2-scavenging enzyme catalase reduces ET formation by wild-type cells derived from LPS-stimulated slugs. In agreement with a reduced production of ROS and ET release, the spore mass of fruiting bodies from AX2 cells lacking all three NOXs displays an increased bacterial contamination compared to the wild-type cells. Studies performed on neutrophils have demonstrated that ROS produced by NADPH oxidases, in tandem with the thiol disulphide oxidoreductase Grx-1, regulate tubulin and actin polymerization by reversible glutathionylation in neutrophils [73, 74]. CGD mutations or pharmacological drugs affecting NADPH oxidase activity all lead to reduced actin and tubulin network formation besides defective degranulation and NET release. Addition of H2O2 restores actin and tubulin polymerization and degranulation. Treatment of neutrophils with the microtubule depolymerizing drug nocodazole or the actin

L. Aubry and B. Lardy

polymerization inhibitor cytochalasin B or disruption of the actin regulator Wiskott-Aldrich Syndrome Protein (WASP) block NET formation [73, 75]. Dictyostelium has been thoroughly used to dissect actin and tubulin network complexity and would be an ideal model to further investigate the interplay between NADPH oxidase-dependent processes, ROS production and cytoskeleton dynamics.

5

Conclusion

The cellular slime mold Dictyostelium expresses a repertoire of three NOX proteins, a p22phox-like protein and a single cytosolic factor, a p67phox-like protein, in a developmentallyregulated manner. Currently, knowledge of their function and regulation is still fragmentary. However, the presence of NADPH oxidase proteins in this amoebozoan organism at the frontier between unicellularity and multicellularity is a unique opportunity to investigate ROS and NADPHdependent signalling from an evolutionary perspective and address the question of the machineries acquired with multicellularity. With its natural advantages and available genetic approaches, it represents a promising model to complement our understanding of canonical NOXs and calciumdependent isoforms and shed light on the core mechanisms underlying their function and the diversity brought about during evolution within this family of proteins.

References 1. Eichinger L, Pachebat JA, Glöckner G et al (2005) The genome of the social amoeba Dictyostelium discoideum. Nature 435:43–57 2. Adl SM, Simpson AG, Lane CE et al (2012) The revised classification of eukaryotes. J Eukaryot Microbio 59:429–493 3. Parent CA, Devreotes PN (1996) Molecular genetics of signal transduction in Dictyostelium. Annu Rev Biochem 65:411–440 4. Aubry L, Firtel R (1999) Integration of signaling networks that regulate Dictyostelium differentiation. Annu Rev Cell Dev Biol 15: 469–517 5. Faix J, Kreppel L, Shaulsky G et al (2004) A rapid and efficient method to generate multiple gene disruptions in Dictyostelium discoideum using a single selectable marker and the Cre-loxP system. Nucleic Acids Res 32:el43 6. Linkner J, Nordholz B, Junemann A et al (2012) Highly effective removal of floxed Blasticidin S resistance cassettes from Dictyostelium discoideum mutants by extrachromosomal expression of Cre. Eur J Cell Biol 91(2):156–160. https://doi.org/10.1016/j. ejcb.2011.11.001 7. Muramoto T, Iriki H, Watanabe J et al (2019) Recent advances in CRISPR/Cas9-mediated genome editing in Dictyostelium. Cells 8(1):46. https://doi.org/10.3390/cells8010046 8. Kuspa A, Loomis WF (1992) Tagging developmental genes in Dictyostelium by restriction enzyme-mediated integration of plasmid DNA. Proc Natl Acad Sci USA 89(18):8803–8807 9. Gruenheit N, Baldwin A, Stewart B et al (2021) Mutant resources for functional genomics in Dictyostelium discoideum using REMIseq technology. BMC Biol 19(1):172. https://doi.org/10.1186/ s12915-021-01108-y

23

NADPH Oxidase-Dependent Processes in the Social Amoeba Dictyostelium discoideum

10. Annesley SJ, Fisher PR (2009) Dictyostelium discoideum-a model for many reasons. Mol Cell Biochem 329:73–91 11. Bozzaro S, Eichinger L (2011) The professional phagocyte Dictyostelium discoideum as a model host for bacterial pathogens. Curr Drug Targets 12:942–954 12. Xu X, Pan M, Jin T (2021) How phagocytes acquired the capability of hunting and removing pathogens from a human body: lessons learned from chemotaxis and phagocytosis of Dictyostelium discoideum. Front Cell Dev Biol 9:724940 13. Dunn JD, Bosmani C, Barisch C et al (2018) Eat prey, live: Dictyostelium discoideum as a model for cell-autonomous defenses. Front Immunol 8:1906. https://doi.org/10.3389/fimmu.2017.01906. eCollection 2017 14. Storey CL, Williams RSB, Fisher PR et al (2022) Dictyostelium discoideum: a model system for neurological disorders. Cells 11: 463. https://doi.org/10.3390/cells1103046 15. Williams RS, Boeckeler K, Graf R et al (2006) Towards a molecular understanding of human diseases using Dictyostelium discoideum. Trends Mol Med 12:415–424 16. Clarke M (2010) Recent insights into host-pathogen interactions from Dictyostelium. Cell Microbiol 12:283–291 17. Morel F, Doussiere J, Vignais PV (1991) The superoxide-generating oxidase of phagocytic cells. Physiological, molecular and pathological aspects. Eur J Biochem 201(3):523–546. https://doi.org/10. 1111/j.1432-1033.1991.tb16312.x 18. Berard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313 19. Lardy B, Bof M, Aubry L et al (2003) NADPH oxidase homologs are required for normal cell differentiation and morphogenesis in Dictyostelium discoideum. Biochim Biophys Acta 1744:199–212 20. Roos D, Kuhns DB, Maddalena A et al (2010) Hematologically important mutations: X-linked chronic granulomatous disease (third update). Blood Cells Mol Dis 45(3):246–265 21. Michelitsch MDM, Weissman JS (2000) A census of glutamine/ asparagine-rich regions: Implications for their conserved function and the prediction of novel prions. Proc Natl Acad Sci USA 97(22): 11910–11915 22. Sumimoto H (2008) Structure, regulation and evolution of Nox-family NADPH oxidase that produce reactive oxygen species. FEBS J 275:3249–3277 23. Fañanás EM, Todesca S, Sicorello A et al (2020) On the mechanism of calcium-dependent activation of NADPH oxidase 5 (NOX5). FEBS J 287(12):2486–2503. https://doi.org/10.1111/febs.15160 24. Tirone F, Radu L, Craescu CT et al (2010) Identification of the binding site for the regulatory calcium-binding domain in the catalytic domain of NOX5. Biochemistry 49(4):761–771. https://doi. org/10.1021/bi901846y 25. Kawahara T, Ritsick D, Cheng G et al (2005) Point mutations in the proline-rich region of p22phox are dominant inhibitors of Nox1- and Nox2-dependent reactive oxygen generation. J Biol Chem 280(36): 31859–31869. https://doi.org/10.1074/jbc.M501882200 26. Dahan I, Issaeva I, Gorzalczany Y et al (2002) Mapping of functional domains in the p22 phox subunit of flavocytochrome b559 participating in the assembly of the NADPH oxidase complex by “peptide walking”. J Biol Chem 277:8420–8432 27. Taylor RM, Burritt JB, Baniulis D et al (2004) Site-specific inhibitors of NADPH oxidase activity and structural probes of flavocytochrome b: characterization of six monoclonal antibodies to the p22phox subunit. J Immunol 173(12):7349–7357. https://doi. org/10.4049/jimmunol.173.12.7349 28. Parkos CA, Dinauer MC, Jesaitis AJ et al (1989) Absence of both the 91kD and 22kD subunits of human neutrophil cytochrome b in two genetic forms of chronic granulomatous disease. Blood 73(6): 1416–1420. https://doi.org/10.1182/blood.V73.6.1416.1416 29. Ambasta RK, Kumar P, Griendling KK et al (2004) Direct interaction of the novel nox proteins with p22phox is required for the

411

formation of a functionally active NADPH oxidase. J Biol Chem 279(44):45935–45941. https://doi.org/10.1074/jbc.M406486200 30. DeLeo FR, Yu L, Burritt JB et al (1995) Mapping sites of interaction of p47-phox and flavocytochrome b with random-sequence peptide phage display libraries. Proc Natl Acad Sci USA 92(15):7110–7114. https://doi.org/10.1073/pnas.92.15.7110 31. Zhu Y, Marchal CC, Casbon AJ et al (2006) Deletion mutagenesis of p22phox subunit of flavocytochrome b558: identification of regions critical for gp91phox maturation and NADPH oxidase activity. J Biol Chem 281(41):30336–30346 32. Vignais PV (2002) The superoxide-generating NADPH oxidase: structural aspects and activation mechanism. Cell Mol Life Sci 59: 1428–1459 33. Bokoch GM, Diebold BA (2002) Current molecular models for NADPH oxidase regulation by Rac GTPase. Blood 100(8): 2692–2696. https://doi.org/10.1182/blood-2002-04-1149 34. Martyn KD, Frederick LM, von Loehneysen K et al (2006) Functional analysis of Nox4 reveals unique characteristics compared to other NADPH oxidases. Cell Signal 18:69–82 35. Banfi B, Tirone F, Durussel I et al (2004) Mechanism of Ca(2+)activation of the NADPH oxidase 5 (NOX5). J Biol Chem 279:18583–18591 36. Vlahou G, Rivero F (2006) Rho GTPase signaling in Dictyostelium discoideum: Insights from the genome. Eur J Cell Biol 85:947–959 37. Lapouge K, Smith SJ, Walker PA et al (2000) Structure of the TPR domain of p67phox in complex with Rac GTP. Mol Cell 6(4): 899–907. https://doi.org/10.1016/s1097-2765(05)00091-2 38. Katoh-Kurasawa M, Hrovatin K, Hirose S et al (2021) Transcriptional milestones in Dictyostelium development. Genome Res. https://doi.org/10.1101/gr.275496.121 39. Stajdohar M, Rosengarten RD, Kokosar J et al (2017) dictyExpress: a web-based platform for sequence data management and analytics in Dictyostelium and beyond. BMC Bioinformatics 18(1):291. https://doi.org/10.1186/s12859-017-1706-9 40. Nasser W, Santhanam B, Miranda ER et al (2013) Bacterial discrimination by Dictyostelid amoebae reveals the complexity of ancient interspecies interactions. Curr Biol 23(10):862–872 41. Kjellin J, Pränting M, Bach F et al (2019) Investigation of the host transcriptional response to intracellular bacterial infection using Dictyostelium discoideum as a host model. BMC Genomics 20: 961. https://doi.org/10.1186/s12864-019-6269-x 42. Lima WC, Hammel P, Cosson P (2020) A recombinant antibody toolbox for Dictyostelium discoideum. BMC Res Notes 13(1):206. https://doi.org/10.1186/s13104-020-05048-8 43. Zhang X, Soldati T (2013) Detecting, visualizing and quantitating the generation of reactive oxygen species in an amoeba model system. J Vis Exp 81:e50717 44. Walk A, Callahan J, Srisawangyong P (2013) Lipopolysaccharide enhances bactericidal activity in Dictyostelium discoideum cells. Dev Comp Immunol 35:850–856 45. Zhang X, Zhuchenko O, Kuspa A et al (2016) Social amoebae trap and kill bacteria by casting DNA nets. Nat Commun 7(10):10938. https://doi.org/10.1038/ncomms10938 46. Nisimoto Y, Jackson HM, Ogawa H et al (2010) Constitutive NADPH-dependent electron transferase activity of the Nox4 dehydrogenase domain. Biochemistry 49(11):2433–2442. https://doi.org/ 10.1021/bi9022285 47. Loomis WF (2014) Cell signaling during development of Dictyostelium. Dev Biol 391:1–16. https://doi.org/10.1016/j.ydbio. 2014.04.001 48. Huber RJ, O’Day DH (2017) Extracellular matrix dynamics and functions in the social amoeba Dictyostelium: A critical review. Biochem Biophys Acta 1861(1):2971–2980. https://doi.org/10. 1016/j.bbagen.2016.09.026 49. Bloomfield G, Pears C (2003) Superoxide signalling required for multicellular development of Dictyostelium. J Cell Sci 116:3387– 3397

412 50. DeLeo FR, Burritt JB, Yu L et al (2000) Processing and maturation of flavocytochrome b558 include incorporation of heme as a prerequisite for heterodimer assembly. J Biol Chem 275(18): 13986–13993. https://doi.org/10.1074/jbc.275.18.13986 51. Aubry L, Mattei S, Blot B et al (2002) Biochemical characterization of two analogues of the apoptosis-linked gene 2 protein in Dictyostelium discoideum and interaction with a physiological partner in mammals, murine Alix. J Biol Chem 277(24):21947–21954 52. Ohkouchi S, Nishio K, Maeda M et al (2001) Identification and characterization of two penta-EF-hand Ca(2+)-binding proteins in Dictyostelium discoideum. J Biochem 130(2):207–215 53. Vito P, Lacanà E, D’Adamio L (1996) Interfering with apoptosis: Ca (2+)-binding protein ALG-2 and Alzheimer’s disease gene ALG-3. Science 271(5248):521–525. https://doi.org/10.1126/science.271. 5248.521 54. Maki M, Takahara T, Shibata H (2016) Multifaceted roles of ALG-2 in Ca2+-regulated membrane trafficking. Int J Mol Sci 17(9):1401. https://doi.org/10.3390/ijms17091401 55. Cubitt AB, Firtel RA, Fischer G et al (1995) Patterns of free calcium in intracellular stages of Dictyostelium expressing jellyfish aquaporin. Development 121:2291–2301. https://doi.org/10.1242/ dev.121.8.2291 56. Kubohara Y, Arai A, Gokan N, Hosaka K (2007) Pharmacological evidence that stalk cell differentiation involves increases in the intracellular Ca(2+) and H(+) concentrations in Dictyostelium discoideum. Dev Growth Differ 49:253–264. https://doi.org/10. 1111/j.1440-169X.2007.00920.x 57. Verkerke-van Wijk I, Brandt R, Bosman L et al (1998) Two distinct signaling pathways mediate DIF induction of prestalk gene expression in Dictyostelium. Exp Cell Res 245(1):179–185. https://doi.org/ 10.1006/excr.1998.4248 58. Berthier S, Paclet MH, Lerouge S et al (2003) Changing the conformation state of cytochrome b558 initiates NADPH oxidase activation: MRP8/MRP14 regulation. J Biol Chem 278(28):25499–25508. https://doi.org/10.1074/jbc.M209755200 59. Smith D, Lloyd L, Wei E et al (2022) Calmodulin binding to the dehydrogenase domain of NADPH oxidase 5 alters its oligomeric state. Biochem Biophys Rep 29:101198. https://doi.org/10.1016/j. bbrep.2021.101198 60. Zimmerman MC, Takapoo M, Jagadeesha DK et al (2011) Activation of NADPH oxidase 1 increases intracellular calcium and migration of smooth muscle cells. Hypertension 58:446–453 61. Benghezal M, Fauvarque MO, Tournebize R et al (2006) Specific host genes required for killing of Klebsiella bacteria by phagocytes. Cell Micobiol 8(1):139–148

L. Aubry and B. Lardy 62. Jauslin T, Lamrabet O, Crespo-Yanez X et al (2021) How phagocytic cells kill different bacteria: a quantitative analysis using Dictyostelium discoideum. MBio 12:e03169–e03120. https://doi. org/10.1128/mBio03169-20 63. Ferling I, Dunn JD, Ferling A et al (2020) Conidial melanin of the human-pathogenic fungus Aspergillus fumigatus disrupts cell autonomous defenses in amoebae. MBio 11(3):e00862–e00820. https:// doi.org/10.1128/mBio00862-20 64. Hilbi H, Weber SS, Ragaz C et al (2007) Environmental predators as models for bacterial pathogenesis. Environ Microbiol 9(3):563–575. https://doi.org/10.1111/j.1462-2920.2007.01238.x 65. Cosson P, Soldati T (2008) Eat, kill or die: when amoeba meets bacteria. Curr Opin Microbiol 11(3):271–276. https://doi.org/10. 1016/j.mib.2008.05.005 66. Katoh M, Chen G, Roberge E et al (2007) Developmental commitment in Dictyostelium discoideum. Eukaryot Cell 6:2038–2045 67. Roberge-White E, Katho-Kurasawa M (2011) Plasticity in the development and dedifferentiation of Dictyostelium discoideum. Dev Growth Differ 53:587–596 68. Chen G, Zhuchenko O, Kuspa A (2007) Immune-like phagocyte activity in the social amoeba. Science 317:678–681 69. Brinkmann V, Reichard U, Goosmann C et al (2004) Neutrophil extracellular traps kill bacteria. Science 303(5663):1532–1535. https://doi.org/10.1126/science.1092385 70. Li T, Zhang Z, Li X et al (2020) Neutrophil extracellular traps: signaling properties and disease relevance. Mediators Inflamm. https://doi.org/10.1155/2020/9254087 71. Bianchi M, Hakkim A, Brinkmann V (2009) Restoration of NET formation by gene therapy in CGD controls aspergillosis. Blood 114(13):2619–2622. https://doi.org/10.1182/blood-200905-221606 72. Fuchs TA, Abed U, Goosmann C et al (2007) Novel cell death program leads to neutrophil extracellular traps. J Cell Biol 176(2): 231–241. https://doi.org/10.1083/jcb.200606027 73. Stojkov D, Amini P, Oberson K et al (2017) ROS and glutathionylation balance cytoskeletal dynamics in neutrophil extracellular trap formation. J Cell Biol 216(12):4073–4090. https://doi.org/10. 1083/jcb.201611168 74. Sakai J, Li J, Subramanian KK et al (2012) Reactive oxygen speciesinduced actin glutathionylation controls actin dynamics in neutrophils. Immunity 37(6):1037–1049. https://doi.org/10.1016/j. immuni.2012.08.017 75. Neeli I, Dwivedi N, Khan S et al (2009) Regulation of extracellular chromatin release from neutrophils. J Innate Immun 1:194–201. https://doi.org/10.1159/000206974

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

24

Michael Breitenbach, Mark Rinnerthaler, Jiri Hasek, Paul J. Cullen, Campbell W. Gourlay, Manuela Weber, and Hannelore Breitenbach-Koller

Abstract

In this chapter, we describe the discovery of the NADPH oxidase gene and protein of the single-celled yeast Saccharomyces cerevisiae, Yno1. This enzyme was characterized with respect to mechanism of action, subcellular location, regulation of gene expression, and physiological function. Yno1 is not involved in defense and is not highly expressed in vegetatively growing cells. However, it is expressed in diverse stress situations. The signaling substance produced by Yno1 in conjunction with the superoxide dismutase Sod1, hydrogen peroxide, consequently leads through a change in the expression of target genes to the modulation of an adaptive cellular response. An example is the formation of pseudohyphae enabling invasive growth of the yeast cells, which is believed to aid in the utilization of new nutrients. The major role of Yno1 is in the switch of the mode of growth from vegetative budding to the formation of pseudohyphae, which are elongated chains of cells. Further examples that are described in this chapter are the

M. Breitenbach (✉) · M. Rinnerthaler · M. Weber · H. Breitenbach-Koller Department of Biosciences and Medical Biology, University of Salzburg, Salzburg, Austria e-mail: [email protected]; [email protected]; [email protected] J. Hasek Laboratory of Cell Reproduction, Institute of Microbiology of the Czech Academy of Sciences, Prague 4, Czech Republic e-mail: [email protected] P. J. Cullen Department of Biological Sciences, University at Buffalo, The State University of New York in Buffalo, Buffalo, NY, USA e-mail: [email protected] C. W. Gourlay School of Biosciences, University of Kent, Canterbury, Kent, UK e-mail: [email protected]

response to osmotic stress and mating. All these pathways have in common that they exit the regular cell cycle and are associated with in parts enormous changes in cell morphology. This is accomplished involving a change in the structure of the actin cytoskeleton. Yno1 was shown to directly modulate the actin cytoskeleton. Keywords

NADPH oxidase · Oxidative stress · Pseudohyphal growth · Mating · Osmotic stress · Actin cytoskeleton · Actin bodies

Abbreviations AMPK BLAST DEPMPO DHE DPI ERAD ESR GEF IMR MAPK MAPKKK NOX NPF PAK PCR PKA PKA pathway

AMP-dependent kinase Basic Local Alignment Search Tool 5-(Diethoxyphosphoryl)-5-methyl-1pyrroline-N-oxide Dihydroethidium Diphenyleneiodonium chloride Endoplasmic reticulum-associated protein degradation Electron spin resonance Guanine nucleotide exchange factor Integrative membrane reductase Mitogen activated protein kinase Mitogen activated protein kinase kinase kinase NADPH oxidase Nucleation promoting factor p21 activated kinase Polymerase chain reaction Protein kinase A RAS-cAMP-protein kinase A pathway

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_24

413

414

PPP RBD WASP

1

M. Breitenbach et al.

Pentose phosphate pathway Ras binding domain Wiskott Aldrich syndrome protein

Introduction

In this chapter we describe how the yeast Saccharomyces cerevisiae gene, YNO1, was identified as the first and up to now only gene in this single-celled yeast coding for a bona fide NADPH oxidase. We also describe the biochemical and genetic experiments providing proof for this assignment and the further experimental results showing that Yno1 is part of a signaling module involved in control of the actin cytoskeleton of the yeast cell. The expression of this gene is elevated under starvation conditions and in response to other stresses, which results in a restructuring of the actin cytoskeleton that is necessary for the formation of invasive filaments or pseudohyphae, for the formation (in part) of shmoos (a mating projection of yeast cells), and for gaining resistance to osmotic stress. Finally, we will speculate about additional functions of Yno1 in regulating yeast signaling networks during stress and removal of respiratory deficient cells from a population. We also include an update of the partial phylogenetic tree of the family of genes coding for NADPH oxidases on the one hand, and ferric reductases on the other, in the kingdom of fungi.

2

History of the Discovery of Yno1

Around the years 2000–2004 we were interested in elucidating the role of oxidative stress in yeast mother cellspecific aging. This was based on the finding that old yeast mother cells isolated by elutriation centrifugation contained a large amount of reactive oxygen species (ROS) localized to mitochondria. These cells showed all the biochemical markers of yeast apoptosis [1], which at the time recently had been discovered by Frank Madeo and colleagues [2]. So, the hypothesis was obvious that aging may be caused by ROS produced in the mitochondria. At the same time we had a strong interest in the biochemical action of the mutation Ras2gly18ala, gly19val. This dominant mutation was shown to be functionally equivalent to the human dominant mutation in Ha-rasgly12val, a mutation found in the homologous position of the human Ha-ras gene in a large percentage of human cancers. Ras2 is a monomeric GTPase that controls cell signaling and global nutrient control. We showed that this yeast mutation (Ras2gly18ala, gly19val) led to a very short mother cell-specific lifespan and to production of ROS in cells [3]. However, a key experiment showed that the ROS did not originate from the mitochondria and the electron

Fig. 24.1 This figure was taken from Heeren et al. [3] with permission. In vivo measurements of the superoxide adducts of the spin trap DEPMPO by ESR. The three strains shown are haploid and congenic: (a) wild type cells, no ESR signal is detectable; (b) RAS2 ala18, val 19 cells, the ESR spectrum clearly shows the DEPMPO superoxide adduct; (c) DEPMPO incubated with RAS2 ala18, val19 rho-zero cells. We see the same amount of DEPMPO superoxide adduct as in (b)

transport chain. Using the spin trap, DEPMPO, we measured the primary product of NADPH oxidases, superoxide radical anion, by in vivo electron spin resonance spectroscopy (ESR). We used an isogenic series of haploid strains that allowed direct comparison of normal or wild type cells to the Ras2gly18ala, gly19val strain, as well as the respiratorydeficient rho-zero strain still carrying the dominant RAS2 mutation. Rho-zero yeast strains are devoid of mitochondrial DNA, lack the mitochondrial electron transport chain, and consequently do not produce oxygen radicals in the mitochondria. The level of steady state superoxide spin adduct was low in the wild-type strain, but equally high in both of the Ras2gly18ala, gly19val strains irrespective of the activity of the mitochondrial respiratory chain (Fig. 24.1). These observations suggested that the high level of superoxide seen in the RAS2 mutant strain was not produced by the

24

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

respiratory complexes of the mitochondria, which did not exist in the rho-zero strain [4]. We therefore decided to search for another source of the superoxide radicals that was independent of mitochondrial respiration. As an obvious possibility, we next wanted to test S. cerevisiae genes that showed sequence similarity with the well-known human phagocytic NADPH oxidase, Nox2. The yeast genome contains 9 such coding regions which were readily identified by a BLAST search. Table 24.1 gives an overview of the 9 yeast genes including standard names, systematic names and the little that is known about physiological function and subcellular localization of these proteins. Seven of the nine open reading frames were named FRE1–7 in the literature and were indicative of a proven or probable function as ferric reductases. The two other open reading frames were not annotated at the time of our work. The functional and structural relationship between the FRE and NOX genes of yeast will be discussed later in this chapter. Next, we cloned all nine open reading frames by polymerase chain reaction (PCR) in a two-step cloning procedure into yeast expression vectors. In the first step, the amplified DNAs were cloned into the mammalian expression vector, pcDNA3.1/HYGRO (ThermoFisher https://www. thermofisher.com/order/catalog/product/V79520?SID=srchsrp-V79520#/V79520?SID=srch-srp-V79520). Starting with these clones, we undertook two approaches to determine the biochemical activity of the cloned and overexpressed genes in yeast. In a first attempt, the doxycycline-inducible yeast vector, pCM297 [14] was used to perform a standard assay with dihydroethidium (DHE ) [12]. Only one of the candidate yeast genes, YGL160W, later re-named YNO1 (Yeast NADPH Oxidase 1), displayed a strong (about tenfold) increase in DHE fluorescence. However, the doxycycline needed to induce expression from this vector displayed a

415

spectrally overlapping fluorescence signal that resulted in a comparatively high background in these experiments making it difficult to discern if other candidate yeast genes might produce superoxide at a lower level. We therefore repeated the measurements using a different yeast overexpression vector, pYES2 (Invitrogen), which depends on galactose for transcriptional induction. The open reading frames now tested were the following: YGL160W (later named YNO1), FRE8, FRE1, FRE3, PaNox1 (positive control with a proven NADPH oxidase activity from Podospoara anserina; [15]), and yeast containing the empty vector (negative control). These experiments were performed in a commonly used laboratory strain background (BY4741), in which was constructed an ordered complete deletion collection of all nonessential yeast genes (http://www.rz.unifrankfurt.de/FB/ fb16/mikro/euroscarf/index.html). In this experiment, we also identified a single open reading frame (YNO1) that yielded a strong signal (tenfold increase over the empty vector control). We first named this gene NOX1 and presented the results in an international meeting in Graz, Austria, in 2009. However, to avoid confusion with the established NADPH oxidase nomenclature, we later renamed the gene YNO1 [12], which is the gene designation used in the literature to the present day. The experiments described so far were performed on standard yeast media with glucose as a carbon source. However, we also wanted to test the influence of Yno1 expression on cell cycle regulation measuring the bud index in stationary phase. To that end, a deletion strain from the collection was used, which was unable to grow on galactose due to an inactivating mutation in the first enzyme of the galactose pathway (GAL1 or galactokinase). The strain was grown in the presence of 2% (wt/vol) raffinose and 3% (wt/vol) galactose. Transcriptional induction with galactose could be

Table 24.1 The table shows an overview of the standard names, systematic names, known or presumed biochemical functions, and the subcellular localizations of the nine homologues of human Nox2 which were found in the yeast genome sequence Standard name FRE1

Systematic name YLR214W

FRE2 FRE3

YKL220C YOR381W

FRE4

YNR060W

Molecular function Main ferric and cupric reductase [5] with specific activity for ferric citrate, catecholate and hydroxamate siderophores (e.g. ferrioxamine B and ferrichrome) [6] and enterobactin [7] Ferric reductase and cupric reductase [8] with identical specificity as FRE1 Ferric reductase with specific activity for ferrioxamine B, ferrichrome, triacetylfusarinine C, and rhodotorulic acid [7] Ferric reductase with low activity for rhodotorulic acid-iron [7]

FRE5 FRE6 FRE7

YOR384W YLL051C YOL152W

Putative ferric reductase Putative ferric reductase Putative ferric reductase

FRE8

YLR047C

Putative ferric reductase [11]; low NOX activity [12]

YNO1

YGL160W

NOX activity [12]

Localization Plasma membrane [5]

Plasma membrane [8] Plasma membrane [7] Most likely plasma membrane [7] Mitochondria [9] Vacuole [10] Most likely plasma membrane Vacuole [13], eventually endoplasmic reticulum Endoplasmic reticulum [12]

416

achieved without the metabolism of this carbon source. The result was that overexpression of Yno1 leads to a large increase in the bud index observed in stationary phase, meaning that the cells started new cell cycles which were not finished and showing that Yno1 overexpression had a strong influence on cell cycle regulation. At that time, a strong belief was prevalent in the community studying the yeast and fungal NADPH oxidases. It was believed that NADPH oxidases existed only in multicellular organisms [16]. Without exception, fungal NOX enzymes known at that time were involved in fruiting body formation in fungi like Aspergillus and Podospora, a function that is clearly multicellular. Because of the strong belief against a NOX enzyme in unicellular yeast, we became interested in defining the detailed biochemical analysis of the activity and subcellular location of the Yno1 protein. We showed that Yno1 was required for the production of superoxide by both the DHE fluorescence method and by in vivo electron spin resonance spectroscopy using the spin trap DEPMPO. The activity depended on the supply of NADPH by an intact pentose phosphate pathway (PPP) of the cells tested. The activity was located in microsomes originating from the endoplasmic reticulum (ER) of the yeast cells. The microsome in vitro system depended on added NADPH and was substantially less active with NADH. The standard NOX inhibitor, diphenyleniodoniumchlorid (DPI), blocked Yno1 activity. Activity was also found if the Yno1 protein was overproduced in yeast, purified, and inserted into artificial lipid vesicles. Altogether, these experiments demonstrated a biochemical function for the Yno1 protein as a functional NOX-type enzyme in a unicellular yeast. The location of the enzyme in vivo was shown by confocal fluorescence microscopy of the Yno1- green fluorescent protein (GFP) fusion protein, which was also functional and enzymatically active. The location coincided with both the peripheral and perinuclear ER in growing cells. The deletion of the YNO1 gene consistently led to a 20% reduction in the level of DHE fluorescence in late exponential yeast cultures growing on complex media containing either glucose, galactose or trehalose and on synthetic complete media with glucose as a carbon source. This indicated a corresponding loss in superoxide production. This fact became important in subsequent studies of Yno1 function, because, as we subsequently learned, the function of ROS comes both from Yno1 and functional mitochondria. Thus, phenotypes found in the yno1Δ deletion mutant were exacerbated by loss of functional mitochondria (rho-zero) at the same time. As shown later by Reddi and Culotta [17], the signal produced by Yno1 depends absolutely on the presence of Sod1, the Cu/Zn superoxide dismutase of yeast. The product hydrogen peroxide (H2O2), accumulates to a transient but high local concentration and stabilizes the plasma membrane casein kinase enzymes Yck1/2. One endpoint of

M. Breitenbach et al.

this regulatory effect is a reduction in the respiratory activity of mitochondria. Probably, the interaction between Yno1 and Sod1 and the stabilization of Yck1/2 requires spatial proximity between the plasma membrane and the peripheral ER, where Yno1 is located. Another finding presented in a paper describing the newly identified NADPH oxidase [12] was the connection between the Yno1 signaling system with the actin cytoskeleton, which led to the discovery of a role for Yno1 in the regulation of polarized growth that occurs during the dimorphic transition to filamentous growth, which we discuss later in this chapter. The deletion of the YNO1 gene led to hypersensitivity to the drug, wiskostatin (known from research on the human disease, Wiskott-Aldrich syndrome, [18]). Growth inhibition and the characteristic changes observed by fluorescence microsopy of the actin cytoskeleton could be reversed by external addition of non-toxic levels of H2O2. This finding again showed that the relevant signaling metabolite is probably H2O2 (see [17]). Among the 7 known human NADPH oxidases, Nox4 displayed the largest sequence similarity with YNO1. The localization of Nox4 can vary dependent on cell type and physiological condition [19, 20]. In human cancer cells Nox4 is located in the ER and its activity is found mainly in transducing signals by the generation of H2O2, which results in the metastasis of cancer cells [21].

3

The Interaction of Yno1 with Sod1 Produces H2O2 as a Signaling Molecule

An important protein that interacts with yeast NADPH oxidases is the anti-oxidative enzyme Sod1. Superoxide dismutases were first described by Irwin Fridovich and Joe McCord in the year 1968 [22]. This class of enzymes converts superoxide into H2O2 and molecular oxygen. In most eukaryotic cells, two SOD enzymes are present that differ in the metal cofactor that is bound: a manganesecontaining enzyme (MnSOD, Sod2) and a copper/zinc containing enzyme (CuZnSod, Sod1). Sod2 is exclusively localized to the matrix of the mitochondria, whereas Sod1 is predominantly cytosolic. Small amounts of the latter enzyme have also been found in the ER, peroxisomes, the nucleus and the mitochondrial intermembrane space [23, 24]. It was shown that in yeast cells, there are two predominant sources of superoxide that can be dismutated by Sod1. These are the mitochondrial respiratory chain and the yeast NADPH oxidase, Yno1 [25]. The main toxicity of superoxide is attributed to the fact that this particular ROS is oxidizing the [4 Fe- 4 S] clusters in the catalytic center of several proteins (e.g. aconitase). As a direct consequence, iron is released that leads to the production of even more dangerous oxygen species (e.g. the hydroxyl radical) [26]).

24

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

Surprisingly, it was found that superoxide dismutases (both Sod1 and Sod2) do not protect yeast cells from oxidative protein damage during either exponential growth or growth in early stationary phase. Such an effect was only observed during late-stationary phase [24]. Additionally, it was demonstrated that small amounts of the cellular Sod1 (content < 1%) is sufficient to protect cells from oxidative damage [27]. Deletion of the SOD1 gene in yeast cells is attributed with a variety of phenotypes, such as accumulation of free iron, inactivation of Aco1 and Leu1 (both are [4 Fe4 S] containing enzymes), increased levels of ROS as measured by DHE, fragmented vacuoles, paraquat sensitivity, lysine auxotrophy and an increased occurrence of apoptotic cells. All of these phenotypes can be quenched by a low dose expression of Sod1 [27]. Taken together, these findings indicate that Sod1 has a function that is independent of its established antioxidation function. This role has turned out to be the production of H2O2 as a chemical messenger that is produced in conjunction with Yno1. It is now well accepted that H2O2 has an important role in redox signaling and is a fast and important second messenger [28]. Although this growing scientific field has been understudied in model systems like yeast, the Sod1-Yno1 couple is now emerging as an important factor in redox signaling in this unicellular organism. Some of the findings [17] clearly show that Yno1 is involved in intracellular signaling by interaction with Sod1. A particularly good example of the role that H2O2 plays in cellular signaling comes from studies of global nutrient control in yeast. In yeast, glucose uptake and sensing is critical for cell growth, Glucose transport is mediated by a series of transmembrane proteins that are found at the plasma membrane. Among them are the receptors Rtg2 and Snf3 which are each composed of 12 membrane-spanning helices . Upon glucose binding, these receptors make contact with two yeast casein kinases, Yck1 and Yck2. Subsequently, these kinases phosphorylate the glucose repressors Mth1 and Std1, which leads to their degradation by the proteasome. Upon degradation of these repressors, the transcription factor Rgt1 is released and mediates the expression of HXT genes [29], which encode for proteins that transport glucose into the cell. Conversely, in the absence of glucose, the Yck1 protein is degraded. Interestingly, the C-terminal region of the Yck1 protein contains a Sod1 binding domain. Sod1 binding and the resulting H2O2 production by this enzyme promotes stabilization of Yck1. The exact underlying mechanism for this stabilization remains enigmatic. As one might expect, the Sod1-dependent stabilization of Yck1 is strictly dependent on the Yno1-derived superoxide production. One of the endpoints of this regulatory effect is lowering of the respiratory activity of mitochondria [17]. Presumably Sod1 controls multiple processes that impact on respiration and some are

417

independent of yno1, perhaps its intermembrane space dismutase role is dominant? Remarkably, Yno1 and Sod1 also regulate carbon metabolism by a second pathway. This example comes from studies of the regulation of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) by the Sod1/Yno1 couple. In yeast cells, three isoforms of GAPDH exist (Tdh1–3) that catalyze the conversion of glyceraldehyde-3-phosphate into 1,3 bis-phosphoglycerate during glycolysis. In yeast, the GAPDH enzyme contains two cysteines, of which cysteine 150 is oxidized by Sod1-derived H2O2. A direct consequence of this modification is a block of glycolysis and a redirection of metabolites into the pentose phosphate pathway (PPP) via the enzyme Zwf1 (glucose-6-phosphate dehydrogenase; G6PD; the gene name is derived from the original German name of the enzyme, Zwischenferment) [25]. In the oxidative part of the PPP, NADPH is produced that fuels a multitude of antioxidative systems (e.g. the glutathione-based system and the thioredoxin system). For a detailed review see [30]. Although it seems counterintuitive at first look, the ROS producing enzyme Yno1 has an anti-oxidative role in the cell. In this way it can also be speculated that Yno1 provides itself with its coenzyme by the PPP-derived NADPH production.

4

Interaction of Yno1 with the Yeast Ras2 Signaling System

Shortly after the discovery of Yno1 as a NADPH oxidase, a report demonstrated that Yno1 interacts with the Ras2 signaling system [31]. In this study, the authors monitored the GTP-bound or active conformation of the Ras2 protein by binding to a protein fusion consisting of the RAS binding domain (RBD) of human Raf oncoprotein and GFP. While the active form of Ras2 in respiratory-positive growing cells is mainly localized to the nucleus, the picture is different in cells deleted for the COX4 gene. In these non-respiring cells under conditions of diauxic shift, active Ras2 is localized mainly to the outer surface of mitochondria. This localization pattern leads to suppression of endoplasmic reticulumassociated protein degradation (ERAD) in which proteins that become mis-folded in the ER are removed and destroyed in the cytosol by the proteasome (see also [17]). This would be expected to result in high levels of Yno1 activity and a high percentage of cells filled with ROS. Interestingly, this unexpected action of Ras2 does not depend on one of its main effector kinases, the protein kinase A (PKA) activity, and is not disturbed by phosphodiesterase mutation, which means that it is independent of the generation of cAMP, which also depends on Ras function. It is therefore clear that the high steady-state levels of ROS in the cox4Δ single mutant and in

418

M. Breitenbach et al.

the cox4Δ rho-zero double mutant is not caused by residual production of ROS by the respiratory chain, but is produced by Yno1. The poor viability of respiratory-deficient cells under these conditions is suppressed by simultaneous deletion of the YNO1 gene. This series of experiments could serve as a model for aging and for a number of human diseases caused by loss of respiratory chain activity. It also fits well into the picture, which was later found for the physiological activity of Yno1, that connects nutritional stress response to filamentous growth, which is also a nutrient-regulated response (see below). The phenotype described above is due to inappropriate activation of a gene (YNO1) whose main function is not in vegetative growth, but functions during nutrient stress under specialized growth conditions.

5

Definition of the YNO1 Subfamily of Yeast NADPH Oxidases in Relation to the Structure and Mechanism of Other NADPH Oxidases and Their Discrimination from Ferric Reductases

Studying features of the Yno1 protein, including by evaluating the three-dimensional structure of the protein, has been instructive to our understanding of the function of this type of enzyme. The Yno1 protein is a characteristic member of the NADPH oxidase family. The FRE (ferric reductase) gene sequences are closely related to the YNO1 sequence in yeast and in the ascomycetes family in general, however the biochemical activity is different. A clearer picture is beginning to emerge by constructing a sequence-based phylogeny among all fungal NOX sequences. The different branches of this phylogeny include not only sequences that are highly similar to YNO1, but also sequences coding for the classical fungal NOX enzymes NoxA, and NoxB, which, based on the phenotypes of deletion mutants, are needed for the formation of fruiting bodies (Fig. 24.2). In the years after the initial discovery of Yno1 as a yeast NADPH oxidase, a number of important papers were published which helped to understand this complex enzyme family. The first experimentally determined threedimensional structure of a NOX enzyme, NOX5 of the bacterium Cylindrospermum stagnale, was published in 2017 [33] and was later discussed with respect to the enzyme mechanism [34]. We discuss here this structure and the suggested mechanism in an attempt to understand the difference between Nox and Fre enzymes. Although the structure contains an EF hand calcium binding domain and Yno1 does not, we will use this structure for our argument about the difference between Fre and Nox. The structure generally confirmed the previously published model structures of Nox enzymes [35]. The structure consists of six transmembrane helices spanning the plasma membrane of the bacterium. On

the cytoplasmic side, the structure of the so-called cytoplasmic domain contains the binding consensus sequences for NADPH (the source of electrons for all known Nox enzymes) and for FAD which is the first acceptor of reduction equivalents from NADPH. The further path of reduction equivalents follows two heme b moieties, both of which are bound by two histidine residues each and which stand edgeon perpendicularly in the cytoplasmic membrane. It is important to note that the heme iron centers do not form part of the substrate (dioxygen) binding site. Rather, they transmit single electrons for the reduction reaction, which takes place on the outer side of the membrane. Based on the bacterial Nox5 structure, the oxygen binding site is located in a loop on the outer aspect of the plasma membrane and consists of propionate 7 of the heme, Arg256, His317 and the iron coordinating His313 [34]. Of these three amino acid side chains, only two are conserved in Yno1, and His 317 is replaced by Phe. The primary product, superoxide, is formed in the periplasmic space of the bacterium, but the subsequent physiological metabolism (presumably in conjunction with a bacterial SOD enzyme) is not known. The single electrons for the reduction reactions are transmitted between the two cytochrome b moieties by a tryptophan side chain (W378). See Chap. 31 by W. Oosterheert, S. Marchese and A. Mattevi. Another relatively recent discovery concerns the Yno1like NADPH oxidase of Candida albicans, which is a commensal fungus, which in immunocompromised patients represents one of the major human fungal pathogens. Among the close homologs of Yno1 which are known to occur in the genome of C. albicans, CFL11 (CR_06670W_A, also known as FRE8) has been shown to encode for a NADPH oxidase [32]. The physiological activity of this gene depends on a special co-localizing SOD protein that localizes near the tip of the cell, which in this species can grow as filaments. The activity of this enzyme is needed for apical growth, as will be discussed in a later section, may connect these enzymes to the regulation of the actin cytoskeleton and filamentous growth. C. albicans is a diploid organism, and the homozygous deletion of the FRE8 gene shows a clear (but partial) defect in the formation and maintenance of hyphae and in virulence in a mouse model. Mice infected with the homozygous fre8Δ/fre8Δ deletion mutant are still killed by infection with the mutant C. albicans strain but the length of the hyphae investigated post mortem was smaller than seen in normal cells. The explanation probably is that the reduced levels of H2O2 that occur in the deletion mutant can be partially supplied by other cellular sources (for instance, the mitochondria). This is reminiscent of the partial defect in invasive growth in yeast mutants deleted for YNO1 that we discuss below (Fig. 24.3). The C. albicans gene CFL11 (CR_06670W_A, also known as FRE8), if overexpressed in P. pastoris, produces ROS and is absolutely inactive as an iron reductase, which is analogous

24

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

Fig. 24.2 Phylogenetic relationships among the fungal members of the integral membrane reductase (IMR) protein superfamily. Only the gene names used in the sequence databases are used in the figure.

419

Explanations of the abbreviations of the genus and species names are shown. The key result of this bioinformatic investigation is that the orthologs of Yno1 form a well-defined protein subfamily in the fungal

420

to the situation seen for Yno1 in S. cerevisiae [12]. It is interesting to consider here that CFL11 of C. albicans is not the closest homolog of YNO1 based on sequence alone. Instead, the closest C. albicans homolog of Yno1 by amino acid sequence is C2_03530W_A. Homozygous deletion of this gene does not show any defect in hyphal growth induced by serum (unpublished work of Arnold Bito). What does this mean for our question of the relation of Fre (ferric reductase) and Nox (NADPH oxidase) activity in the homologs of Yno1 of fungi? We think that the mechanism for the two reactions involved (Fre activity and Nox activity) is extremely similar, depending on the co-factors NADPH, FAD, and the two b-type cytochromes. Even the details of single electron transport within the cytoplasmic domain of the protein and in the membrane part of the enzyme are virtually the same. The topology of the two hemes is one on top of the other orthogonally to the lipid bilayer of the membrane. Perhaps the main difference between the Nox and Fre activities is in the substrate binding site situated on the other side of the membrane. These enzymes display the binding sites of the two different substrates (dioxygen and the iron siderophore complexes, respectively) which require a different binding pocket, which in total comprises only a short part of the protein sequence. Therefore, it is in our view impossible to predict enzyme activity (Fre vs Nox activity) based solely on amino acid sequence (over 500 amino acids) of these enzymes. Therefore, the names for Fre enzymes and Yno1-like Nox enzymes should be based on experimentally determined enzyme activities and not on sequence comparisons alone. We show in Fig. 24.2 a sequence-based phylogeny of fungal Yno1-like and NoxA-like enzymes. As there are two different enzyme activities which are encoded in Yno1-like sequences, namely Nox and Fre activities, it has been suggested to call this family of enzymes, the IMR (Integrative Membrane Reductase) family [37]. As the figure shows, the closely related subfamily (marked yellow in Fig. 24.2) of enzymes encoded by Yno1-like sequences, contains both enzymes with NADPH oxidase activity (but not ferric

Fig. 24.2 (Continued) kingdom, which is different from the previously known NoxA, B, and C subfamilies and that biochemical activities (either NADPH oxidase or ferric reductase) cannot be discriminated with confidence from the sequence based phylogenetic tree (see also text for a discussion of this point). The bold face proteins in the phylogenetic tree illustrate this point: S. c. Yno1 and C.a. C2_03530_WA are located in the same subfamily, however, the Candida gene does not code for a NADPH oxidase; C. a. CFL11 and S. c. FRE1 are sequence-wise closely related and lie in the ferric reductase subfamily, however C.a. CFLl1 is a proven NADPH oxidase [32] and S.c. FRE1 is a proven ferric reductase [5]. The abbreviated species names in the cladogram are in alphabetical order: A. g. Ashbya gossypii, A. c. Acremonium chrysogenum, A. f. Aspergillus fumigatus, C. a. Candida albicans, C. g. Candida glabrata, D. d. Dictyostelium

M. Breitenbach et al.

reductase activity) and enzymes displaying ferric reductase (but not NADPH oxidase) activity. The two sequences coding for fungal NoxA and NoxB enzymes are closely related to each other. Although they are clearly members of the IMR superfamily of proteins, they are only distantly related with the (yellow) homologs of Yno1. In the case of the NoxA and NoxB enzymes, candidate genes have been found which are believed to encode regulatory proteins related to the mammalian cytosolic components regulating the phagocyte Nox2 enzyme, like Rac and p67phox [38]. See Chap. 25 by D. Takemoto and B. Scott. However, the genome sequences of S. cerevisiae and C. albicans do not reveal open reading frames that show homologs to those regulatory proteins. Very recently, methods have become available for protein structure prediction based on the alphaFold “deep learning” algorithm [39], and a database was created comprising a few hundred thousand predicted protein structures. It is claimed by the authors that these predictions are more than 90% correct. We have compared the predicted structures of the Yno1 subfamily of proteins in S. cerevisiae and C. albicans. However, in the presumed “outside” substrate binding parts of these large proteins, no obvious structural differences were found between an established Nox protein and a proven Fre activity (data not shown in detail).

6

The Interaction of Yno1 Signaling with the Actin Cytoskeleton of the Yeast Cell

In yeast cells, the actin cytoskeleton can either be observed as patches, cables or the actomyosin ring in dividing cells. Actin patches can be considered a meshwork of branched actin filaments that are nucleated by the action of the Arp2/3 complex and a regulatory nucleation promoting factor (NPF), called Las17. The nucleation of branched actin filaments by the Arp2/3 complex and the regulation by an NPF is highly conserved, with Las17 showing homology to

discoideum, D. h. Debaryomyces hansenii, F. o. Fusarium oxysporum, F. v. Fusarium verticillioides, G. z. Gibberella zeae, K. l. Kluyveromyces lactis, L. e. Lodderomyces elongisporus, L. b. Laccaria bicolor, L. t. Lachancea thermotolerans, M. g. Magnaporthe grisea, M. o. Magnaporthe oryzae, M. t. Myceliophthora ahliaele, N. c. Neurospora crassa, N. h. Nectria haematococca, N. p. Neofusicoccum parvum, P.a. Podospora ahliae, R. o. Rhizopus oryzae, S.c. Saccharomyces cerevisiae, S. char. Stachybotrys chartarum, S. k. Saccharomyces kudriavzevii, S. m. Sordaria macrospora, S. s. Sporothrix schenckii, T.d. Torulaspora delbrueckii, T. m. Togninia minima, T. t. Thielavia terrestris, V. d. Verticillium ahlia, V.p. Vanderwaltozyma polyspora, Z. r. Zygosaccharomyces rouxii

24

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

421

Fig. 24.3 (a) Plate wash assays show that the YNO1 deletion causes a defect in invasive growth (middle panel) which can be reversed by external H2O2 (lower panel). The remaining yeast cells on the plates were photographed and quantified by optical analysis. (b) Plate wash assays of the yno1Δ mutant and suppression of the defect by Las17 overexpression The yno1-deletion in a respiring strain prevents invasive

growth (middle panel, 60% of wild type), but this effect can be compensated by overexpression of Las17 (lower panel, 100% of wild type) even in the absence of added H2O2. This is a key result prompting us to include a Yno1-independent branch in the regulatory scheme shown in Fig. 24.5. Results taken from Weber et al. [36], with permission

mammalian WASP proteins. Linear actin filaments are generated by the action of nucleating proteins called formins, which add actin monomers to a growing end. In yeast linear filaments are arranged by crosslinking proteins into cytoplasmic actin cables or the contractile cytokinetic ring. Both actin cables and patches are important for the growth of yeast cells but differ in their function. The cables primarily serve as tracks for polarized vesicular transport to the daughter cell and patches fulfil their function in close proximity to the plasma membrane. Actin patches form predominantly at sites of polarized growth and are an important component of the endocytotic machinery [40]. In exponentially growing cells both actin cables and patches form highly dynamic structures, with cables found within mother and daughter cells and endocytotic actin patches concentrated in the bud. The addition of an actin-depolymerizing agent, such as latrunculin B, prevents bud growth, demonstrating that the actin cytoskeleton is essential for membrane extension to complete construction of a new membrane protrusion. The polarization of yeast cells is dependent on the small G-protein Cdc42, a member of the RAS superfamily and Rho subfamily of GTPases. Cdc42 is highly conserved among species—human and yeast Cdc42 proteins are more than 80% identical—and the protein is widely considered to function as a master regulator of cell polarity and signal transduction. Like many monomeric G-proteins, Cdc42 is activated by a guanine nucleotide exchange factor (GEF), which in

yeast is called Cdc24. Once activated, Cdc42 binds effector proteins to induce growth at a particular site and induce polymerization of the actin-based cytoskeleton. One way that Cdc42 induces actin polymerization is by binding to the formin, Bni1 [41]. Formins induce actin polymerization and direct oriented growth in many organisms. Cdc42 also activates two p21 activated kinases [PAKs Ste20 and Cla4], which phosphorylate myosins that control vesicle delivery along the actin cytoskeleton. Ste20-like protein kinases are required for normal cell growth and the transition from apical to isotropic growth in budding yeast [42]. The dynamic nature of the actin cytoskeleton is sensitive to redox regulation. This can come either from the regulation of actin regulatory proteins, or via redox control of conserved methionine or cysteine side chains. The oxidation of Cys-374 is reported to lead to a decrease in the polymerization and elongation rate of F-actin filaments [43], while oxidation of the Met-44 residue of actin by the flavoprotein monooxygenase Mical causes disassembly and decreased polymerization of actin filaments [44]. NADPH oxidase enzymes have been found to localize with actin leading to the proposal that the cytoskeleton can be both regulated by ROS and play a role in its localized production [45]. The first indication that Yno1 may be involved in the regulation of the actin cytoskeleton came from a large-scale study. The yeast whole genome deletion collection was used to test the growth fitness of each deletion strain in the

422

presence of small molecules (millions of assays based on more than 1000 chemicals) using a microarray-based approach. The yno1Δ deletion mutant was found to be hyper-sensitive against two drugs that block actin polymerization: wiskostatin and latrunculin B [46]. The drug wiskostatin specifically binds WASP in human cells and thus blocks actin nucleation [47]. Cells lacking YNO1 were themselves sensitive to wiskostatin, suggesting an interaction with the yeast WASP homolog Las17. By combining the YNO1 deletion with a second gene deletion at the SOD1 locus this sensitivity was further increased. Following the idea that the Yno1/Sod1 couple forms H2O2, the addition of this second messenger to the experiment could in fact reverse the above mentioned phenotype [12, 36]. The addition of non-lethal amounts of H2O2 suppressed the mutant phenotype (Fig. 24.3). Addition of low amounts of latrunculin B, an inhibitor of actin polymerization, did not effect actin cable morphology in wild type yeast cells, but led to a complete disappearance of the long actin filaments in the YNO1 deletion mutant and appearance of actin patches exclusively in the daughter cells. Both phenotypes, the appearance of patches and disappearance of cables in YNO1 deletion cells treated with Latrunculin B, could be reverted by the addition of H2O2 [12]. These data suggested that the ability of Yno1 and Sod1 to produce H2O2 can affect the nature or stability of actin filaments in yeast cells. This was further supported by the appearance of actin morphology defects in yno1Δ cells where actin patches we replaced by cortical actin cables. In addition that the overexpression of Yno1 led to the aggregation of actin patches within large and static F-actin bodies. It can be speculated that Yno1, in concert with Sod1, promotes actin nucleation at sites of polarized growth. During the stationary phase of growth F-actin can be found within stable structures that have been termed “actin bodies” [48]. This is a deliberate and reversible structure that presumably exists to ensure that quiescent cells can re-use actin when nutrition arrives and rapidly initiate growth [49]. Interestingly the activity of Yno1 appears to be linked to the prevalence of actin bodies in quiescent yeast cells [36]. The oxidation of actin can also lead to irreversible aggregate formation and, unlike the formation of reversible actin bodies, these have been shown to lead to high levels of Yno1 derived ROS and a regulated form of cell death in yeast [31]. As Yno1 levels accumulate in cells that have lost respiratory function one can speculate that Yno1 forms part of a regulated cell death pathway in yeast. In this model Yno1 activity is required for the regulation of healthy actin body formation, presumably via the effects of discrete ROS production. However, in cells that have lost electron transport chain function and which accumulate high levels of Yno1 this leads to damaging levels of ROS, more actin aggregation and a loss of viability. Such regulated cell mechanisms have

M. Breitenbach et al.

been proposed in yeast as a method to allow for the removal of damaged yeast cells from a population.

7

Studies of Gene Expression of YNO1 in Relation to Filamentous Growth, Mating, and the Osmotic Stress Response

To begin to define how Yno1-derived H2O2 impacts cellular functions, two reporter assays were developed. In the first approach, a reporter plasmid was constructed that contained a chromosomal fragment immediately upstream of the YNO1 gene (500 bp) that was connected to a functional lacZ gene from Escherichia coli. The lacZ gene encodes for the ß-galactosidase protein, whose enzymatic activity can be used to measure changes in transcriptional activity by the 500 bp YNO1 promoter. As a control, a similar plasmid without the YNO1 promoter sequence was used. Under a number of different growth conditions tested, the 500 bp YNO1 promoter sequence significantly repressed reporter activity by about threefold [50]. This result fits with the idea that the YNO1 gene is only expressed under select conditions and may therefore have specific functions in the cell in specific environments. The second reporter assay was constructed in the following way: the endogenous YNO1 open reading frame was replaced by the gene encoding GFP. Accordingly, cells grown under environmental conditions that induce the expression of the YNO1 promoter show a bright green fluorescent signal. Environmental conditions that stimulated YNO1-GFP gene expression included the addition of high concentrations (non-physiological, e.g. 1 M) of sodium chloride, the yeast pheromone alpha-factor, fusel alcohols, and the poor or non-preferred carbon source galactose (instead of glucose as a carbon source). Because H2O2 is also produced by the mitochondria, the differential signal to noise ratio of this assay was further improved in petite or rho-zero strains that are devoid of mitochondrial DNA. When the reporter assays were performed in rho-zero strains, the output signal was substantially higher. This means that the mitochondrial respiratory deficiency in conjunction with the inducing conditions mentioned above induces expression of the YNO1 gene more than seen in respiring cells. This finding also confirms (or does not contradict) the results discussed below [36] that the phenotypes of cells lacking YNO1 are stronger if the cell is simultaneously rho-zero, which indicates that probably the superoxide and subsequent H2O2 originating from either mitochondria or Yno1 are equivalent. Pathways that are affected by the above-mentioned conditions include the osmotic stress response (induced by high concentrations of solutes like sodium chloride) [51], mating (induced by alpha-factor) and filamentous growth

24

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

(induced by growth of cells in galactose as carbon source [52]). All of these responses have a common denominator: the Rho-type GTPase Cdc42, which is a master regulator of cell polarity and signaling in yeast and other organisms [53, 54]. We therefore examined the role that Yno1 plays in each of these responses. In yeast cells, the vacuole is essential for osmoregulation. Salt stress induces a fragmentation of vacuoles [55]. After removal of salt stress, osmoregulation takes place and the vacuolar fragments start to fuse, regaining their typical morphology in the timespan of thirty minutes. Upon deletion of the YNO1 gene, the vacuolar collapse is increased. In the same mutant, the regeneration of vacuolar morphology is not completely abolished, but is significantly delayed [36]. Thus, Yno1 may play an important role in the cellular adaptation to the osmotic stress response. During the life cycle of yeasts and many other fungal species, cells exist in both haploid and diploid states [56]. The haploid forms differ in their mating type, which can be either MATa or MATα. Under optimal conditions, complementary cell types can fuse to produce a diploid zygote. In order to sense cells of the opposite mating type, the two cell types secrete and detect peptide pheromones. The detection of pheromones by cells of complementary mating types leads to the activation of a signaling pathway that induces characteristic morphological changes in the haploid cells (shmoos) to allow mating. Despite the fact that the addition of the pheromone alpha factor to MATa cells induced a large increase in expression of the YNO1 gene, deletion of the YNO1 gene did not impact the formation of mating structures and did not impact mating efficiency [36]. It is possible that the superoxide levels originating from the mitochondria are sufficient to regulate the pheromone response. It would be revealing to see if, in future experiments, shmoo formation and mating efficiency are perturbed in strains that lack both the YNO1 gene and functional mitochondria (e.g. rho-zero). The third known biological process that is regulated by Yno1 is filamentous growth. Filamentous (or invasive/ pseudohyphal) growth is a growth pattern that occurs in many fungal species. During filamentous growth, cells develop in elongated patterns that resemble tubes or filaments. In some species of fungi, filamentous growth occurs as a mycelial mat where long hyphae containing many nuclei grow outward from a parent cell in many directions. Filamentous growth also occurs in many species of fungal pathogens, such as C. albicans discussed above as well as numerous other pathogens that can cause infections in humans [57]. In many cases, filamentous growth is required for virulence of fungal pathogens [58]. In yeast, filamentous growth can be induced by nutrient limitation (carbon and nitrogen) and can also be triggered by fusel alcohols or the addition of galactose as the sole carbon

423

source [52, 59]. Galactose is less readily available as a source of energy and requires extensive metabolic adjustments before it can be used for metabolism. As a result of the execution of the filamentous growth program, yeast cells significantly change their morphology. Unlike filamentous fungi, S. cerevisiae does not form true hyphae, where cells form an interconnected syncytium, but yeast does form pseudohyphae, where cells grow away from their parents and become elongated, and the cells remain attached to each other. All of these changes allow cells to penetrate or invade into substrates, called invasive growth [60], presumably in search for pools of available nutrients. Filamentous growth has many of the hallmarks of eukaryotic cell differentiation. During filamentous growth, cells change their shape and polarity or orientation of growth. Cells also change their adherence properties. Genetic screens and high-throughput screens that capitalize on ordered collections of deletion mutants or gene overexpression constructs have identified more than 600 genes that impact filamentous growth [61–63]. Many of these genes comprise signaling pathways that are required for cell fate determination in eukaryotes, such as the Mitogen Activated Protein Kinase (MAPK) pathways, which are evolutionarily conserved signaling pathways that regulate the response to stress, cell differentiation, and other morphogenetic responses in many eukaryotes [64]. Filamentous growth also requires the RAS-cAMP-protein kinase A (PKA) pathway, the target of rapamycin (TOR) pathway, and the AMP-dependent kinase (AMPK) Snf1p, as well as other pathways not discussed here. These pathways do not operate in isolation but co-regulate each other’s target genes, and in some cases each others’ activities, to coordinately regulate filamentous growth. Given that Yno1 has homologs in many fungal species, including C. albicans, and given that YNO1 expression is induced by growth in galactose, we became interested in exploring whether Yno1 might regulate filamentous growth. We found that in cells that lack Yno1, the number of hyperpolarized cells is reduced and the capability for invasive growth is limited [36]. Conversely, overproduction of Yno1 led to enhanced polarized growth. Therefore, Yno1 is required for proper filamentous growth in yeast. It may be interesting to consider that Yno1 controls filamentous growth, which some speculate is related to multicellular behaviors seen in other fungal species and more distantly metazoans, and which connects Yno1 in this way to the regulation of biological responses in multicellular organisms. Collectively, our data indicate that Yno1 acts downstream of Cdc42. However, potential downstream targets of Yno1, as indicated in the previous parts of this chapter, include Ras2 and Yck1 [17, 31]. Previously, both Ras2 and Yck1 in its active form were shown to promote pseudohyphal growth [62, 65]. However, the regulatory mechanism involving Yno1 is not the only one acting on F-actin during start of

424

M. Breitenbach et al.

Fig. 24.4 Fluorescence micrography of F-actin in strains under conditions of pseudohyphae induction. The effect of Yno1p on F-actin distribution at the cell cortex. F-actin structures were visualized by expression from an Abp140-mRFP fusion protein that is integrated into the chromosome. Note the large actin bodies in the strain overexpressing Yno1. Pictures were taken in late exponential phase about 7 h after shifting to galactose medium. Video frames show that the actin bodies are relatively static. Results taken from Weber et al. [36], with permission

pseudohyphal growth. We found that overexpression of the actin nucleation factor Las17 restored invasive growth to cells lacking Yno1 [36]. Figure 24.3 shows that overexpression of Las17 can compensate the defect of invasive growth which is seen in the yno1Δ deletion mutant. Besides the small G-protein Cdc42, there are other elements shared between the osmotic stress response, filamentous growth, and the pheromone response pathway. These shared proteins comprise components of MAP kinase pathways. The utilization of common or shared proteins is a common feature of signaling pathways in eukaryotes. All three Cdc42p-dependent MAP kinase pathways require the same p21-activated protein kinase (PAK, Ste20) and MAPKKK (Ste11). Nevertheless, each pathway has a different job in the cell and induces a particular set of target genes and cellular response [64, 66]. Future work will be needed to understand whether Yno1 is performing the same task in all of the MAPK-dependent responses, or whether it has the effect of controlling some cellular outputs over others. Each of the yeast MAPK pathways alters the regular progression of the cell cycle. In fact, by employing a vector-based reporter system, we could show that some elements of the YNO1 promoter act in an inhibitory manner during exponential growth (see above). It might be noteworthy that all of these processes take place in the cortical regions of the cell: Yno1 in vegetatively-growing cells is

found in the peripheral ER, which is in close proximity to the plasma membrane. It is unknown if this location is maintained after induction of the filamentous growth program. Ras2, Cdc42 and Yck1 are associated with the plasma membrane by lipid modification, at sites where cell growth and actin cytoskeleton polymerization are initiated. It can be speculated that upon activation, all of these processes as well as Yno1 localization are concentrated at sites of polarized growth. We investigated the changes induced in the actin cytoskeleton during induction and in late exponential phase close to reaching the stationary phase of cells growing in the pseudohyphal mode. As shown in Fig. 24.4, overexpression, but not deletion of YNO1 leads to an aggregation of F-actin in the cells, which up to now has been noticed, but not analyzed in detail in the literature on the actin cytoskeleton. These actin bodies are larger than the actin patches seen in daughter cells in growing yeast and certainly need more intensive investigations in the future, also with respect to similar structures in higher cells. At present we can only make some educated guesses about the structure and function of actin bodies. We believe that they (1) result from oxidation of the actin molecule [67], (2) are not leading to apoptosis, (3) are reversible, and (4) perhaps represent a storage form of actin in conditions of low energy supply of the cell.

24

8

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

Conclusions: A Model for Yno1 Functions

To summarize, work by many laboratories which we have referenced above has established Yno1 as an NADPH oxidase in yeast. The universality of NADPH oxidases in practically all eukaryotic cells indicates that they might have existed early in the evolution of the eukaryotic tree of life and therefore likely perform an important general function for cellular health and signaling. With no exception, all of the presently available experimental results point to the fact that the Yno1 subfamily of fungal NADPH oxidases functions in signal transmission, and not in defense. Indeed, Yno1 is an important signaling molecule by the ultimate production of H2O2 (by Sod1) to generate changes in cellular physiology. Among many possible functions for this protein a prominent function is the regulation of the actin cytoskeleton. We have shown that by regulating the actin cytoskeleton, Yno1 plays a critical role in fungal cell differentiation to filamentous growth. Yno1 also regulates aspects of mating and changes to the vacuole that occur during an osmotic response. It will

Fig. 24.5 Model for the regulation of the actin cytoskeleton by Yno1 based on our own work and the work of many other laboratories, referenced in part below. In the model, the polymerization of filamentous actin (F-actin) is regulated by the action of evolutionarily conserved proteins. One of these proteins is the Rho GTPase Cdc42. When activated during normal cell cycle progression, such as through positional cues or bud-site-selection proteins, Cdc42 in the GTP or active conformation binds to the formin, Bni1, which functions to polymerize actin at sites of growth. Cdc42 also binds to PAK kinases, which phosphorylate myosin (Myo2) to direct vesicle delivery to growth sites. Cdc42 can also be activated in response to external signals, like fluctuating nutrient levels, which induce the activation of Ras2 and the signaling mucin-like glycoprotein Msb2. F-actin is also produced in a parallel pathway by the WASP homologue, Las17, which together with verprolin (Vrp1) directs actin assembly and branching through the Arp2/

425

be interesting to further define in more mechanistic detail how Yno1 exerts these various changes to cellular processes. Although the picture of the regulation of filamentous growth in S. cerevisiae remains incomplete, we present here a tentative model for how Yno1 and other proteins impact the actin cytoskeleton in response to extrinsic cues. One point of regulation is that the expression of the YNO1 gene is controlled to bring about changes in response to fluctuating environments. As a result, the actin cytoskeleton becomes regulatable by nutrient control during the overall cellular restructuring that occurs during filamentous growth (Fig. 24.5). In the model presented here, we review known and new mechanisms for the regulation of the actin cytoskeleton. In a simplified view, the regulatory scheme shown here represents a binary switch (a toggle switch) that can lead the cell either to a normal budding cell cycle if glucose is an abundant carbon source, or to filamentous growth if the less preferred carbon source galactose is utilized. The branches of this scheme are different in their dependence on YNO1. One part of this model is essentially independent of YNO1 and

3 complex (Number 5). Overlaid on this regulatory scheme is the Yno1 protein, which through Sod1-dependent production of H2O2, regulates the actin cytoskeleton, perhaps by direct modification by oxidation (Number 6). This function of Yno1 is regulated by changes in expression of the YNO1 gene, which itself can be controlled by Cdc42dependent MAPK pathways. We also note the potential function of Yno1 in feedback inhibition of Ras2. In this way, the regulation and reformation of F-actin is effected both by a pathway including Yno1, and by pathway(s) independent of Yno1.Many of these functional relationships are evolutionarily conserved and occur in S. cerevisiae and C. albicans. 1: [31]; 2: [36]; 3: [68]; 4: [17]; 5: necessary for actin filament branching / Pseudohyphal growth; 6: motor proteome for vesicle transport on actin filaments; 7: [54]; 8: [69]; 9: [70]; 10: [41]; 11: [71]; 12: [72]

426

H2O2, while in the other part, the regulation of the NADPH oxidase, Yno1, and the effects of H2O2 are depicted. The actin cytoskeleton is regulated by the Rho GTPase Cdc42, which controls formin-dependent actin polymerization [41, 73]. Cdc42p also regulates PAK kinases to control myosin-1 phosphorylation and activity. The WASP homolog Las17 also regulates the actin cytoskeleton. Las17 forms a tight complex with Vrp1. Based on our results, the direct target of Wiskostatin in yeast is Vrp1, but overexpression of Las17 can override a defect of Yno1. The regulatory pathway we show here begins to explain how signals starting with nutrient levels connect to RAS and Cdc42. Ras2 is a major nutrient-regulatory GTPase in the cell, and it might not be surprising that Ras2 is connected to Yno1 regulation [70]. Nutrients tie into Cdc42 activity in two ways. One is by spatial landmarks or bud-site-selection proteins, which through the Ras-type GTPase Rsr1 activate Cdc42 at different sites on the cell cortex depending on nutrient levels [54, 74]. The second way is by the mucin Msb2 [69], which is differentially glycosylated under different nutrient states and is proteolytically processed in its underglycosylated state, which stimulates its function in the MAPK pathway [75, 76]. When underglycosylated, Msb2 activates Cdc42 (by some manner) to induce the MAP kinase pathway that controls filamentous growth. Interestingly, this same MAPK pathway appears to influence YNO1 gene expression, which may represent a feedback loop. Collectively, Yno1, Cdc42, and many other proteins presumably work together in a coordinated manner to promote the filamentous growth mode of S. cerevisiae. Although future work will be needed to explore this model further, it is a significant finding that a ROS-dependent pathway is important for establishing a eukaryotic cell differentiation response, like filamentous growth. Acknowledgements The work presented here was supported by grant P26713 of the Austrian Science Fund FWF to M.B., P33511 of the FWF (to M.R.), GM098629 of the NIH, NIGMS to P.J.C. and 67985971 of the RVO and MEYS CR (8J20AT023) to J.H.

References 1. Laun P, Pichova A, Madeo F et al (2001) Aged mother cells of Saccharomyces cerevisiae show markers of oxidative stress and apoptosis. Mol Microbiol 39:1166–1173 2. Madeo F, Frohlich E, Frohlich KU (1997) A yeast mutant showing diagnostic markers of early and late apoptosis. J Cell Biol 139:729– 734 3. Heeren G, Jarolim S, Laun P et al (2004) The role of respiration, reactive oxygen species and oxidative stress in mother cell-specific ageing of yeast strains defective in the RAS signalling pathway. FEMS Yeast Res 5:157–167 4. Flury U, Mahler HR, Feldman F (1974) A novel respirationdeficient mutant of Saccharomyces cerevisiae. I. Preliminary

M. Breitenbach et al. characterization of phenotype and mitochondrial inheritance. J Biol Chem 249:6130–6137 5. Dancis A, Klausner RD, Hinnebusch AG et al (1990) Geneticevidence that ferric reductase is required for iron uptake in Saccharomyces cerevisiae. Mol Cell Biol 10:2294–2301 6. Shatwell KP, Dancis A, Cross AR et al (1996) The FRE1 ferric reductase of Saccharomyces cerevisiae is a cytochrome b similar to that of NADPH oxidase. J Biol Chem 271:14240–14244 7. Yun CW, Bauler M, Moore RE et al (2001) The role of the FRE family of plasma membrane reductases in the uptake of siderophoreiron in Saccharomyces cerevisiae. J Biol Chem 276:10218–10223 8. Georgatsou E, Alexandraki D (1994) Two distinctly regulated genes are required for ferric reduction, the first step of iron uptake in Saccharomyces cerevisiae. Mol Cell Biol 14:3065–3073 9. Sickmann A, Reinders J, Wagner Y et al (2003) The proteome of Saccharomyces cerevisiae mitochondria. Proc Natl Acad Sci USA 100:13207–13212 10. Huh WK, Falvo JV, Gerke LC et al (2003) Global analysis of protein localization in budding yeast. Nature 425:686–691 11. De Freitas JM, Kim JH, Poynton H et al (2004) Exploratory and confirmatory gene expression profiling of mac1Δ. J Biol Chem 279: 4450–4458 12. Rinnerthaler M, Buttner S, Laun P et al (2012) Yno1p/Aim14p, a NADPH-oxidase ortholog, controls extramitochondrial reactive oxygen species generation, apoptosis, and actin cable formation in yeast. Proc Natl Acad Sci USA 109:8658–8663 13. Wiederhold E, Gandhi T, Permentier HP et al (2009) The yeast vacuolar membrane proteome. Mol Cell Proteomics 8:380–392 14. Klinger H, Rinnerthaler M, Lam YT et al (2010) Quantitation of (a)symmetric inheritance of functional and of oxidatively damaged mitochondrial aconitase in the cell division of old yeast mother cells. Exp Gerontol 45:533–542 15. Brun S, Malagnac F, Bidard F et al (2009) Functions and regulation of the Nox family in the filamentous fungus Podospora anserina: a new role in cellulose degradation. Mol Microbiol 74:480–496 16. Lalucque H, Silar P (2003) NADPH oxidase: an enzyme for multicellularity? Trends Microbiol 11:9–12 17. Reddi AR, Culotta VC (2013) SOD1 integrates signals from oxygen and glucose to repress respiration. Cell 152:224–235 18. Rajmohan R, Meng L, Yu SJ et al (2006) WASP suppresses the growth defect of Saccharomyces cerevisiae las17Δ strain in the presence of WIP. Biochem Biophys Res Commun 342:529–536 19. Block K, Gorin Y, Abboud HE (2009) Subcellular localization of Nox4 and regulation in diabetes. Proc Natl Acad Sci USA 106: 14385–14390 20. Hilenski LL, Clempus RE, Quinn MT et al (2004) Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 24:677–683 21. Auer S, Rinnerthaler M, Bischof J et al (2017) The human NADPH oxidase, Nox4, regulates cytoskeletal organization in two cancer cell lines, HepG2 and SH-SY5Y. Front Oncol 7 22. McCord JM, Fridovich I (1969) Superoxide dismutase an enzymic function for erythrocuprein (hemocuprein). J Biol Chem 244:6049– 6055 23. Medinas DB, Rozas P, Traub FM et al (2018) Endoplasmic reticulum stress leads to accumulation of wild-type SOD1 aggregates associated with sporadic amyotrophic lateral sclerosis. Proc Natl Acad Sci USA 115:8209–8214 24. O’brien KM, Dirmeier R, Engle M et al (2004) Mitochondrial protein oxidation in yeast mutants lacking manganese- (MnSOD) or copper- and zinc-containing superoxide dismutase (CuZnSOD)— Evidence that MnSOD and CuZnSOD have both unique and overlapping functions in protecting mitochondrial proteins from oxidative damage. J Biol Chem 279:51817–51827 25. Montllor-Albalate C, Kim H, Jonke AP et al (2021) Sod1 integrates oxygen availability to redox regulate NADPH production and the

24

Discovery and Functional Analysis of the Single-Celled Yeast NADPH Oxidase, Yno1

thiol redoxome. Proc Natl Acad Sci U S A. https://doi.org/10.1073/ pnas.2023328119 26. Liochev SI, Fridovich I (1999) Superoxide and iron: Partners in crime. IUBMB Life 48:157–161 27. Montllor-Albalate C, Colin AE, Chandrasekharan B et al (2019) Extra-mitochondrial Cu/Zn superoxide dismutase (Sod1) is dispensable for protection against oxidative stress but mediates peroxide signaling in Saccharomyces cerevisiae. Redox Biol 21 28. Rhee SG, Kang SW, Jeong W et al (2005) Intracellular messenger function of hydrogen peroxide and its regulation by peroxiredoxins. Curr Opin Cell Biol 17:183–189 29. Moriya H, Johnston M (2004) Glucose sensing and signaling in Saccharomyces cerevisiae through the Rgt2 glucose sensor and casein kinase I. Proc Natl Acad Sci USA 101:1572–1577 30. Aung-Htut MT, Ayer A, Breitenbach M et al (2012) Oxidative stresses and ageing. Subcell Biochem 57:13–54 31. Leadsham JE, Sanders G, Giannaki S et al (2013) Loss of cytochrome c oxidase promotes RAS-dependent ROS production from the ER resident NADPH oxidase, Yno1p, in yeast. Cell Metab 18: 279–286 32. Rossi DCP, Gleason JE, Sanchez H et al (2017) Candida albicans FRE8 encodes a member of the NADPH oxidase family that produces a burst of ROS during fungal morphogenesis. PLoS Pathog 13 33. Magnani F, Nenci S, Millana Fananas E et al (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci U S A 114:6764–6769 34. Magnani F, Mattevi A (2019) Structure and mechanisms of ROS generation by NADPH oxidases. Curr Opin Struct Biol 59:91–97 35. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313 36. Weber M, Basu S, Gonzalez B et al (2021) Actin cytoskeleton regulation by the yeast NADPH oxidase Yno1p impacts processes controlled by MAPK pathways. Antioxidants 10 37. Grissa I, Bidard F, Grognet P et al (2010) The Nox/ferric reductase/ ferric reductase-like families of eumycetes. Fungal Biol 114:766– 777 38. Takemoto D, Tanaka A, Scott B (2007) NADPH oxidases in fungi: Diverse roles of reactive oxygen species in fungal cellular differentiation. Fungal Genet Biol 44:1065–1076 39. Jumper J, Evans R, Pritzel A et al (2021) Highly accurate protein structure prediction with AlphaFold. Nature 596:583+ 40. Mishra M, Huang J, Balasubramanian MK (2014) The yeast actin cytoskeleton. FEMS Microbiol Rev 38:213–227 41. Evangelista M, Blundell K, Longtine MS et al (1997) Bni1p, a yeast formin linking Cdc42p and the actin cytoskeleton during polarized morphogenesis. Science 276:118–122 42. Wu C, Lytvyn V, Thomas DY et al (1997) The phosphorylation site for Ste20p-like protein kinases is essential for the function of myosin-I in yeast. J Biol Chem 272:30623–30626 43. Dalle-Donne I, Carini M, Vistoli G et al (2007) Actin Cys374 as a nucleophilic target of alpha,beta-unsaturated aldehydes. Free Radic Biol Med 42:583–598 44. Hung RJ, Pak CW, Terman JR (2011) Direct redox regulation of F-actin assembly and disassembly by Mical. Science 334:1710– 1713 45. Valdivia A, Duran C, San Martin A (2015) The role of Nox-mediated oxidation in the regulation of cytoskeletal dynamics. Curr Pharm Des 21:6009–6022 46. Hillenmeyer ME, Fung E, Wildenhain J et al (2008) The chemical genomic portrait of yeast: Uncovering a phenotype for all genes. Science 320:362–365 47. Peterson JR, Bickford LC, Morgan D et al (2004) Chemical inhibition of N-WASP by stabilization of a native autoinhibited conformation. Nat Struct Mol Biol 11:747–755

427

48. Sagot I, Pinson B, Salin B et al (2006) Actin bodies in yeast quiescent cells: an immediately available actin reserve? Mol Biol Cell 17:4645–4655 49. Vasicova P, Lejskova R, Malcova I et al (2015) The stationary-phase cells of Saccharomyces cerevisiae display dynamic actin filaments required for processes extending chronological life span. Mol Cell Biol 35:3892–3908 50. Greslehner G (2016) Diploma thesis: Construction of a lacZ reporter system for measuring the activity of the promoter of YNO1, encoding a recently discovered yeast NADPH oxidase. University of Salzburg, Salzburg 51. Hohmann S (2002) Osmotic stress signaling and osmoadaptation in yeasts. Microbiol Mol Biol Rev 66:300–372 52. Gimeno CJ, Ljungdahl PO, Styles CA et al (1992) Unipolar cell divisions in the yeast S. cerevisiae lead to filamentous growth: regulation by starvation and RAS. Cell 68:1077–1090 53. Chen RE, Thorner J (2007) Function and regulation in MAPK signaling pathways: Lessons learned from the yeast Saccharomyces cerevisiae. Biochim Biophys Acta Mol Cell Res 1773:1311–1340 54. Park HO, Bi E (2007) Central roles of small GTPases in the development of cell polarity in yeast and beyond. Microbiol Mol Biol Rev 71:48–96 55. Lagrassa TJ, Ungermann C (2005) The vacuolar kinase Yck3 maintains organelle fragmentation by regulating the HOPS tethering complex. J Cell Biol 168:401–414 56. Sprague GF, Blair LC, Thorner J (1983) Cell-interactions and regulation of cell type in the yeast Saccharomyces cerevisiae. Annu Rev Microbiol 37:623–660 57. Mitchell AP (1998) Dimorphism and virulence in Candida albicans. Curr Opin Microbiol 1:687–692 58. Lo HJ, Kohler JR, Didomenico B et al (1997) Nonfilamentous C-albicans mutants are avirulent. Cell 90:939–949 59. Lorenz MC, Cutler NS, Heitman J (2000) Characterization of alcohol-induced filamentous growth in Saccharomyces cerevisiae. Mol Biol Cell 11:183–199 60. Roberts RL, Fink GR (1994) Elements of a single MAP kinase cascade in Saccharomyces cerevisiae mediate two developmental programs in the same cell type: mating and invasive growth. Genes Dev 8:2974–2985 61. Ryan O, Shapiro RS, Kurat CF et al (2012) Global gene deletion analysis exploring yeast filamentous growth. Science 337:1353– 1356 62. Shively CA, Eckwahl MJ, Dobry CJ et al (2013) Genetic networks inducing invasive growth in Saccharomyces cerevisiae identified through systematic genome-wide overexpression. Genetics 193: 1297+ 63. Palecek SP, Parikh AS, Kron SJ (2000) Genetic analysis reveals that FLO11 upregulation and cell polarization independently regulate invasive growth in Saccharomyces cerevisiae. Genetics 156:1005– 1023 64. Bardwell L (2006) Mechanisms of MAPK signalling specificity. Biochem Soc Trans 34:837–841 65. Gimeno CJ, Ljungdahl PO, Styles CA et al (1992) Unipolar cell divisions in the yeast Saccharomyces cerevisiae lead to filamentous growth—regulation by starvation and Ras. Cell 68:1077–1090 66. Saito H (2010) Regulation of cross-talk in yeast MAPK signaling pathways. Curr Opin Microbiol 13:677–68367 67. Farah ME, Amberg DC (2007) Conserved actin cysteine residues are oxidative stress sensors that can regulate cell death in yeast. Mol Biol Cell 18:1359–1365 68. Kowalewski GP, Wildeman AS, Bogliolo S et al (2021) Cdc42 regulates reactive oxygen species production in the pathogenic yeast Candida albicans. J Biol Chem 297 69. Cullen PJ, Sabbagh W, Graham E et al (2004) A signaling mucin at the head of the Cdc42- and MAPK-dependent filamentous growth pathway in yeast. Gene Dev 18:1695–1708

428 70. Zaman S, Lippman SI, Zhao X et al (2008) How Saccharomyces responds to nutrients. Annu Rev Genet 42:27–81 71. Lechler T, Jonsdottir GA, Klee SK et al (2001) A two-tiered mechanism by which Cdc42 controls the localization and activation of an Arp2/3-activating motor complex in yeast. J Cell Biol 155:261–270 72. Lechler T, Shevchenko A, Li R (2000) Direct involvement of yeast type I myosins in Cdc42-dependent actin polymerization. J Cell Biol 148:363–373 73. Evangelista M, Pruyne D, Amberg DC et al (2002) Formins direct Arp2/3-independent actin filament assembly to polarize cell growth in yeast. Nat Cell Biol 4:32–41

M. Breitenbach et al. 74. Chant J, Pringle JR (1995) Patterns of Bud-Site Selection in the Yeast Saccharomyces cerevisiae. J Cell Biol 129:751–765 75. Vadaie N, Dionne H, Akajagbor DS et al (2008) Cleavage of the signaling mucin Msb2 by the aspartyl protease Yps1 is required for MAPK activation in yeast. J Cell Biol 181:1073–1081 76. Adhikari H, Vadaie N, Chow J et al (2015) Role of the unfolded protein response in regulating the mucin-dependent filamentousgrowth mitogen-activated protein kinase pathway. Mol Cell Biol 35:1414–1432

NADPH Oxidases in Fungi

25

Daigo Takemoto and Barry Scott

Abstract

Synthesis of reactive oxygen species (ROS) by specific NADPH oxidases (Nox) can serve both defense and differentiation signalling roles in animals and plants. Fungi have three subfamilies of NADPH oxidase, NoxA, NoxB and NoxC. NoxA and NoxB have a structure very similar to the human gp91phox whereas NoxC has a Ca2+ binding motif similar to that found in the human Nox5 and plant Rboh families of NADPH oxidases. Specific isoforms of Nox have been shown by genetic analysis to be required for various fungal physiological processes and cellular differentiations, including development of sexual fruiting bodies, ascospore germination, hyphal defense, hyphal growth in both mutualistic and antagonistic plant-fungal interactions. A survey of 65 fungal genomes identified up to four Nox genes in some fungal species, reflecting the diverse morphologies and life cycles of fungal species. The presence of nox genes in fungi from the Chytridiomycota to Ascomycota suggests that Nox is an ancestral enzyme for fungi. This chapter provides an overview of our current knowledge of fungal NADPH oxidases, including Nox distribution in the fungal kingdom, Nox structure and regulation, and known biological functions of this important group of enzymes. Keywords

Fungi · NADPH oxidase · Cellular differentiations · Symbiosis · Pathogenesis · Genomes

D. Takemoto Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan e-mail: [email protected] B. Scott (✉) School of Natural Sciences, Massey University, Palmerston North, New Zealand e-mail: [email protected]

1

Introduction

Reactive oxygen species (ROS) generated by NADPH oxidases (Nox) play a crucial role in cell defense and signalling in multi-cellular organisms. The most well studied member of this group of enzymes is the mammalian gp91phox (also known as Nox2), which is responsible for the phagocytic ‘oxidative burst’, a hallmark of the mammalian defense response to microbial pathogens. An additional six NADPH oxidases, including Nox1, Nox3–Nox5 and Duox1–2 (dual oxidase), have been discovered in mammals. The Duoxes are similar in structure to Nox5 except they contain an N-terminal peroxidase domain. These enzymes have been proposed to control many different physiological functions including epithelial cell host defense, maintenance of vascular tone, regulation of hormone biosynthesis, and modulation of cell proliferation and differentiation [1–3]. Plant cells are also capable of an ‘oxidative burst’ in response to pathogen recognition [4–6]. This defense associated ROS is produced by a group of plant NADPH oxidases, known as respiratory burst homologs (Rboh) but specific isoforms are known to have key roles in developmental signalling ([7–9]. See Chap. 26 by G. Miller and R. Mittler). Arabidopsis thaliana possesses ten Rboh isoforms, which have been shown by genetic analysis to be involved in a range of physiological and developmental responses, including suppression of salicylic acid-induced programmed cell death, abscisic acid signalling, root hair tip expansion and pollen tube growth [10–13]. Filamentous fungi have three different subfamilies of NADPH oxidases, corresponding to NoxA, NoxB and NoxC [14]. The NoxA and NoxB isoforms have been shown by genetic analysis to be required for various cellular differentiations, including development of sexual fruiting bodies and germination of ascospores (Fig. 25.1) [14– 16]. NADPH oxidase catalysed ROS production also serves a defense role against other fungi [17, 18]. No functional role for the NoxC isoform has yet been established [19].

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_25

429

430

D. Takemoto and B. Scott

A1

A2

B

A3

C

D

WT

E

WT Fig. 25.1 Various phenomena regulated by NADPH oxidases in fungi. (a) Formation of multicellular fruiting body of Neurospora crassa [20]. (a1) Protoperithecium; (a2) Developing perithecium; (a3) Mature perithecium. Scanning electron microscopy (SEM) image. (b) Ascospore germination of Podospora anserina [21]. (c) Cripped growth (CG) of P. anserina. NG, Normal growth [22]. (d) Hyphal cell fusion (arrowheads) of grass-endophytic fungus Epichloë festucae. Cell wall of hyphal cells were stained with Calcofluor white. noxA knockout mutant showed no cell fusion and irregular hyphal growth on nutrient-poor

media. Bars = 10 μm [23]. (e) Appressorium-mediated penetration of rice blast pathogen Magnaporthe oryzae. Localization of LifeAct-RFP indicates the organization of F-actin beneath the appressorium where penetration peg is formed. nox2 (noxB) knockout mutant showed unorganized patch of actin located at the appressorium pore. (a) https:// journals.plos.org/plosone/article?id=10.1371/journal.pone.0110398. (b) https://journals.plos.org/plosone/article?id=10.1371/journal.pone. 0037488. (c) https://www.mdpi.com/2309-608X/4/3/85. (e) https:// www.pnas.org/doi/10.1073/pnas.1217470110

These mammalian, plant and fungal studies indicate that ROS production by NADPH oxidases is a universal defense and cell differentiation signalling system among multicellular organisms. The diffusible nature of superoxide and H2O2 make them ideal second messengers for signalling within the cell, and in the case of H2O2, which can traverse the cell membrane, inter-cellular signalling. The battery of

ROS scavenging systems present in cells including ascorbate peroxidases, glutathione, superoxide dismutases and catalases, ensures rapid turnover of the ROS to maintain ROS homeostasis [12, 24]. However, it is not clear what the targets of ROS are and how specificity is generated to regulate such a variety of cellular physiological and differentiation processes.

25

2

NADPH Oxidases in Fungi

Role of Nox in Fruiting Body Development and Ascospore Germination

Genetic analysis of fungal homologs of gp91phox has shown that these enzymes have a key role in defense and multicellular development (Fig. 25.1a). Deletion of the single NADPH oxidase gene, noxA, found in Aspergillus nidulans blocks differentiation of sexual fruiting bodies (cleistothecia) but has no effect on vegetative hyphal growth or asexual development [15]. Deletion mutants of the noxA homologues from Neurospora crassa (nox-1), Podospora anserina (Nox1) and Sordaria macrospora (nox1) are also defective in sexual development, demonstrating that NADPH oxidase catalysed production of ROS is critical for sexual fruiting body development in filamentous fungi [16, 25, 26]. While noxA mutants of A. nidulans can initiate sexual development to the point of formation of cleistothecia initials, further development of the fruiting body is blocked [15]. Cytochemical analysis showed that NoxA generates superoxide in young primordial and Hülle cells, and H2O2 and possibly other ROS in peridial cells. Lara-Ortiz et al. [15] propose that H2O2 acts as a second messenger to regulate differentiation of ascomata tissue. Nox1 mutants of P. anserina are blocked in the differentiation of ascogonia to perithecia [16]. Nox1 was identified from a suppressor screen for mutations that impair the development of a phenomenon called Crippled Growth (CG) (Fig. 25.1c), which is an epigenetic induced cell degeneration process of P. anserina caused by C, a hereditary unit that resembles a prion [17, 27, 28]. Crippled Growth can be observed in the wild-type strain grown under special conditions but the frequency is increased in anti-suppressor strains which have mutations that enhance translational accuracy [29]. P. anserina Nox1 mutants are also defective in aerial hyphae formation and pigmentation, other cell differentiations associated with entry into stationary phase growth [16]. Analysis of additional mutants that suppress CG has shown that signalling through the cell wall integrity mitogen activated protein kinase (MAPK) pathway is required for these stationary phase differentiations [27, 28]. Interestingly, the fertility defect of the Nox1 mutant can be partially rescued by serial transfer to a nutrient rich media, suggesting that ROS production by Nox1 is a signal to mobilize nutrients to the cells involved in building fructifications. Internal mobilization of nutrients is crucial, as the trigger for perithecia development in P. anserina is nutrient limitation in the external environment. In addition to the stationary phase differentiation defects, both Nox1 and ASK1 (encoding the MAPK kinase of the cell wall integrity MAPK pathway) deletion mutants are impaired in hyphal interference [18]. This is a defense reaction characterized by

431

accumulation of H2O2 and death of either the competitor fungus (when confronted with Penicillium chrysogenum) or of P. anserina itself (when confronted with Coprinopsis cinerea). Taken together these results indicate that the processes of sexual reproduction, host defense and cell degeneration in P. anserina are all interconnected [17]. As in P. anserina, a genetic screen led to the identification of Nox1 in Sordaria macrospora. Using UV mutagenesis a series of mutants were isolated blocked in fruiting body development at the stage where protoperithecia transition to mature perithecia; the so called pro mutants. One of these, pro32, was shown to have a mutation in nox1 [26]. Targeted deletion mutants of nox1 were also generated and these had the same phenotype as pro32 i.e. all were sterile, being able to develop protoperithecia but no mature perithecia. In addition, these Δnox1 mutants were defective in hyphal cell-cell fusion and had a significant reduction in vegetative hyphal growth. The lack of cell-cell fusion is a defining phenotype of noxA/ nox1 mutants (Fig. 25.1d), a phenotype initially identified in Neurospora crassa through the inability of conidial anastomosis tubes (CATs) to fuse [30]. Using nitro blue tetrazolium (NBT) to detect superoxide S. macrospora Δnox1 mutants were shown to have enhanced levels of ROS in vegetative hyphae and protoperithecia [26]. Deletion of nox-1 in Neurospora crassa also leads to a female sterility phenotype but in contrast to the closely related S. macrospora, Δnox-1 mutants are unable to form protoperithecia [25]. However, they do develop ascogonia, which is the coiled female reproductive structure that gives rise to protoperithecia. N. crassa Δnox-1 mutants are also impaired in aerial hyphal development and conidiation. Most filamentous fungi contain a second NADPH oxidase, NoxB (=Nox2) that is very similar in structure to NoxA, except it contains an N-terminal extension of unknown function (Fig. 25.2) [16, 31]. However, NoxB/ Nox2 has no role in sexual development or in the CG phenomenon observed for P. anserina (Fig. 25.1c) [16, 25, 26]. Instead Nox2 has been shown to be essential for ascospore germination in several fungi (Fig. 25.1b) [16, 25, 26]. NoxB/Nox2 catalysed production of ROS in the ascospore may act either directly on the cell wall or indirectly as a second messenger to weaken it and thereby to re-establish polarised growth [16]. The absence of a NoxB in Aspergillus may reflect the fact that nutritional signalling is not required to trigger ascospore germination in this genus. Interestingly, a P. anserina pls1 mutant, which is defective in the synthesis of a Pls1-type tetraspanin, is also defective in ascospore germination [32]. However, a mutation in the orthologous gene in the rice blast pathogen, Magnaporthe oryzae, had no effect on ascospore germination but disrupted appressorium penetration of its plant host, a pathogenicity defect identical to that seen for the M. oryzae ΔNox2 mutant (Fig. 25.1e) [33]. A

432

D. Takemoto and B. Scott

Animal

Homo sapiens Nox2 Transmembrane H. sapiens Nox5

Transmembrane (heme binding motif) EF hand

Epichloë festucae NoxA FAD binding NADPH binding

Fungi

E. festucae NoxB

100 aa Podospora anserina Nox3 (NoxC)

Plant

Oomycete

Magnaporthe oryzae Nox3 (NoxC)

Phytophthora sojae Nox

Arabidopsis thaliana RbohD

Fig. 25.2 Domain structure of NADPH oxidases in animals, fungi, oomycetes and plants

good explanation for these apparently conflicting reports is the fact that P. anserina ascospores and M. oryzae appressoria are both heavily melanised structures whereas the ascospores of the latter lack melanin, suggesting that NoxB/Nox2 catalysed ROS production is required to weaken or break-down the melanin to initiate these two different morphogenetic processes [32]. A shared feature of both processes is the emergence of polarised hyphae; through a specific pore in the ascospore of P. anserina and by way of a penetration peg at the base of the appressorium of M. oryzae (Fig. 25.1e). Another interesting phenotype observed for the P. anserina Nox mutants was a defect in the ability of these mutants to penetrate layers of cellophane [19]. On closer examination Brun et al. [19] discovered that P. anserina forms specialised cellular structures that facilitate penetration of this cellulose substrate. Hyphae growing on the surface first reorientate their growth perpendicularly toward the cellophane to form bulges that establish contact with the surface.

Then very thin needle-like structures emerge from these bulges to enable the fungus to penetrate the cellophane layers. Once deep within the layers the needle-like hyphae differentiate into palm-like structures from which more needle-like structures emerge, enabling the layers of cellophane to be crossed. Nox2 and Pls1 mutants are unable to reorientate to form bulges whereas Nox1 mutants are unable to form the needle-like structures, suggesting that penetration of the cellophane requires the sequential action of Nox2 followed by Nox1. This stage-wise recruitment of Nox is very similar to what is observed for appressorium differentiation (Nox2) and formation of the penetration peg (Nox1) in M. oryzae [33, 34]. A key regulatory component of both Nox complexes is NoxR, which is a homologue of the mammalian p67phox (Fig. 25.3) [35]. NoxR is required for both NoxA and NoxB activated cellular differentiations [25, 36]. A. nidulans noxR deletion mutants are defective in both sexual and asexual development [37]. The common

25

NADPH Oxidases in Fungi

433

Human Nox

NOXO1

O2-

O2

Nox1

NOXA1

O2

p22phox

p22phox

O2-

Rac

p47phox

O2-

Nox2

O2

Nox5

Rac

p67phox

p40phox

Fungi Nox

2 Cdc42

BemA

O2

O2-

Pls1

NoxD/Pro41

O2-

NoxA

NoxR

RacA

2 Cdc42

Cdc24 (GEF)

BemA

O2

NoxB

NoxR

O2-

O2

NoxC

RacA

Cdc24 (GEF)

Fig. 25.3 Similarities and differences of Nox family NADPH oxidases and (predicted) regulatory components of human (top) and fungi (bottom)

block in sexual development observed for both noxR and noxA mutants provides strong evidence that ROS production by NoxR activation of NoxA is necessary for the differentiation of cleistothecia [15, 37]. Why NoxR but not NoxA is required for conidiophore development in A. nidulans requires further investigation. Deletion of the noxR homologue in N. crassa (nor-1) resulted in female sterility, production of non-viable ascospores and defects in both asexual development and hyphal growth [25]. These multiple phenotypes provide genetic evidence that NOR-1 is required for activation of both NOX1 and NOX2 in N. crassa but for different differentiation processes.

3

Nox Controlled Differentiations Associated with Plant Pathogenesis and Symbiosis

NoxA and NoxB are also important for several differentiations associated with fungal symbiosis and pathogenesis. In a genetic screen to identify fungal symbiotic genes that

control the mutualistic symbiotic interaction between Epichloë festucae and perennial ryegrass (Lolium perenne), Tanaka et al. [31] identified a noxA insertional mutant that changed the interaction from mutualistic to antagonistic. In wild-type associations E. festucae grows systemically in the intercellular spaces as infrequently branched hyphae parallel to the axis of the leaf to form a hyphal network within the host aerial tissue [31, 38–40]. Within the stem of the grass the hyphae grow by tip growth but once they colonise the leaf primordia they become attached to the plant cell walls and switch to a unique pattern of growth known as intercalary growth [39, 41]. This enables the endophyte to synchronise its growth with the host and avoid mechanical shear as the cells of the leaf (both pseudostem and blade) rapidly expand. Inactivation of E. festucae noxA resulted in unregulated growth of the hyphae in meristematic and mature leaf tissue giving rise to a dramatic increase in fungal biomass in all tissues [31], suggesting that both tip- and intercalary growth are disrupted in this mutant. Plants infected with the noxA deletion mutant lose apical dominance, become severely stunted and undergo precocious senescence. Transmission

434

electron microscopy showed that deposits of electron-dense cerium chloride perhydroxides, products of a reaction between cerium chloride and H2O2, were significantly reduced in the endophyte extracellular matrix and associated plant cell walls of meristematic tissue infected with the noxA mutant. In contrast, deletion of E. festucae noxB had no effect on the plant symbiotic phenotype. Deletion mutations in noxR and racA also result in a plant interaction phenotype similar to that observed for noxA, suggesting both are required for NoxA activation [35, 42]. The inability of noxA, noxR and racA mutants to undergo cell-cell fusion (Fig. 25.1d) may be the underlying defect that gives rise to the dramatic plant-interaction phenotype seen with these and other symbiotic mutants [23, 43, 44]. These results demonstrate that fungal production of ROS by the NoxA complex is an important signalling mechanism to maintain the mutualistic interaction between E. festucae and perennial ryegrass. E. festucae is also able to form an appressorium-like structure, called an expressorium, within the plant host thereby facilitating exit of endophytic hyphae from the intercellular spaces to form a hyphal network on the surface of the leaf [45]. Genetic analysis showed that NoxA, NoxB and NoxR are all required for the proper development of this novel structure. Major remodelling of the hyphal cell wall occurs following exit from the leaf [45, 46]. Fungal NADPH oxidase catalysed production of ROS is also an important signalling mechanism for plant pathogenesis. Colonization of rye florets by Claviceps purpurea shows a very similar pattern of growth in the host as E. festucae in perennial ryegrass, with hyphae that seldom branch. Deletion of C. purpurea nox1 blocks the ability of this fungus to colonise the host and become pathogenic [47]. While the nox1 mutant can still penetrate the epidermis, it is impaired in colonization of the plant ovarian tissue. As described above both Nox1 and Nox2 complexes are required for M. oryzae to cause rice blast disease [33]. Δnox1 and Δnox2 mutants are unable to cause this disease because of an inability to bring about appressorium mediated cuticle penetration. Deletion of the noxR regulatory gene resulted in mutant phenotypes identical to the Δnox1Δnox2 double mutant, providing strong evidence that NoxR is required for activation of both Nox isoforms. Electron microscopy revealed that while appressorium development was normal, re-establishment of polarised growth at the base to initiate formation of a penetration peg was blocked. Subsequent work by the same group showed that formation of a penetration peg involves a two-step remodelling of the cytoskeleton involving the sequential action of first Nox2 then the Nox1 complex [34]. The Nox2 complex is required in the first step for septin-mediated F-actin reorientation at the appressorium pore. Formation of this hetero-oligomeric septin ring is crucial for establishing strong physical contact between the appressorium and the plant cuticle to support the strong

D. Takemoto and B. Scott

mechanical forces generated from the turgour pressure within the appressorium. The Nox1 complex is required for the second step that involves maintenance of the cortical F-actin network to enable polarised hyphal growth and host cell penetration. Interestingly, both Δnox1 and Δnox1Δnox2 mutants failed to grow within the host when inoculated into wounded rice seedlings suggesting that Nox1 is required for ongoing growth of M. oryzae within the host plant [33]. While P. anserina and the plant pathogen Botrytis cinerea do not develop true appressoria, they do form appressorialike structures when in contact with an appropriate surface [19, 48]. As described above P. anserina penetration of cellophane involves the sequential action of Nox2 and Nox1 complexes. Both ΔnoxA and ΔnoxB mutants of B. cinerea are defective in their ability to form primary and secondary lesions on young bean seedlings, with the latter showing a more severe phenotype [36]. ΔnoxR mutants had a similar phenotype to the ΔnoxAΔnoxB double mutant. Subsequent microscopic analysis of the ΔnoxA mutant showed that it was unable to form infection cushions on a glass surface [48]. B. cinerea ΔnoxA, ΔnoxB and ΔnoxR mutants are also defective in the development of sclerotia, structures that are crucial for initiation of the sexual cycle in this fungus [36]. Using RNAi Kim et al. [49] showed that Nox1 and Nox2 have important roles in virulence and pathogenicity (sclerotia development) in Sclerotinia sclerotiorum. Host lesion development is also impaired in ΔnoxA, ΔnoxB and ΔnoxR mutants of both Alternaria alternata [50] and Fusarium graminearum [51, 52]. ΔnoxA and ΔnoxR mutants of F. graminearum are also defective in perithecia development and therefore are female sterile. All these studies highlight the importance of NADPH oxidase generated ROS for key fungal differentiations associated with plant host infection and colonisation. In the case of the plant pathogens a lack of a functional NoxA or NoxB complex leads to a dramatic reduction in virulence on the host. However, and almost paradoxically, inactivation of the NoxA complex in E. festucae leads to an increase in virulence on the plant host and a breakdown in the symbiotic interaction. Whether this is related to the specialised form of growth (intercalary) this endophyte exhibits within the host remains to be determined.

4

Other Differentiations

One other very interesting differentiation is the injury induced formation of asexual structures in recovering cells of Trichoderma atroviride [53]. Genetic analysis confirmed that NADPH oxidase dependent ROS production is required for asexual development in response to this injury. Both NoxA and NoxR were shown to be required for this response.

25

NADPH Oxidases in Fungi

Soil-inhabiting fungal pathogens use chemical signals to locate and colonise their host plants. The vascular wilt fungus, Fusarium oxysporum exhibits hyphal chemotropism towards tomato roots, a response triggered by secreted plant peroxidases (Prx) [54]. Initiation of this chemotropic response requires regulated synthesis of ROS by NoxB [55]. Deletion of noxB or noxR specifically abolishes chemotropism of F. oxysporum toward Prx gradients. Exogenous treatment with H2O2 rescued chemotropic growth with either mutant but not a mutant lacking the G protein coupled receptor Ste2. Furthermore, phosphorylation of the cell wall integrity MAPK, Mpk1, triggered by peroxidase, was severely attenuated in the ΔnoxB and ΔnoxR mutants. Taken together these results suggest that Ste2 and the cell wall integrity MAPK pathway function downstream of the NoxB complex.

5

Structure of Fungal NADPH Oxidases

Production of ROS by human gp91phox (Nox2) proceeds in two discrete steps in a cell free system [56]. Electrons are initially transferred from NADPH to cytochrome b-associated FAD (step 1), followed by transfer to the haem groups of cytochrome b (step 2), which directly reduce molecular oxygen to superoxide. The small GTPase, Rac2, is involved in both steps but requires interaction with p67phox in step 2 [57]. Human Nox2 has six trans-membrane alpha helices, two of which, helix three and five, each contain two conserved His residues, which serve as coordination sites for two haem moieties, forming a channel for electrons to cross the plasma membrane [2, 58]. Two FAD and four NADPH binding motifs, are located in the C-terminus of the protein and form the central part of a cytosolic domain known as the dehydrogenase region. All NADPH oxidase proteins analysed to date, including those from animals, fungi and plants, have the six transmembrane-spanning domains, and putative FAD and NADPH binding cytosolic domains (Fig. 25.2). The three different subfamilies of NADPH oxidase found in the fungal kingdom have all these domains (Fig. 25.2) [14, 59]. NoxA has domains for the catalytic core but no additional motifs, and is very similar in structure to human Nox2 [15]. NoxB is very similar in structure to NoxA but in addition has an N-terminal extension of approximately 40 amino acids [16, 31]. Although this extension is conserved among fungal NoxB proteins, implying that this region is functionally important, a bioinformatics analysis (http:// smart.embl-heidelberg.de/smart/) of this sequence failed to identify any obvious structural or functional motifs. NoxC has an even longer N-terminal extension, of between 170 and 250 amino acids, which contains a putative calcium-binding EF-hand motif [60], similar to that found in human Nox5 and the plant Rboh enzymes (Fig. 25.2) [59]. Interestingly, the fungal-like organisms that comprise the oomycete, including

435

Phytophthora spp. and Hyaloperonospora parasitica, have NADPH oxidases that are most similar in structure to the plant Rboh enzymes, and include two N-terminal EF-hand motifs (Fig. 25.2) [59]. NADPH oxidases, which were originally considered to be ferric reductases, have been identified in the hemiascomycetous yeast fungi, Saccharomyces cerevisiae [61] and Candida albicans [62]. Because ferric reductases share all the structural features of NADPH oxidases, including six/seven transmembrane domains and sequences for binding haem, FAD and NADPH, it is difficult by sequence comparisons alone to confirm which group within the larger Nox/integral membrane reductase (IMR) superfamily they belong, without phylogenetic and biochemical analyses. Biochemical analysis of S. cerevisiae Yno1 (yeast NADPH oxidase 1) showed that this protein exhibited all the biochemical properties, including ROS production in vitro, of a fungal NADPH oxidase rather than a ferric reductase ([61]; see Chap. 24 by M. Breitenbach et al.). The presence of a distinct motif in the predicted NADPH binding site that is characteristic of the DUOX enzymes, suggests that Yno1 is structurally more similar to the mammalian Nox5 and fungal NoxC, than the fungal NoxA and NoxB. More recently, Fre8 in C. albicans has been shown to be a NADPH oxidase required for hyphal morphogenesis [62]. A Fre8-dependent ROS burst is associated with the yeast to hyphal morphological transition. Mutants of fre8Δ exhibit a morphogenesis defect in vitro, and are specifically impaired in the development and maintenance of elongated hyphae. Like Yno1, Fre8 contains all the features predicted for a member of the superfamily of the NADPH family of oxidoreductases.

6

Activation Mechanism of Fungal NADPH Oxidases

Generation of ROS by the phagocytic NADPH oxidase requires formation of a multi-enzyme complex composed of the catalytic subunit gp91phox, an adaptor protein p22phox, regulatory subunits p40phox, p47phox, p67phox, and the small GTPase Rac (Fig. 25.3) [2, 57]. gp91phox is associated in the membrane with p22phox, which stabilizes the complex and provides a docking site for the regulatory subunit p47phox. Together, gp91phox and p22phox constitute flavocytochrome b558 [63], which functions catalytically in neutrophils and monocytes in conjunction with the regulatory subunits and Rac In the resting state the regulatory subunits are located in the cytosol. Rac in resting cells is complexed to GDP and is associated in the cytosol with the inhibitor protein, RhoGDP Dissociation Inhibitor (RhoGDI) [64]. When the cells become exposed to bacteria or mediators of inflammation, p47phox becomes phosphorylated in the autoinhibitory region [65, 66], allowing the tandem Src homology 3 (SH3) domain

436

to bind to a proline rich region (PRR) in the C-terminus of p22phox, resulting in translocation to the membrane and assembly with the flavocytochrome b558 [67, 68]. Similarly, cell activation results in activation of one or more guanine nucleotide exchange factors (GEF), causing exchange of GDP for GTP on Rac dissociation from RhoGDI, and translocation to the membrane where Rac binds with the flavocytochrome [69]. Both p47phox and Rac have specific domains for interaction with p67phox, the regulatory subunit that is essential for activating electron flow within the flavocytochrome, thereby activating the NADPH oxidase. While p40phox is not essential for activation, it facilitates translocation of p47phox and p67phox to the membrane, resulting in increased Nox2 activity. Both p40phox and p47phox have a phox homology (PX) domain that interacts with specific phosphoinositides in the plasma membrane [70, 71]. Although we now have a reasonable understanding of the composition of the fungal NoxA and NoxB complexes little is still known about the biochemistry of activation of these complexes in response to growth and differentiation signals. As discussed above NoxA and NoxB have a very similar domain structure to the mammalian gp91phox, but a key outstanding question was whether they had a homologue of p22phox to stabilise these oxidoreductases in the cell membrane. While there was no obvious fungal homologue identified by BLAST type searches, a key step forward was the identification of a scaffold-like protein with some similarity to p22phox, encoded by a gene that showed up in a P. anserina CG genetic screen [72]. This Impaired in the Development of Crippled Growth (IDC) mutant (IDC509) has an identical phenotype to the Nox1 mutant (IDC343) i.e. it was defective in development of CG, aerial hyphae formation, development of colony pigmentation, formation of fruiting bodies and the ability to undergo cell-cell fusion. This gene was named PaNoxD. Remarkably, PaNoxD was found to be an orthologue of the previously characterised SmPro41, which encodes an integral membrane protein required for fruiting body development in S. macrospora [73]. As for P. anserina, the phenotype of the SmPro41 mutant was identical to that of PaNox1. Both P. anserina NoxD and S. macrospora Pro41 were predicted to have a similar membrane topology to p22phox [74]. This work enabled the identification of a homologue of PaNoxD in the more distant Leotiomycete fungus, B. cinerea [48]. As for P. anserina and S. macrospora, the phenotype of the B. cinerea noxD mutant was identical to noxA; both genes are required for the fusion of CATs, formation of infection cushions on glass slides, host pathogenicity, and formation of conidia and sclerotia [36, 75]. In yeast two-hybrid split ubiquitin and co-immunoprecipitation assays, NoxA and NoxD physically interact [48], suggesting fungi as well as mammals form a membrane flavocytochrome b558 complex (Fig. 25.3)

D. Takemoto and B. Scott

[76]. The presence of homologues of noxD in the genomes of some basal fungi and in unicellular organisms related to both fungi and animals, such as Rosella allomycis from the Cryptomycota, suggest NoxD is an ancestral trait to all fungi [48]. Given the tetraspanin Pls1, a highly conserved integral membrane protein, is required for the same differentiation steps as NoxB [32], this may be the corresponding membrane scaffold protein for this NADPH oxidase isoform (Fig. 25.3) [72]. However, evidence for a direct physical interaction between these proteins is still lacking. As discussed above fungi also contain a homologue of p67phox, which has been designated NoxR [35]. The N-terminal domain structure of NoxR is similar to p67phox, and includes four tetratricopeptide repeat (TPR) motifs for Rac binding, and a putative NADPH oxidase activation domain (Fig. 25.4) [35, 59]. In contrast, the C-terminus of NoxR lacks the SH3 protein-protein interaction domain found in p67phox that is required for interaction with p47phox, and although there is a recognisable Phox and Bem1 (PB1) domain, it is very different to the PB1 domain found in p67phox that is required for interaction with p40phox [2]. The absence of these domains in NoxR is consistent with the apparent absence of p47phox and p40phox homologues in fungal genome databases [59]. These observations suggested that fungi have distinct regulatory components for the recruitment of NoxR to the membrane and activation of NoxA or NoxB. Protein-protein interaction experiments showed that filamentous fungal homologues of the yeast polarity proteins, Bem1 and Cdc24, interact with one another and with NoxR, both in vitro and in vivo, providing strong evidence that these proteins are components of the fungal NoxA and NoxB complexes (Fig. 25.4) [77]. The other key regulatory component of the gp91phox complex is Rac. Homologues of this conserved protein are found across the fungal kingdom and have been shown in several filamentous fungi to play a key role in hyphal growth and development [42, 78, 79]. Fungal RacA is very similar in structure to mammalian Rac containing G1 to G5 domains for guanine nucleotide binding and hydrolysis, and a C-terminal CaaX (C represents cysteine, a is an aliphatic amino acid, and X is a terminal amino acid) motif for isoprenylation [80]. Although structurally very similar to Cdc42, another member of the Rho family of GTPases, RacA contains the conserved Ala32 and Gly35 (corresponding to Ala27 and Gly30 of Rac1 and Rac2), which are crucial for proteinprotein interactions between mammalian Rac and p67phox, and fungal RacA and NoxR [81, 82]. The inability of a noxR construct, encoding a NoxR with a single amino acid substitution in the predicted RacA binding site (R101E; echoing the similar R102E substitution in mammalian p67phox), to complement a noxR deletion mutant, suggests that interaction between NoxR and RacA in E. festucae hyphae in the plant is required to activate NoxA.

25

NADPH Oxidases in Fungi

Fig. 25.4 Similarities and differences of domain composition of Nox regulatory components of human (top) and fungi (bottom). TPR, tetratricopeptide repeats; AD, activation domain; SH3, Src homology 3; PB1, Phox and Bem 1; AIR, auto-inhibitory region; PX, Phox homology; CH, calponin homology; GEF, guanine nucleotide exchange factor

437

Human TPRs

AD SH3

PB1

SH3

p67phox PX

SH3 SH3

PX

SH3

AIR

p47phox PB1

p40phox TPRs

AD

PB1

SH3

NOXA1 SH3 SH3

PX NOXO1

Fungi TPRs

AD

PB1

NoxR

CH

Rho GEF

PB1

Cdc24 SH3

SH3

PX

PB1

BemA

In contrast to NoxA and NoxB, the fungal NoxC has an EF-hand motif in the N-terminal extension of the protein (Figs. 25.2 and 25.3) [14]. Human Nox5, which has EF-hand motifs, has a mechanism of activation that is distinct from the Nox1–4, as it does not require either p67phox or Rac [83]. Organisms that only have NADPH oxidases of the EF-hand type, such as plants, insects and oomycetes, have no p67phox/NoxR-like gene in their genome. These observations suggest that fungal NoxC will have a mechanism of activation that is distinct from NoxA and NoxB, which is probably mediated by calcium signalling [59]. The absence of both noxR and noxD homologues in the hemiascomycetous yeast fungi [59], suggest that activation of the Nox enzymes found in this group of fungi [61, 62], will also be different to that of NoxA and NoxB. While we now have considerable insight into the components of the fungal Nox complex and their biological role, links between these complexes and signalling pathways are still rather limited. However, a link between Nox and cAMP signalling has been shown for S. macrospora [26]. Nox links with the three MAPK pathways including stress-activation [15], cell wall integrity [84] and pheromone/ pathogenesis [84] signalling, all of which play key roles in development and fungal-host interactions, have also been established. The presence of EF hands in NoxC implies this

Nox is linked to calcium signalling. Little is also known about the signalling hubs where these complexes come together. In mammals, IQ motif containing GTPaseactivating protein (IQ GAP) scaffold proteins have been shown to have a key role in Nox complex signalling [85]. These scaffold proteins are involved in actin binding [86], cell adhesion and migration as well as cytokinesis in mammals [87, 88], and septation, cytokinesis and re-polarisation of the cytoskeleton in Ashbya gossypii and Dictyostelium [89, 90]. An analysis of the single IQGAP homologue found in B. cinerea, showed that it was important for several developmental processes and interacted physically with RacA and NoxD, suggesting it may have a key scaffold role in bringing together signalling complexes [91]. Our knowledge of the cellular localisation of the active (ROS-generating) Nox in fungal cells has also not been clearly defined even though GFP and mCherry tagging of some of the components has been examined. In P. anserina and B. cinerea, NOX1-GFP has been detected mainly in the ER and vesicles derived from the ER [72, 75], while vacuolar and plasma membrane-localization of Nox1-mCherry was observed in N. crassa [92]. Besides these post-translational activation mechanisms, at least some fungal Nox genes are also regulated at the transcriptional level. In A. nidulans noxA expression is induced

438

D. Takemoto and B. Scott

during sexual development, and peaks at the time of cleistothecia differentiation [15]. A key regulator of sexual development in S. macrospora and N. crassa is PRO1/ADV1, a Zn(II)2Cys6 zinc cluster transcription factor [93]. ChIPseq analysis in S. macrospora identified around 30 target genes of PRO1 that were previously shown to be involved in sexual development in A. nidulans, N. crassa and S. macrospora, including genes encoding components of the NOX complexes (nox1, nox2, pls1 and pro41) as well as the cell wall integrity and pheromone signalling MAPK pathways [94]. A conserved motif for PRO1 binding was found in the promoter regions of these genes. Taken together, these results suggest that NADPH oxidase activation in fungi is controlled at three different levels; (1) transcriptional activation of nox, (2) translocation of cytosolic NoxR and other Nox regulator components to the plasma membrane, and (3) activation of RacA by a specific GEF which promotes Rac binding with Nox.

7

Distribution of NoxA, NoxB, NoxC and Predicted Regulatory Components of Fungal NADPH Oxidases

A survey of 65 fungal genomes from diverse taxonomic groups revealed considerable variation in nox gene composition within the fungi, from a complete absence in some lineages (Saccharomycotina and Mucoromycotina) with up to four copies in Pucciniomycotina species, reflecting the diverse morphologies and life cycles of fungal species (Fig. 25.5). Given that nox genes can be found in a wide range of fungi from Chytridiomycota to Ascomycota, Nox is an ancestral enzyme for fungi, likely to have been lost in some lineages during their evolution. As discussed above, disruption of the noxA (nox1) gene in various fungi disrupts multicellular fruiting body formation. Consistent with this observation, the distribution of NoxA correlates well with the ability of these species to form fruiting bodies in their life cycle [95]. pro41/noxD homologues are found in all fungal species that have NoxA, except a basal Blastocladiomycota species, Allomyces macrogynus, providing strong support that fungal Pro41/ NoxD functions as the scaffold equivalent of p22phox for NoxA [48]. NoxB is found in most of the fungi that have NoxA, except Aspergillus species. As discussed above disruption of noxB (nox2) compromises ascospore (sexual spore) germination [15, 16]. Another key observation from the genome survey is the strong correlation, with the exception of the basal lineages, between the presence of plsA homologues with noxB, supporting the hypothesis that Pls1 might function as the membrane scaffold protein (like p22phox) specific for NoxB [72].

In contrast to the widespread distribution of NoxA and NoxB across the fungal kingdom, NoxC (which has EF-hand motifs like human Nox5) is infrequently found among the fungal genomes analysed (Fig. 25.5). NoxC was found in just 11 out of the 65 fungal species analysed. While attempts have been made to analyse the function of NoxC in a number of filamentous fungi by gene disruption [19], no clear phenotypes have been observed for these disrupted strains, leaving the role of this isozyme still elusive. Because NoxC was not found in primitive fungal species, it is unclear whether the fungal NoxC shares an ancestral origin with mammalian Nox5, despite the structural similarities. Although ferric reductase-like Nox in S. cerevisiae and C. albicans (described above) show structural similarity with mammalian Nox [61, 62], they were not identified as significant homologues by the search using mammalian Nox or fungal NoxA, B and C as queries, suggesting that these Noxs evolved independently. Homologues of noxR were found in all fungal genomes that had either noxA or noxB, providing further support that NoxR is an essential component for activation of the Nox complex (Fig. 25.5). The absence of a noxR homologue in two Ustilaginomycotina species but the presence of a noxC homologue, suggests that NoxR is not required for NoxC activation, which is consistent with the very different structure of this isozyme. As for the mammalian p67phox, fungal NoxR has been shown to interact directly with the small GTPase, RacA, via the N-terminal TPR motif [35, 82]. While some fungal species devoid of both noxA and noxB have a noxR-like gene, the encoded NoxR-like product lacks a Nox activation domain. However, these NoxR-like proteins appear to have retained the N-terminal TPR domain required for interaction with RacA [59]. Furthermore, this domain contains a conserved arginine residue between the 3rd and 4th TPR motif for Rac-binding, corresponding to R102 in the 3rd TPR motif in mammalian p67phox [35, 96], suggesting that they are indeed Rac-binding proteins. Homologues of racA were found across the fungal kingdom and in general correlated with the presence of homologues of noxA/noxB and noxR, with the exception of some species within the Ustilaginomycotina lineage (Fig. 25.5). Given that Plant Rboh’s have N-terminal EF-hand motifs required for the binding of Rac for their activation [97], the RacA homologues found within Ustilaginomycotina might be involved in the activation of NoxC. BemA (Bem1), Cdc24 and Cdc42 were originally identified as central components for the establishment of polarized growth in the yeast, Saccharomyces cerevisiae [98]. Interestingly, protein-protein interaction studies suggest these same proteins are key regulators of the fungal Nox complex [77]. Given BemA and Cdc24 both interact

25

NADPH Oxidases in Fungi Division

Cla ss/Subphy lum

Ascomycota

Sordariomycetes

Leotiomycetes

Lecanoromycetes Eurotiomycetes

Dothideomycetes

Pezizomycetes

Orbiliomycetes

Saccharomycotina

Taphrinomycotina

Basidiomycota

Agaricomycotina

Ustilaginomycotina

Pucciniomycotina

Mucoromycota

Glomeromycotina

Fungi

Mortierellomycotina

Mucoromycotina

Zoopagomycota Blastocladiomycota Chytridiomycota

Entomophthoromycotin Kickxellomycotina

439 Spe cie s

str a in

Type

NoxA

NoxB

NoxC

NoxR

Pro41 /NoxD

Pls1

BemA

Cdc24

RacA

Neurospora crass Podospora anserina Magnaporthe oryzae Fusarium graminearum Cryphonectria parasitica Epichloe festucae Metarhizium anisopliae Cordyceps fumosorosea Ophiocordyceps australis Botrytis cinerea Blumeria graminis f. sp. triticale Sclerotinia sclerotiorum Pseudogymnoascus verrucosus Hymenoscyphus fraxineus Physcia stellaris Heterodermia speciosa Aspergillus fumigatus Aspergillus oryzae Aspergillus terras Penicillium expansum Bipolaris maydis Alternaria alternata Neofusicoccum parvum Tuber melanosporum Tirmania nivea Terfezia claveryi Orbilia oligospora Dactylellina haptotyla Arthrobotrys flagrans Saccharomyces cerevisiae Saccharomyces eubayanus Naumovozyma castellii Yarrowia lipolytica Tortispora caseinolytica Candida orthopsilosis Schizosaccharomyces pombe Saitoella complicata Pneumocystis jirovecii Cryptococcus neoformans Cryptococcus wingfieldii Rhizoctonia solani Fomitiporia mediterranea Kwoniella dejecticola Ustilago maydis Tilletia walkeri Meira miltonrushii Sporisorium graminicola Puccinia triticina Puccinia striiformis f. sp. tritici Melampsora larici-populina Cronartium quercuum f. sp. fusiforme Glomus cerebriforme Rhizophagus irregularis Gigaspora margarita Mortierella antarctica Gamsiella multidivaricata Actinomortierella wolfii Mucor circinatus Syncephalastrum racemosum Rhizopus delemar

OR74A

S

1

1

-

1

1

1

1

1

1

1

S mat+

S

1

1

1

1

1

1

1

1

1

1

1

1

1

1

70-15

P

1

1

1

1

Cdc42

1

1

PH1

P

1

1

1

1

1

1

1

1

1

1

EP115

P

1

1

-

1

1

1

1

1

1

1

Fl1

E

1

1

E6

I

1

1

-

1

1

1

1

1

1

1

ARSEF 2679

I

1

1

-

1

1

1

1

1

1

1

1

1

-

1

1

1

1

1

Map64

I

1

1

-

1

1

1

1

1

1

1

B05.10

P

1

1

-

1

1

1

1

1

1

1

THUN-12

P

1

1

-

1

1

1

1

1

1

1

1980

P

1

1

-

1

1

1

1

1

1

1

UAMH 10579

S

1

1

1

1

1

1

1

1

1

1

1

1

C0375214F

L

1

1

-

1

1

1

1

1

1

1

lichen metagenome

L

1

1

-

1

1

1

1

1

1

1

Af293

A

1

-

-

1

1

1

1

1

1

1

1

1

1

1

1

1

nf4

P

1

1

-

1

1

1

1

1

MD-8

P

1

-

-

1

1

-

1

1

1

1

ATCC 48331

P

1

1

1

1

1

1

1

1

1

1

SRC1lrK2f

P

1

1

1

1

1

1

1

1

1

1 1

RIB40

S

1

-

-

1

1

NIH2624

A

1

-

1

1

1

1

UCRNP2

P

1

1

1

1

1

1

1

1

1

Mel28

EM

1

1

-

1

1

1

1

1

1

1

G3

EM

1

1

-

1

1

1

1

1

1

1

T7

EM

1

1

-

1

1

1

1

1

1

1

ATCC 24927

N

1

1

-

1

1

1

1

1

1

1

CBS 200.50

N

1

1

-

1

1

1

1

1

1

1

CBS H-5679

N

1

1

-

1

1

1

1

1

1

1

S288C

S

-

-

-

-

-

1

1

-

1

1

-

1

1

1

-

Cdc42-like

1

1

1

1

1

1

1

1

1

-

Cdc42-like

-

R-like

NRRL Y-17796

S

-

-

-

R-like

Co 90-125

S

-

-

-

-

-

972h-

S

-

-

-

-

-

-

1

1

-

1

NRRL Y-17804

S

1

1

-

1

1

1

1

1

1

1

FM1318

S

-

-

-

-

CBS 4309

S

-

-

-

-

CLIB122

S

-

-

1

1

RU7

A

-

-

-

-

-

-

1

1

-

JEC21

A

-

-

-

-

-

-

1

1

-

1

CBS 7118

S

1

1

-

1

1

1

1

1

1

1

AG-1 IB

P

1

1

-

1

1

1

1

1

1

1

1

MF3/22

P

1

1

-

1

1

1

1

1

1

1

CBS 10117

S

1

1

1

1

1

1

1

1

1

1

-

-

-

1

1

1

1

DAOMC 238049

P

-

-

3

-

1

1

-

-

1

-

1

1

1

1

CBS 10092

P

-

-

-

-

-

1

EP

-

1

MCA 3882

1

1

1

1

1-1 BBBD Race 1

P

2

1

-

1

1

1

1

1

1

1

PST-78

P

2

1

-

1

1

1

1

1

1

1

2

2

-

1

1

2

1

1

1

521

98AG31

P

P

-

-

-

2

1

G11

P

2

2

-

1

1

1

1

1

1

DAOM 227022

AM

1

2

-

1

1

1

1

1

1

1

DAOM 197198w

AM

2

1

-

1

1

1

1

1

1

1 1

NC121

AM

1

2

-

1

1

2

1

1

1

KOD1229

S

1

1

-

1, R-like

1

2

1

2

1

2

NRRL 6456

S

1

1

-

1

1

1

2

2

2

1

NRRL 6351

S

1

1

-

1

1

1

1

1

1

1

CBS 142.35

S

-

-

-

R-like

-

2

-

1

2

1

NRRL 2496

S

-

-

-

R-like x2

-

-

1

3

3

1

RA 99-880

S

-

-

-

R-like x2

-

-

2

2

2

2

CBS 931.73

A

-

2

-

1, R-like x 2

-

3

2

3

3

4

Smittium simulii

SWE-8-4

IS

-

1

-

1

-

1

1

1

1

1

Allomyces macrogynus

ATCC 38327

S

1

2

-

2

-

-

2

2

4

2

-

1, R-like

-

-

1

1

2

1

2

-

-

1

1

2

1

Basidiobolus meristosporus

Batrachochytrium dendrobatidis

JAM81

A

-

1

Batrachochytrium salamandrivorans

A MFP1 5 /1

A

-

1

Metazoa

Fig. 25.5 Distribution of Noxs and predicted regulators in Ascomycota fungi. The types of fungi were categorized as follows. S, saprophyte; P, plant pathogen; E, endophyte; I, insect pathogen; L, lichen; A, animal pathogen; EM, ectomycorrhiza; N, nematode-trapping; EP, epiphyte;

AM, arbuscular mycorrhiza; IS, insect symbiont. The numbers represent the number of copies of the genes (or their absence) found in the genome of each of the various fungal species

physically with NoxR and themselves, via a Phox and Bem1 (PB1) protein interaction module that is conserved in animals, fungi, protists and plants [99], we have proposed that these proteins are key regulators of the fungal Nox complex. The presence of analogues of these and other protein-protein interaction domains in mammalian Nox regulators (Fig. 25.4), suggest a potential link between the protein machinery required for fungal polarity establishment and the Nox complex [77]. Interrogation of the 65 fungal genomes showed that BemA and Cdc24 are consistently

conserved in all species, including those without a nox gene; a result that is perhaps not too surprising given polarised growth is a fundamental growth feature of the filamentous fungi. In addition, the small GTPase Cdc42, which is the closest homologue of RacA, and can bind specifically to BemA [82], is also highly conserved in species of the fungal kingdom. This analysis suggests that fungal species may have lost Nox catalytic functions during their evolution but have retained or modified Nox regulators, that may now confer other functions.

440

8

Glossary

Conclusions

Although a role for ROS in regulating various fungal differentiations has been clearly established many questions remain unanswered. Are NoxD and PlsA the scaffold equivalents of the mammalian p22phox for the NoxA and NoxB enzyme complexes respectively? What is the mechanism for the recruitment of NoxR to the cell membrane and activation of NoxA and NoxB? The physical interaction of Cdc24 (a GEF) and BemA with NoxR suggests these two cell polarity regulators play a key role but are there still other regulators to be discovered? What is the function of NoxC in fungal species that contain this isoform and how is it activated? What is the mechanism for activation and recruitment of RacA to the plasma membrane to activate Nox? In animals Rho GDIs are known to be key regulators of Rho GTPase function [64]. There is increasing evidence from mammalian systems that kinases acting on either Rho GDIs or Rho GTPases themselves, act to regulate formation of complexes. Lipids and phosphatases may also act in concert with kinases, to control the overall specificity and dynamics of Rho GTPase action [64]. Further insights into how the fungal Rho GTPase-GDI cycle is regulated will help bridge our gap in knowledge between fungal sensing of differentiation signals and transduction of those signals within the cell to activate Nox. How ROS signal within the cell and in what molecular form, remains a key question to understand ROS mediated differentiation processes. Both superoxide and hydrogen peroxide are able to affect protein function by reacting with sulphur-containing groups, such as the invariant cysteine residue found in protein tyrosine phosphatases and cysteinerich regions of transcription factors [100, 101]. Tyrosine phosphorylation is widely used by multicellular organisms to control cell fate decisions and cytoskeleton dynamics [102]. Both tyrosine kinases and protein tyrosine phosphatases (PTPs) exert signal specificity through subcellular targeting as well as catalytic domain specificity. An analysis of fungal genomes identified homologues of the mammalian protein tyrosine phosphatases but whether they have a role in ROS signalling remains to be tested. MAP kinase pathways are also regulated by ROS through the signalling intermediates, thioredoxin and glutaredoxin, which serve as both sensors and transducers of oxidative stress [103, 104]. Oxidants are also known to trigger the generation of Ca2+ signals, in part through activation of calcium channels

[105]. Genetic evidence supports the hypothesis of a link between ROS production and Ca2+ signalling for at least two developmental systems; root hair development in A. thaliana [106, 107] and multi-cellular development in Dictyostelium discoideum [108]. Whether there is cross talk between Nox generated ROS and Ca2+ signalling in filamentous fungal cell development remains to be determined. In summary, NADPH oxidase generated ROS is an important signalling system in fungi to control various cell differentiation processes but also has a role in defense. Activation of the fungal Nox has been shown to require a p67phox-like regulator, NoxR, and the small GTPase RacA, functions analogous to those in the well-studied human phagocytic gp91phox. However, there are structural differences between NoxR and p67phox. Protein-protein interaction studies have shown that NoxR interacts with the polarity proteins Cdc24 and BemA, providing an interesting link between Nox and cell polarity in fungi. The survey here of 65 fungal genomes indicates that Nox have been retained in a wide range of fungal taxa suggesting Nox is an ancestral enzyme. While it has been lost in some lineages, it is almost universally retained in fungi that undergo multi-cellular development. The powerful genetic systems available for many fungal species make them an ideal group of organisms to investigate the fundamental role of NADPH oxidase generated ROS signalling in eukaryotic cells.

Glossary Ascogonium (plural ascogonia) Female sexual organ in ascomycetous fungus from which asci develop. Ascospore A sexual spore that develops within an ascus in ascomycetous fungi. Ascus (plural asci) A sac like structure from which sexual spores (ascospores) develop. Cleistothecium (plural cleistothecia) A globose, closed fruiting body (ascocarp) from which ascospores develop and are released. Conidiation A biological process in which filamentous fungi reproduce asexually from spores. Peridium Cell wall of ascomata comprised of densely woven hyphae. Perithecium A round or flask shaped fruiting body with a pore through which ascospores are discharged. Sclerotium A persistent, vegetative, resting stage of certain fungi.

References

References 1. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87(1):245–313 2. Lambeth JD (2004) NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol 4(3):181–189 3. Suh YA, Arnold RS, Lassegue B, Shi J, Xu X, Sorescu D et al (1999) Cell transformation by the superoxide-generating oxidase Mox1. Nature 401(6748):79–82 4. Doke N (1983) Involvement of superoxide anion generation in the hypersensitive response of potato tuber tissues to infection with an incompatible race of Phytophthora infestans and to the hyphal wall components. Physiol Plant Pathol 23:345–357 5. Auh CK, Murphy TM (1995) Plasma membrane redox enzyme Is involved in the synthesis of O2– and H2O2 by Phytophthora elicitor-stimulated rose cells. Plant Physiol 107(4):1241–1247 6. Grant M, Brown I, Adams S, Knight M, Ainslie A, Mansfield J (2000) The RPM1 plant disease resistance gene facilitates a rapid and sustained increase in cytosolic calcium that is necessary for the oxidative burst and hypersensitive cell death. Plant J 23(4): 441–450 7. Simon-Plas F, Elmayan T, Blein J-P (2002) The plasma membrane oxidase NtrbohD is responsible for AOS production in elicited tobacco cells. Plant J 31(2):137–147 8. Torres MA, Dangl JL, Jones JD (2002) Arabidopsis gp91 phox homologues AtrbohD and AtrbohF are required for accumulation of reactive oxygen intermediates in the plant defense response. Proc Natl Acad Sci U S A 99(1):517–522 9. Yoshioka H, Numata N, Nakajima K, Katou S, Kawakita K, Rowland O et al (2003) Nicotiana benthamiana gp91phox homologs NbrbohA and NbrbohB participate in H2O2 accumulation and resistance to Phytophthora infestans. Plant Cell 15(3): 706–718 10. Forman HJ, Fukuto JM, Torres M (2004) Redox signaling: thiol chemistry defines which reactive oxygen and nitrogen species can act as second messengers. Am J Physiol Cell Physiol 287:246–256 11. Kwak JM, Mori IC, Pei ZM, Leonhardt N, Torres MA, Dangl JL et al (2003) NADPH oxidase AtrbohD and AtrbohF genes function in ROS-dependent ABA signaling in Arabidopsis. EMBO J 22(11):2623–2633 12. Torres MA, Dangl JL (2005) Functions of the respiratory burst oxidase in biotic interactions, abiotic stress and development. Curr Opin Plant Biol 8(4):397–403 13. Kaya H, Nakajima R, Iwano M, Kanaoka MM, Kimura S, Takeda S et al (2014) Ca2+-activated reactive oxygen species production by Arabidopsis RbohH and RbohJ is essential for proper pollen tube tip growth. Plant Cell 26(3):1069–1080 14. Aguirre J, Ríos-Momberg M, Hewitt D, Hansberg W (2005) Reactive oxygen species and development in microbial eukaryotes. Trends Microbiol 13(3):111–118 15. Lara-Ortíz T, Riveros-Rosas H, Aguirre J (2003) Reactive oxygen species generated by microbial NADPH oxidase NoxA regulate sexual development in Aspergillus nidulans. Mol Microbiol 50(4): 1241–1255 16. Malagnac F, Lalucque H, Lepère G, Silar P (2004) Two NADPH oxidase isoforms are required for sexual reproduction and ascospore germination in the filamentous fungus Podospora anserina. Fungal Genet Biol 41(11):982–997 17. Haedens V, Malagnac F, Silar P (2005) Genetic control of an epigenetic cell degeneration syndrome in Podospora anserina. Fungal Genet Biol 42(6):564–577 18. Silar P (2005) Peroxide accumulation and cell death in filamentous fungi induced by contact with a contestant. Mycol Res 109(Pt 2): 137–149

441 19. Brun S, Malagnac F, Bidard F, Lalucque H, Silar P (2009) Functions and regulation of the Nox family in the filamentous fungus Podospora anserina: a new role in cellulose degradation. Mol Microbiol 74(2):480–496 20. Lehr NA, Wang Z, Li N, Hewitt DA, López-Giráldez F, Trail F et al (2014) Gene expression differences among three Neurospora species reveal genes required for sexual reproduction in Neurospora crassa. PLoS One 9(10):e110398 21. Coppin E, Berteaux-Lecellier V, Bidard F, Brun S, RuprichRobert G, Espagne E et al (2012) Systematic deletion of homeobox genes in Podospora anserina uncovers their roles in shaping the fruiting body. PLoS One 7(5):e37488 22. Gautier V, Tong LCH, Nguyen T-S, Debuchy R, Silar P (2018) PaPro1 and IDC4, two genes controlling stationary phase, sexual development and cell degeneration in Podospora anserina. J Fungi (Basel) 4(3) 23. Kayano Y, Tanaka A, Akano F, Scott B, Takemoto D (2013) Differential roles of NADPH oxidases and associated regulators in polarized growth, conidiation and hyphal fusion in the symbiotic fungus Epichloë festucae. Fungal Genet Biol 56:87–97 24. Nordberg J, Arner ES (2001) Reactive oxygen species, antioxidants, and the mammalian thioredoxin system. Free Radic Biol Med 31(11):1287–1312 25. Cano-Domínguez N, Álvarez-Delfín K, Hansberg W, Aguirre J (2008) The NADPH oxidases NOX-1 and NOX-2 require the regulatory subunit NOR-1 to control cell differentiation and growth in Neurospora crassa. Eukaryot Cell 7(8):1352–1361 26. Dirschnabel DE, Nowrousian M, Cano-Domínguez N, Aguirre J, Teichert I, Kück U (2014) New insights into the roles of NADPH oxidases in sexual development and ascospore germination in Sordaria macrospora. Genetics 196:729–744 27. Kicka S, Silar P (2004) PaASK1, a mitogen-activated protein kinase kinase kinase that controls cell degeneration and cell differentiation in Podospora anserina. Genetics 166(3):1241–1252 28. Kicka S, Bonnet C, Sobering AK, Ganesan LP, Silar P (2006) A mitotically inheritable unit containing a MAP kinase module. Proc Natl Acad Sci U S A 103(36):13445–13450 29. Silar P, Haedens V, Rossignol M, Lalucque H (1999) Propagation of a novel cytoplasmic, infectious and deleterious determinant is controlled by translational accuracy in Podospora anserina. Genetics 151(1):87–95 30. Read ND, Goryachev AB, Lichius A (2012) The mechanistic basis of self-fusion between conidial anastomosis tubes during fungal colony initiation. Fungal Biol Rev 26:1–11 31. Tanaka A, Christensen MJ, Takemoto D, Park P, Scott B (2006) Reactive oxygen species play a role in regulating a fungusperennial ryegrass mutualistic association. Plant Cell 18:1052– 1066 32. Lambou K, Malagnac F, Barbisan C, Tharreau D, Lebrun MH, Silar P (2008) The crucial role during ascospore germination of the Pls1 tetraspanin in Podospora anserina provides an example of the convergent evolution of morphogenetic processes in fungal plant pathogens and saprobes. Eukaryot Cell 7:1809–1818 33. Egan M, Wang Z-Y, Jones MA, Smirnoff N, Talbot NJ (2007) Generation of reactive oxygen species by fungal NADPH oxidases is required for rice blast disease. Proc Natl Acad Sci U S A 104 (July 10):11772–11777 34. Ryder LS, Dagdas YF, Mentlak TA, Kershaw MJ, Thornton CR, Schuster M et al (2013) NADPH oxidases regulate septin-mediated cytoskeletal remodeling during plant infection by the rice blast fungus. Proc Natl Acad Sci U S A 110(8):3179–3184 35. Takemoto D, Tanaka A, Scott B (2006) A p67phox-like regulator is recruited to control hyphal branching in a fungal-grass mutualistic symbiosis. Plant Cell 18(10):2807–2821 36. Segmüller N, Kokkelink L, Giesbert S, Odinius D, van Kan J, Tudzynski P (2008) NADPH oxidases are involved in

442 differentiation and pathogenicity in Botrytis cinerea. Mol Plant Microbe Interact 21(6):808–819 37. Semighini CP, Harris SD (2008) Regulation of apical dominance in Aspergillus nidulans hyphae by reactive oxygen species. Genetics 179(4):1919–1932 38. Christensen MJ, Bennett RJ, Schmid J (2002) Growth of Epichloë/ Neotyphodium and p-endophytes in leaves of Lolium and Festuca grasses. Mycol Res 106:93–106 39. Christensen MJ, Bennett RJ, Ansari HA, Koga H, Johnson RD, Bryan GT et al (2008) Epichloë endophytes grow by intercalary hyphal extension in elongating grass leaves. Fungal Genet Biol 45: 84–93 40. Tan YY, Spiering MJ, Scott V, Lane GA, Christensen MJ, Schmid J (2001) In planta regulation of extension of an endophytic fungus and maintenance of high metabolic rates in its mycelium in the absence of apical extension. Appl Environ Microbiol 67(12): 5377–5383 41. Voisey CR (2010) Intercalary growth in hyphae of filamentous fungi. Fungal Biol Rev 24:123–131 42. Tanaka A, Takemoto D, Hyon GS, Park P, Scott B (2008) NoxA activation by the small GTPase RacA is required to maintain a mutualistic symbiotic association between Epichloë festucae and perennial ryegrass. Mol Microbiol 68(5):1165–1178 43. Becker Y, Eaton CJ, Brasell E, May KJ, Becker M, Hassing B et al (2015) The fungal cell wall integrity MAPK cascade is crucial for hyphal network formation and maintenance of restrictive growth of Epichloë festucae in symbiosis with Lolium perenne. Mol Plant Microbe Interact 28:69–85 44. Tanaka A, Cartwright GM, Saikia S, Kayano Y, Takemoto D, Kato M et al (2013) ProA, a transcriptional regulator of fungal fruiting body development, regulates leaf hyphal network development in the Epichloë festucae-Lolium perenne symbiosis. Mol Microbiol 90(3):551–568 45. Becker M, Becker Y, Green K, Scott B (2016) The endophytic symbiont Epichloë festucae establishes an epiphyllous net on the surface of Lolium perenne leaves by development of an expressorium, an appressorium-like leaf exit structure. New Phytol 211:240–254 46. Noorifar N, Savoian MS, Ram A, Lukito Y, Hassing B, Weikert TW et al (2021) Chitin deacetylases are required for Epichloë festucae endophytic cell wall remodeling during establishment of a mutualistic symbiotic interaction with Lolium perenne. Mol Plant Microbe Interact 34(10):1181–1192 47. Giesbert S, Schurg T, Scheele S, Tudzynski P (2008) The NADPH oxidase Cpnox1 is required for full pathogenicity of the ergot fungus Claviceps purpurea. Mol Plant Pathol 9(3):317–327 48. Siegmund U, Marschall R, Tudzynski P (2015) BcNoxD, a putative ER protein, is a new component of the NADPH oxidase complex in Botrytis cinerea. Mol Microbiol 95:988–1005 49. Kim H-j, Chen C, Kabbage M, Dickman MB (2011) Identification and characterization of Sclerotinia sclerotiorum NADPH oxidases. Appl Environ Microbiol 77(21):7721–7729 50. Yang SL, Chung KR (2013) Similar and distinct roles of NADPH oxidase components in the tangerine pathotype of Alternaria alternata. Mol Plant Pathol 14(6):543–556 51. Wang L, Mogg C, Walkowiak S, Joshi M, Subramaniam R (2014) Characterization of NADPH oxidase genes NoxA and NoxB in Fusarium graminearum. Can J Plant Pathol 36:12–21 52. Zhang C, Lin Y, Wang J, Wang Y, Chen M, Norvienyeku J et al (2016) FgNoxR, a regulatory subunit of NADPH oxidases, is required for female fertility and pathogenicity in Fusarium graminearum. FEMS Microbiol Lett 363(1):fnv223 53. Hernández-Oñate MA, Esquivel-Naranjo EU, Mendoza-MendozaA, Stewart A, Herrera-Estrella AH (2012) An injury-response mechanism conserved across kingdoms determines entry of the

References fungus Trichoderma atroviride into development. Proc Natl Acad Sci U S A 109(37):14918–14923 54. Turrà D, El Ghalid M, Rossi F, Di Pietro A (2015) Fungal pathogen uses sex pheromone receptor for chemotropic sensing of host plant signals. Nature 527(7579):521–524 55. Nordzieke DE, Fernandes TR, El Ghalid M, Turrà D, Di Pietro A (2019) NADPH oxidase regulates chemotropic growth of the fungal pathogen Fusarium oxysporum towards the host plant. New Phytol 224(4):1600–1612 56. Pick E (2020) Cell-free NADPH oxidase activation assay: A triumph of reductionism. 3rd ed. Neutrophil methods and protocols. Springer, New York 57. Diebold BA, Bokoch GM (2001) Molecular basis for Rac2 regulation of phagocyte NADPH oxidase. Nat Immunol 2(3):211–215 58. Groemping Y, Rittinger K (2005) Activation and assembly of the NADPH oxidase: a structural perspective. Biochem J 386(Pt 3): 401–416 59. Takemoto D, Tanaka A, Scott B (2007) NADPH oxidases in fungi: diverse roles of reactive oxygen species in fungal cellular differentiation. Fungal Genet Biol 44:1065–1076 60. Lewit-Bentley A, Rety S (2000) EF-hand calcium-binding proteins. Curr Opin Struct Biol 10(6):637–643 61. Rinnerthaler M, Buttner S, Laun P, Heeren G, Felder TK, Klinger H et al (2012) Yno1p/Aim14p, a NADPH-oxidase ortholog, controls extramitochondrial reactive oxygen species generation, apoptosis, and actin cable formation in yeast. Proc Natl Acad Sci U S A 109(22):8658–8663 62. Rossi DCP, Gleason JE, Sanchez H, Schatzman SS, Culbertson EM, Johnson CJ et al (2017) Candida albicans FRE8 encodes a member of the NADPH oxidase family that produces a burst of ROS during fungal morphogenesis. PLoS Pathog 13(12):e1006763 63. Segal AW, West I, Wientjes F, Nugent JH, Chavan AJ, Haley B et al (1992) Cytochrome b-245 is a flavocytochrome containing FAD and the NADPH-binding site of the microbicidal oxidase of phagocytes. Biochem J 284:781–788 64. DerMardirossian C, Bokoch GM (2005) GDIs: central regulatory molecules in Rho GTPase activation. Trends Cell Biol 15(7): 356–363 65. Hoyal CR, Gutierrez A, Young BM, Catz SD, Lin JH, Tsichlis PN et al (2003) Modulation of p47phox activity by site-specific phosphorylation: Akt-dependent activation of the NADPH oxidase. Proc Natl Acad Sci U S A 100(9):5130–5135 66. Inanami O, Johnson JL, McAdara JK, El-Benna J, Faust LR, Newburger PE et al (1998) Activation of the leukocyte NADPH oxidase by phorbol ester requires the phosphorylation of p47PHOX on serine 303 or 304. J Biol Chem 273(16):9539–9543 67. Leto TL, Adams AG, de Mendez I (1994) Assembly of the phagocyte NADPH oxidase: binding of Src homology 3 domains to proline-rich targets. Proc Natl Acad Sci U S A 91(22): 10650–10654 68. Sumimoto H, Kage Y, Nunoi H, Sasaki H, Nose T, Fukumaki Y et al (1994) Role of Src homology 3 domains in assembly and activation of the phagocyte NADPH oxidase. Proc Natl Acad Sci U S A 91(12):5345–5349 69. Heyworth PG, Bohl BP, Bokoch GM, Curnutte JT (1994) Rac translocates independently of the neutrophil NADPH oxidase components p47phox and p67phox. Evidence for its interaction with flavocytochrome b558. J Biol Chem 269(49):30749–30752 70. Ago T, Takeya R, Hiroaki H, Kuribayashi F, Ito T, Kohda D et al (2001) The PX domain as a novel phosphoinositide-binding module. Biochem Biophys Res Commun 287(3):733–738 71. Kanai F, Liu H, Field SJ, Akbary H, Matsuo T, Brown GE et al (2001) The PX domains of p47phox and p40phox bind to lipid products of PI3K. Nat Cell Biol 3(7):675–678

References 72. Lacaze I, Lalucque H, Siegmund U, Silar P, Brun S (2015) Identification of NoxD/Pro41 as the homologue of the p22phox NADPH oxidase subunit in fungi. Mol Microbiol 95:1006–1024 73. Nowrousian M, Frank S, Koers S, Strauch P, Weitner T, Ringelberg C et al (2007) The novel ER membrane protein PRO41 is essential for sexual development in the filamentous fungus Sordaria macrospora. Mol Microbiol 64(4):923–937 74. Meijles DN, Howlin BJ, Li JM (2012) Consensus in silico computational modelling of the p22phox subunit of the NADPH oxidase. Comput Biol Chem 39:6–13 75. Siegmund U, Heller J, van Kan JA, Tudzynski P (2013) The NADPH oxidase complexes in Botrytis cinerea: evidence for a close association with the ER and the tetraspanin Pls1. PLoS One 8(2):e55879 76. Scott B (2015) Conservation of fungal and animal nicotinamide adenine dinucleotide phosphate oxidase complexes. Mol Microbiol 95(6):910–913 77. Takemoto D, Kamakura S, Saikia S, Becker Y, Wrenn R, Tanaka A et al (2011) Polarity proteins Bem1 and Cdc24 are components of the filamentous fungal NADPH oxidase complex. Proc Natl Acad Sci U S A 108(7):2861–2866 78. Boyce KJ, Hynes MJ, Andrianopoulos A (2003) Control of morphogenesis and actin localization by the Penicillium marneffei RAC homolog. J Cell Sci 116(Pt 7):1249–1260 79. Chen C, Dickman MB (2004) Dominant active Rac and dominant negative Rac revert the dominant active Ras phenotype in Colletotrichum trifolii by distinct signalling pathways. Mol Microbiol 51(5):1493–1507 80. Paduch M, Jelen F, Otlewski J (2001) Structure of small G proteins and their regulators. Acta Biochim Pol 48:829–850 81. Lapouge K, Smith SJ, Walker PA, Gamblin SJ, Smerdon SJ, Rittinger K (2000) Structure of the TPR domain of p67phox in complex with Rac.GTP. Mol Cell 6(4):899–907 82. Kayano Y, Tanaka A, Takemoto D (2018) Two closely related Rho GTPases, Cdc42 and RacA, of the endophytic fungus Epichloë festucae have contrasting roles for ROS production and symbiotic infection synchronized with the host plant. PLoS Pathog 14: e1006840 83. Kamiguti AS, Serrander L, Lin K, Harris RJ, Cawley JC, Allsup DJ et al (2005) Expression and activity of NOX5 in the circulating malignant B cells of hairy cell leukemia. J Immunol 175(12): 8424–8430 84. Fu C, Ao J, Dettmann A, Seiler S, Free SJ (2014) Characterization of the Neurospora crassa cell fusion proteins, HAM-6, HAM-7, HAM-8, HAM-9, HAM-10, AMPH-1 and WHI-2. PLoS One 9(10):e107773 85. Ikeda S, Yamaoka-Tojo M, Hilenski L, Patrushev NA, Anwar GM, Quinn MT et al (2005) IQGAP1 regulates reactive oxygen speciesdependent endothelial cell migration through interacting with Nox2. Arterioscler Thromb Vasc Biol 25(11):2295–2300 86. Erickson JW, Cerione RA, Hart MJ (1997) Identification of an actin cytoskeletal complex that includes IQGAP and the Cdc42 GTPase. J Biol Chem 272(39):24443–24447 87. Adachi M, Kawasaki A, Nojima H, Nishida E, Tsukita S (2014) Involvement of IQGAP family proteins in the regulation of mammalian cell cytokinesis. Genes Cells 19(11):803–820 88. Briggs MW, Sacks DB (2003) IQGAP proteins are integral components of cytoskeletal regulation. EMBO Rep 4(6):571–574 89. Faix J, Clougherty C, Konzok A, Mintert U, Murphy J, Albrecht R et al (1998) The IQGAP-related protein DGAP1 interacts with Rac and is involved in the modulation of the F-actin cytoskeleton and control of cell motility. J Cell Sci 111(Pt 20):3059–3071

443 90. Wendland J, Philippsen P (2002) An IQGAP-related protein, encoded by AgCYK1, is required for septation in the filamentous fungus Ashbya gossypii. Fungal Genet Biol 37(1):81–88 91. Marschall R, Tudzynski P (2016) BcIqg1, a fungal IQGAP homolog, interacts with NADPH oxidase, MAP kinase and calcium signaling proteins and regulates virulence and development in Botrytis cinerea. Mol Microbiol 101(2):281–298 92. Cano-Dominguez N, Bowman B, Peraza-Reyes L, Aguirre J (2019) Neurospora crassa NADPH oxidase NOX-1 Is localized in the vacuolar system and the plasma membrane. Front Microbiol 10:1825 93. Masloff S, Pöggeler S, Kück U (1999) The pro1+ gene from Sordaria macrospora encodes a C6 zinc finger transcription factor required for fruiting body development. Genetics 152(1):191–199 94. Steffens EK, Becker K, Krevet S, Teichert I, Kück U (2016) Transcription factor PRO1 targets genes encoding conserved components of fungal developmental signaling pathways. Mol Microbiol 102(5):792–809 95. Lalucque H, Silar P (2003) NADPH oxidase: an enzyme for multicellularity? Trends Microbiol 11(1):9–12 96. Koga H, Terasawa H, Nunoi H, Takeshige K, Inagaki F, Sumimoto H (1999) Tetratricopeptide repeat (TPR) motifs of p67phox participate in interaction with the small GTPase Rac and activation of the phagocyte NADPH oxidase. J Biol Chem 274(35):25051–25060 97. Wong HL, Pinontoan R, Hayashi K, Tabata R, Yaeno T, Hasegawa K et al (2007) Regulation of rice NADPH oxidase by binding of Rac GTPase to its N-terminal extension. Plant Cell 19(12): 4022–4034 98. Butty AC, Perrinjaquet N, Petit A, Jaquenoud M, Segall JE, Hofmann K et al (2002) A positive feedback loop stabilizes the guanine-nucleotide exchange factor Cdc24 at sites of polarization. EMBO J 21(7):1565–1576 99. Sumimoto H, Kamakura S, Ito T (2007) Structure and function of the PB1 domain, a protein interaction module conserved in animals, fungi, amoebas, and plants. Sci STKE 2007(401):re6 100. Finkel T (2003) Oxidant signals and oxidative stress. Curr Opin Cell Biol 15(2):247–254 101. Liu H, Colavitti R, Rovira II, Finkel T (2005) Redox-dependent transcriptional regulation. Circ Res 97(10):967–974 102. Alonso A, Sasin J, Bottini N, Friedberg I, Friedberg I, Osterman A et al (2004) Protein tyrosine phosphatases in the human genome. Cell 117(6):699–711 103. Finkel T (2000) Redox-dependent signal transduction. FEBS Lett 476(1–2):52–54 104. Fujino G, Noguchi T, Takeda K, Ichijo H (2006) Thioredoxin and protein kinases in redox signaling. Semin Cancer Biol 16(6): 427–435 105. Waring P (2005) Redox active calcium ion channels and cell death. Arch Biochem Biophys 434(1):33–42 106. Carol RJ, Takeda S, Linstead P, Durrant MC, Kakesova H, Derbyshire P et al (2005) A RhoGDP dissociation inhibitor spatially regulates growth in root hair cells. Nature 438(7070): 1013–1016 107. Foreman J, Demidchik V, Bothwell JH, Mylona P, Miedema H, Torres MA et al (2003) Reactive oxygen species produced by NADPH oxidase regulate plant cell growth. Nature 422(6930): 442–446 108. Lardy B, Bof M, Aubry L, Paclet MH, Morel F, Satre M et al (2005) NADPH oxidase homologs are required for normal cell differentiation and morphogenesis in Dictyostelium discoideum. Biochim Biophys Acta 1744(2):199–212

26

Plant NADPH Oxidases Gad Miller and Ron Mittler

Abstract

Reactive oxygen species (ROS) are signaling molecules that play important roles in many processes in plants, including development, immunity, and acclimation to different environmental stimuli. The intensity, duration and localization of ROS signals are determined by a delicate equilibrium between ROS production, scavenging, and transport mechanism within the ROS network. Within this network of genes, plant NADPH oxidases (NOXs), or respiratory burst oxidase homologues (RBOHs), comprise a small, but highly important, conserved gene family that generates ROS. RBOH-generated ROS drive signaling cascades regulating key cellular processes, tissue specific programs, and systemic responses via cell-to-cell communication. In addition, RBOH proteins serve as regulatory hubs for calcium signaling and other secondary messengers, such as NO. In this chapter we summarize the current understanding of the activity, regulation and different roles of RBOH proteins in plants. Keywords

RBOH · Plant · ROS · Development · Stress · NADPH oxidase · Oxidative burst · Signal transduction · Systemic signaling

G. Miller (✉) The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan, Israel e-mail: [email protected] R. Mittler The Division of Plant Sciences and Technology and Interdisciplinary Plant Group, College of Agriculture, Food and Natural Resources, Christopher S. Bond Life Sciences Center University of Missouri, Columbia, MO, USA e-mail: [email protected]

1

Introduction

The reactive oxygen species (ROS) network is an evolutionary conserved network found in all aerobic organisms [1]. This network includes an array of genes and proteins that control ROS production, perception, transport and scavenging in cells [2]. In photosynthetic eukaryotes, the electron transport chains in chloroplasts and mitochondria together with peroxisomes are the primary sites of ROS formation. Although ROS are considered toxic byproducts of aerobic metabolism, they also play prominent roles in cellular signaling [3, 4]. ROS are also produced at almost any cellular compartment of cells by proteins or molecules with a sufficiently high redox potential to excite or donate an electron to atmospheric oxygen [5]. Early research into ROS metabolism primarily focused on their toxic potential that cause cell death by indiscriminate oxidation during stress conditions or disease [6, 7]. However, during the last two decades multiple studies focused on ROS as important signal transduction molecules acting in stress responses as well as other processes during favorable growth conditions. ROS including singlet oxygen (1O2), superoxide (O2•–) and hydrogen peroxide (H2O2) serve as secondary messengers to control a large array of biological processes ranging from the regulation of development and growth to responses to biotic and/or abiotic stimuli [1]. Oxidative stress only results when ROS accumulation exceeds the cellular capacity for ROS detoxification and the activity of damage repair systems in cell [2]. Plants employ a large array of ROS scavenging enzymes including superoxide dismutase (SOD), ascorbate peroxidase (APX), and catalase, as well as many small antioxidants such as ascorbic acid and glutathione [6, 8]. The capacity of cells to generate ROS on the one hand and detoxify them on the other under any given condition or situation is key to their function as signaling molecules [5, 9].

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_26

445

446

G. Miller and R. Mittler

The ROS network also include important ROS producing enzymes. These include xanthine oxidase, glucose oxidase, and different classical peroxidases that generate O2•– and H2O2 [9]. But perhaps the most important ROS producing enzymes considered as the engines of ROS signaling are NADPH oxidases (NOXs) [10]. Here we describe the discovery of plant NOXs, their structure, function, genetics and regulation, and summarize the current understanding of their roles in plants.

2

The Discovery of Plant NOXs

Early studies of plant NOXs activity and regulation have focused on their role in plant immunity in response to pathogens. NOX activity is tightly associated with and essential for proper immunity of the plant in responses to pathogens. The hypersensitive response (HR), a form of cell death associated with plants resistance to pathogens, was first described in 1902 in wheat cultivars in response to leaf rust (Puccinia disersa) by H. Marshal Ward [11]. Yet, it took eight decades until the studies made by the Doke group in 1983 described superoxide (O2•–) production in response to Phytophtora infestans in potato tubers during HR development [12, 13]. This rapid generation O2•– was shown to be NADPH-dependent [14]. Termed ‘oxidative burst’, this rapid and transient production of large amounts of ROS is one of the earliest noticeable aspects of a plant’s defense response that occurs prior to cell death and has been associated with many forms of HR [11, 15]. H2O2 generated during the oxidative burst was shown to play a plethora of functions, driving cross-linking of structural cell wall proteins, local trigger of programmed cell death, and also acting as a diffusible signal inducing expression of genes for cellular protection in unchallenged cells adjacent to the region of infection [16]. Questions regarding the origin of the oxidative burst received growing attention during the following decade but were still somewhat controversial. The debate over the source of ROS in the oxidative burst focused on two options: (1) the pH-dependent generation of H2O2 by cell-wall peroxidase, and (2) the action of analogs of the phagocyte NADPH oxidase system. The production of H2O2 by peroxidases involving oxidation of NAD(P)H and other reductants by molecular oxygen have been characterized in plant cell walls at the late 1950s. These publications may have set the ground searching for peroxidases as the origin of the oxidative burst [17–22]. The arguments against NOXs as the source of the oxidative burst relied mainly on differences in its definition between mammalian cell systems and plants, and that

phagocytic NOXs are resistant to azides and cyanides whereas in plants the oxidative burst is sensitive to these agents [23–25]. Furthermore, the reliance of some studies supporting the involvement of plant NOXs on using diphenyleneiodonium (DPI) as a NOX inhibitor was a matter for concern, as it is not specific to NOXs [23, 24, 26, 27]. In the absence of direct evidence for the involvement plant NOXs, an alternative system based on cell wall peroxidases that can rapidly generate H2O2 in vitro under alkaline conditions, which were shown to be essential for the oxidative burst was suggested [28]. Yet, in addition to the fact that DPI was able to completely block the oxidative burst similar to its inhibitory activity in human cells [16], evidence showing the involvement of small G-proteins in activating it in soybean cell cultures, as it was known to function in mammalian cells, was another compelling evidence in support of the involvement of plant NOX homologues [29, 30]. The debate continued even after the cloning of the plant RBOHs. RBOHs were identified on the basis of similarity to the mammalian NOX2 (gp91phox) subunit cloned first from rice in 1996 (RBOHA; [31]) followed by six homologues (RBOHA-F) in Arabidopsis thaliana (At) in 1998 [32, 33]. Ten RBOH gene homologous to gp91phox were eventually identified in the Arabidopsis genome (AtRBOHA-J) [32, 34, 35], and 9 in rice [36]. The RBOH sequences contained an extended N-terminal region that contained putative Ca2+ EF hand binding motifs and similarity to a human GTPase activating protein, suggesting direct regulatory involvement for Ca2+ and G-proteins activation of the oxidative burst. Initial reports that antibodies against the mammalian p22phox, p47phox, and p67phox-proteins cross-react with plant proteins of similar size [37] have been dismissed [38], and their absence in plants was finally concluded with the publication of the Arabidopsis genome [34]. In vitro assays showed that plant RBOHs have an NADPH-dependent O2– production activity that doesn’t require additional cytosolic subunit components and was stimulated in the presence of Ca2+ [39]. Genetic evidence in Arabidopsis showing that the two highly expressed RBOHD and RBOHF are required for the extracellular production of ROS triggered by microbial pathogens, finally provided the final conclusive evidence in favor of NOX activity-mediated oxidative burst in plants [40]. Over the years since, the multigenic RBOH family have expanded to around 130 members in nearly 30 plant species providing a growing body of evidence to the different roles of RBOH proteins in diverse biological processes [41– 46]. Nonetheless, the RBOH gene family is best

26

Plant NADPH Oxidases

characterized in the model plant Arabidopsis due to the availability of molecular genetic tools [47].

3

Structure of RBOH Proteins

All RBOH proteins present the same domain structure, with a core C-terminal region containing six α-helical transmembrane domains connected by five loops (loops A-E), and a C-terminal FNR (ferredoxin-NADP+ reductase) domain, containing NADPH-binding and the electron carrier flavin adenine dinucleotide (FAD)-binding motifs that are crucial for transferring electrons to generate superoxide anions (Fig. 26.1) [48, 49]. The third and fifth transmembrane domains contain two pairs of histidine residues, which are attachment sites for the two heme groups that that are responsible for two-step electron transfer from FAD to oxygen [10, 50]. The N-terminus is highly variable compared to the transmembrane and C-terminal domains [46]. It contains two Ca2+-biding EF-hand motifs and multiple regulatory target sites for Rac GTPase(s) [36, 51], calciumdependent protein kinases (CDPKs), mitogen activated protein (MAP) kinase(s), GIRAFFE heme oxygenase, phospholipase Dα1, and phosphatidic acid [48, 52]. Although RBOHs were cloned based on their homology to the phagocytic NOX2, they most resemble NOX5 that has an extended cytosolic N-terminal containing four EF-hand motifs, making its activation directly dependent on Ca2+ [10, 53]. In contrast, activation of the animal Nox1–4 members, which lack a regulatory C-terminal domain, requires forming a stable heterodimer with the membranal protein p22phox, forming the catalytic core of the NOX complex. The p22phox functions as a docking site for the regulatory cytosolic protein p47phox [49]. The small GTPase Rac binds to p67phox (or its homologous protein NOXA1), serving as a switch for NOX activation [49, 53]. None of the known regulators of mammalian NOXs have obvious homologs in plants, except for the small GTPase Rac protein [54]. The N-terminal region of RBOHs contains two EF-handlike motifs upstream to the two EF-hand-motifs [54, 55]. Crystal structure of the conserved N-terminal region of rice OsRBOHB revealed that the two EF-hand-like motifs together with the swapped EF-hands form a structure similar to that found in calcineurin B. The swapping of the EF-hands creates a coiled-coil region that was shown to be necessary for the interaction of OsRBOHB with OsRac1 [55]. In addition, it was demonstrated that an intramolecular interaction between the N-terminal and the C-terminal, similar to the intramolecular interaction in mammalian Nox5, result in its stimulation [55].

447

3.1

Regulation of Activation

ROS and calcium signaling are integrated in regulatory networks of multitude processes in plant biology via RBOHs activation. The N-terminal region of RBOHs function as a regulatory hub for integrating ROS and calcium homeostasis and signaling, with the binding of calcium to the EF-hands promoting ROS production, which subsequently activate calcium channels [55–57]. A heterologous expression system based on human embryonic HEK293T kidney cells lacking NOX2, which does not produce significant amounts of extracellular ROS, was used for quantifying the activity of RBOHs [58– 62]. Binding of Ca2+ to the EF-hand motifs induce conformational changes that activates the ROS producing activity [60]. The two EF-motifs were shown crucial for the Ca2+dependent activation [58, 59]. Point mutation in the conserved glutamate in the first EF-hand Ca2+ binding loop demonstrated that this motif has greater contribution to Ca2+ binding than the 2nd EF-hand motif and it is commonly essential for NOX activity, as the reduced Ca2+ affinity nearly abolished the ROS producing activity in all RBOHs [59, 60, 63–65]. The specific ROS-producing activity varied 100-fold among the different AtRBOHs in HEK293T cells with the highest activity produced by AtRBOHH, followed by AtRBOHJ, and -C, with AtRBOHA, -D, -E and -G producing around ten times less ROS, and even less by AtRBOHG, -F, -I and -B [64]. In addition to direct Ca2+ binding, the N-terminal of RBOHs is also target for Ca2+ network regulatory components. Calcineurin B-like (CBL) Ca2+ sensor proteins together with their interacting partners CBL-interacting protein kinases [CIPKs] function to govern Ca2+-signaling networks in plants [66]. CIPK26 and CIPK11 were shown to interact with the N-terminal of RBOHF and together with either CBL1 or CBL9 strongly enhance ROS production by RBOHF in HEK293T cells [61, 62]. While doing so, the combined activity of CIPKs with other kinases can sometimes fine-tune NOX activity, as in combination with the serine/threonine protein kinase open stomata 1 (OST1) triggering the complementary activation of RBOHF [61]. In the HEK293T cell system Ca2+-dependent activation of RBOHs is inhibited by the Ser/Thr protein kinase inhibitor K-252a, indicating that protein phosphorylation is a prerequisite for the Ca2+-dependent activation of RBOHs [59, 67]. Thus, phosphorylation was suggested as the initial trigger for Ca2+-dependent ROS signaling network [59]. It was shown that Ca2+ binding and phosphorylation act synergistically to activate RBOH-dependent ROS production, most likely by increasing the binding affinities of Ca2+ to the EF-hands resulting in a synergistic activation [58–60].

448

G. Miller and R. Mittler

Fig. 26.1 Structure of NADPH oxidases in plants and mammals. A conserved C-terminal core region consisting of six transmembrane a-helices (cylinders) and two heme groups (indicated by ‘H’ and ‘Fe’) is found in all RBOHs. (a) Plant RBOH; Two EF hand motifs are present in the N-terminal region. (b) Mammalian Nox1–Nox4; This Nox subfamily forms heterodimer with p22phox that contains 2 transmembrane a-helices and a proline rich region (PRR). (c) Mammalian

Nox5; Four EF hand motifs are present in the N-terminal region. (d) Mammalian Duox; Peroxidase domain in the N-terminal region is involved in H2O2 generation. This figure was reprinted with permission from Suzuki N, Miller G, Morales J, Shulaev V, Torres MA, Mittler R: Respiratory burst oxidases: the engines of ROS signaling. Curr Opin Plant Biol 2011, 14:691–699 [10]. Copyright Elsevier (2011)

Genetic evidence and in vivo assays indicated that phosphorylation is involved in self-amplification loops that contributes to the amplification of signals in defense and wound responses and stomatal closure [68–70]. Several serine residues at the N-terminal of RBOHs have been identified as phosphorylation target sites [54, 63, 71–

73]. Ser163 on AtRBOHD that is conserved in RBOHF (Ser174) and in other RBOH family members, was shown to be phosphorylated in response to the bacterial elicitor flagellin peptide 22 (flg22) [74]. In contrast, AtRBOHD Ser8, Ser39 and Ser152 that were found to be phosphorylated in response to flg22 are not conserved in AtRBOHF

26

Plant NADPH Oxidases

[71, 74]. The differences in environments of these serine residues suggested that AtRBOHD and AtRBOHF might be differentially regulated in response to divergent stimuli [72]. Other positive feedback amplification of RBOH activation involves calmodulin dependent activation of mitogenactivated protein kinase 8 [75], phospholipase Dα1-generated phosphatidic acid [76], or binding of small Rac GTPases to its N-terminal extension [36]. Some of these regulatory components, such as receptor-like cytoplasmic kinases and CDPKs, may regulate RBOH via direct interaction, by modulating other regulators, or both [73, 77–79]. ROS production by RBOHs may subsequently induce cytosolic Ca2+ elevation, which inhibits Rac binding, and thus terminate the oxidative burst [36]. RBOH activity can also be turned off post-translationally by nitric oxide (NO) via S-nitrosylation, the addition of an NO moiety to a reactive cysteine thiol to form an S-nitrosothiol (SNO) [80]. Activity of sweet pepper RBOHs inhibited in the presence of NO donors, peroxynitrite (ONOO-) and glutathione (GSH), suggested that RBOH proteins can undergo S-nitrosation, Tyr-nitration, and glutathionylation, respectively [81]. In addition, Ca2+-independent regulation may also play a critical role in specific response programs. For example, RBOHD can be activated via specific phosphorylation by BRI1-ASSOCIATED RECEPTOR KINASE 1 (BIK1) during pathogen immunity response in the absence of Ca2+ activation [54]. Thus, there is a large regulatory repertoire that mainly relies on Ca2+-dependent but also on Ca2+-independent mechanisms that can contribute to the tight control of ROS production by plant RBOHs to allow their signaling function [10].

4

Cellular Distribution of RBOH Proteins

In plants, RBOH proteins are mainly localized to the plasma membrane [63, 82]. Inclusion of RBOH proteins in lipid microdomains within the plasma membrane can enhance their functional specificity [1]. Proteomic studies revealed that RBOHs are present in detergent-insoluble fractions of the PM, suggesting that like their mammalian homologs they could be associated in vivo with sphingolipid- and sterolenriched rafts [83–85]. Such clustering of RBOHs in microdomains restrict ROS production during cell elongation to the growth points and is essential for polarized cell responses such as that in the apical membrane of elongating root hairs or the growing tip of pollen tubes, [63, 86]. Maintaining RBOH activity in the PM requires requirement of new proteins trough endosomes trafficking from the Golgi pool [63, 87]. However, NOX activity was also detected in intracellular vesicles in response to salt stress or during abscisic acid

449

(ABA)-induced stomatal closure [88, 89]. This could allow ROS trafficking between different compartments. Thus, specific subcellular localization could define the precise location of ROS accumulation and function and contribute to the specificity in function of plant NADPH oxidases [10]. Some mammalian NOXs, including NOX2, 4, and 5 are localized to other subcellular localization in addition to their localization at the plasma membrane [90–95]. NOX4 is active in mitochondria [92], and NOX5 has been identified in in various intra-cellular locations, including mitochondria, endoplasmic reticulum, and perinuclear region [90]. In plant cells, thus far there are no reports indicating intracellular localization of RBOH proteins or activation of oxidative burst within a specific organelle. However, additional studies are required before such possibilities for any of the ten RBOH proteins could be rule out.

5

RBOH Gene Expression

The spatio-temporal control of RBOH protein expression is important for their functions [49]. Yet, relatively little temporal transcriptional regulation was observed for RBOH genes, which implies that most of their regulation in response to different stimuli may be at the post-transcriptional level. Nevertheless, RBOHD is unique among the ten different RBOHs of Arabidopsis showing relatively high abundance in roots, shoots and reproductive tissues and a high degree of responsiveness to a wide range of stresses [10]. RBOHF is the second abundant RBOH that is expressed mostly in roots and shoots [10]. The expression of RBOHD and RBOHF is elevated by wounding, pathogens or different abiotic stress stimuli [40, 41, 89, 96–98]. The differential expression of RBOHs in different plant organs, tissues, and cell types, indicates a high degree of specialization (Fig. 26.2). However, co-expression of some RBOHs in a particular tissue may suggest some redundancy or other mode of cooperation in their activity (Fig. 26.3). RBOHD, which is expressed in most tissues, may function together with other RBOH proteins (e.g., RBOHA-C, G or H) to form a vertical continuum of oxidative burst potential (from the tip of roots to the top of the shoots and flowers) that function in systemic long-distance cell-to-cell communication (discussed below).

5.1 5.1.1

Biological Function Tissue Specific Function of RBOHs

Roots Roots constantly need to grow and develop to support the plant’s demand for water and nutrients. Plant roots

450

G. Miller and R. Mittler

Fig. 26.2 Multiplicity of functions of the plant RBOH-NADPH oxidases. ROS produced by RBOH proteins mediate multiple processes in plants. This figure was adapted with permission from Suzuki N,

Miller G, Morales J, Shulaev V, Torres MA, Mittler R: Respiratory burst oxidases: the engines of ROS signaling. Curr Opin Plant Biol 2011, 14:691–699 [10]. Copyright Elsevier (2011)

accumulate relatively high levels of ROS, especially at areas of high activity such as the root meristem, the zone with active cell divisions (meristematic zone), cell expansion (elongation zone) and cell differentiation (maturation zone; Fig. 26.4). ROS control cell expansion and cell differentiation processes such as root hair formation and lateral root development both as oxidants and as signalling factors that regulate transcriptional network. ROS also regulate crosstalk with phytohormone signals in controlling root development. Therefore, maintaining the root’s ROS homeostasis is important for balancing between cell proliferation and differentiation at the tip of the root [99].

The role of RBOHC, RBOHD, and RBOHF in root has been established and found to be consistent with their expression level in the root (Fig. 26.2) [10]. Arabidopsis rbohF mutant seedlings grown on nutrient media show reduced growth of the primary root, while the rbohD root length was comparable to the wild type. Addition of ABA inhibited the length of the rbohD root similar to the wild type, whereas the rbohF was insensitive to ABA, not showing further root growth decrease [56]. These results indicated that RBOHF has a specific role in mediating ABA signaling in primary roots of young plants. Root growth is directed downwards by gravitropism. In contrast, the

Fig. 26.3 Expression profile comparison of the RBOH family members in different organs and tissues in Arabidopsis. The plot was generated using CoNekT, the web platform for visualization and analysis of plant co-expression and co-function networks [201]. Gene expression is scaled to range from 0 to 1

26 Plant NADPH Oxidases 451

452

Fig. 26.4 The rhd2-6/rbohc mutant has impaired root hair elongation compared to wild-type (Col-0). (a) An Arabidopsis root stained with propidium iodide shows cell shapes found in the meristematic, transition, and elongation zones. Scale bar = 100 μm. (b) The tips of young Arabidopsis roots are compared between wild-type (Col-0) and the root hair defective (rhd2) mutant, which has a defect in the RBOHC gene. The area above the dashed lines indicates the maturation zone, where root hairs form. Scale bar = 200 μm. This figure was adapted with permission from Chapman JM, Muhlemann JK, Gayomba SR, Muday GK: RBOH-dependent ROS synthesis and ROS scavenging by plant specialized metabolites to modulate plant development and stress responses. Chem Res Toxicol 2019, 32:370–396 [47]. Copyright (2019) American Chemical Society

hydrotropic response of the root directs root growth away from low water potential (towards increased water availability) and attenuates the gravitropic growth. Growth towards gravitropic stimulus, induced by rotating the plates of vertically grown seedlings, is accompanied by transient asymmetric accumulation of ROS (higher at the epidermis inner to the curvature) at the elongation zone of the root. Hydrotropically responding roots show no transient asymmetric ROS distribution. Depletion of H2O2 in the root apex by antioxidants or DPI treatment increased the hydrotropic bending of the root. Similar response occurred in rbohC but not the rbohD mutant, indicating the hydrotropic response of the root is tuned by RBOHC [100]. RBOHF also plays an additional specific role important for the function of roots. Uptake of water by the root occurs via root epidermal hairs inwards through the cortex consisting mostly of parenchyma cells to the inner endodermis cell layer surrounding the vascular bundle of xylem and phloem. The endodermis contains a thickened ring-like cell-wall modifications composed of lignin and suberin impervious to water called the ‘Casparian strips’. that function as a barrier that prevent water

G. Miller and R. Mittler

from moving freely through the apoplast (i.e., the space outside the plasma membrane, between cells, within which material can diffuse freely). Casparian strips control the transportation of water and inorganic salts between the cortex and the vascular bundle, thereby aiding selective movement of solutes inwards into the vasculature trough plasma membrane transporters [101, 102]. Formation of the Casparian strips requires deposition of lignin polymerized from monolignol monomers in the apoplast of endodermal cells by coupled action of NADPH oxidase, SOD and peroxidases [103]. Building of the Casparian strips as a median ring surrounding the plasma membrane is mediated in Arabidopsis by RBOHF, which is brought into proximity of localized peroxidases by Casparian strip domain proteins (CASPs) [104]. Absence of RBOHF therefore leads to a very strong delay in the formation of the Casparian strips and impairment of the apoplastic barrier. This role is unique to RBOHF, since no other mutant of root expressed NADPH oxidase, including RBOHB, which is strongly expressed in endodermal tissue, display any delay in Casparian strips formation [104]. Lateral Roots (LRs) Lateral root primordia are initiated in the pericycle, the cell layer surrounding the vascular bundle, adjacent to the endodermis. Thus, the lateral root must emerge through several cell layers, including the epidermis, to emerge out of the primary root [47]. Treatment of Arabidopsis roots with exogenous ROS promote the emergence of LRs, whereas DPI delayed the development of LR [105]. Moreover, peroxidase activity and ROS signaling are specifically required during lateral root emergence but not for generation of the lateral root primordia [106]. Interestingly, double rbohD/F mutant had remarkably higher lateral root density, more than the wild-type or the single mutants. Surprisingly, the absence of RBOHD and RBOHF promote an increase of O2•– and a decrease in H2O2 accumulation at the mature area of the primary root that was caused by increased peroxidase activity [107]. It was concluded that RBOHD and RBOHF negatively modulate lateral root development by changing peroxidase activity and increasing local production of O2•– in the primary root. Differentiation of LR primordia (LRP) by division of specific pericycle founder cells is controlled by the phytohormone auxin [106]. However, the control of local generation of O2•– and peroxidase activity by RBOHD and RBOHF in the LRP promoting the emergence of LR function is independent of auxin [106, 107]. RBOHA-F promoter-GUS reporter histology showed these RBOHs to have variable expression and distribution within the LRP, suggesting some specificity in LR formation [105]. Next to RBOHD, RBOHE exhibited the most

26

Plant NADPH Oxidases

widespread expression in the LRP, with strong expression in the endodermis, cortex, the overlying epidermis cells, and in the basal meristem (also known as Transition zone— Fig. 26.4a) where LR priming occurs. Correspondingly, higher-order mutants lacking RBOHE and RBOHD showed delay in the rate of LR emergence. Root Hair Elongation Root hairs are long tubular-shaped outgrowths from root epidermal cells that can grow up to 1 mm or more vastly increasing the root surface area and effectively increasing the root diameter. Root hairs are generally thought to aid plants in nutrient acquisition, anchorage, and microbe interactions [108]. The role of RBOHC in root hair development was first identified in a genetic screen in the root hair defective 2 (rhd2) mutant that has short root hairs (Fig. 26.4) [57, 109]. ROS accumulation at the growing tip of root hairs is nearly abolished in the rbohC mutants [57]. Thus, RBOHC activity is essential for polarized ROS accumulation in root epidermis cells that facilitates cell expansion. RBOHC dependent ROS, particularly hydroxyl radicals OH•– that develops downstream of O2•– production via the “Fenton reaction” is important for cell expansion. OH•– induce cell wall loosening attacking cell-wall matrix polymers that allow the cell to stretch [110]. In addition, OH•– activates hyperpolarization-activated Ca2+ channels facilitating Ca2+ influx from extracellular stores producing the tip-focused Ca2+ gradient required for the process of cell elongation [57]. The increase in cytosolic Ca2+, in turn, activates RBOHC NOX activity creating a positive feedback loop that maintain the polarity and the directional growth of the root hair cell [63]. Auxin is known to promote root hair elongation [111, 112]. Auxin-induced growth of root hairs is abolished in the rbohC mutant [113]. Root hair initiation is controlled by ROOT HAIR DEFECTIVE 6 (RHD6) transcription factor through auxin and ethylene associated processes [114]. The Auxin-induced elongation phase of the root hair is mediated by activating the RHD6 LIKE 4 (RSL4) that controls the final root hair size [115, 116]. RSL4 directly regulate the transcription of RBOHC and several peroxidases that together regulates apoplastic ROS homeostasis in the root hair cell [117]. Similarly, RSL4 also regulates the expression of RBOHJ, which together with RBOHH implicated in polar growth of pollen tubes [65, 117]. Interestingly, double rbohH/J mutants display moderate to low ROS level in root hairs and up to 40% reduction in root hair growth. Therefore, although RBOHC is the main RBOH involved in root hair growth, RBOHJ and RBOH are also important ROS producing enzymes in this process [117].

453

Pollen Fertility ROS production by RBOHs play important role in plant reproduction. Antisense strategy in tomato (Lycopersicon esculentum) targeted to the C-terminal regions of whitefly-induced (Wfi1) and RBOH1 (homologues of RBOHD and RBOHF, respectively), resulted in developmental deformations in leaves and flowers [41]. These observations suggested that RBOHs play important role in different tissues and developmental stages of reproduction with redundancy, as the antisense may have targeted multiple RBOHs, and no single or double mutants in Arabidopsis have demonstrated such severe pleiotropic phenotype. In addition, the deficiency in RBOH activity also resulted in male sterility that led to the development of parthenocarpic tomato fruits (i.e., seedless fruits) [41], which indicated the importance of RBOH activity to reproductive development and function. Pollen development begins in the anther locules when a diploid microspore mother cell undergoes meiosis to produce a tetrad microspore. Each microspore divides mitotically once or twice (depending on the species). This is followed by a maturation stage when the pollen desiccates in preparation for dehiscing of the pollen sac and release of the mature pollen [118]. The tapetum is a layer of cells in the anther that surround and nourish the developing pollen. The tapetum cells undergo programmed cell death, culminating at stage 10 and 11 in Arabidopsis coinciding with mitotic divisions and the maturation of the developing pollen grain [119]. Tapetal degradation releases lipids, flavonoids (specialized plant metabolites with antioxidant activity) and proteins that are utilized by developing pollen and necessary for the formation of the pollen coat [47]. AtRBOHE is specifically expressed in tapetum cells from stage 6 until tapetum degeneration is visible at stage 10 [120]. Loss of function in AtRBOHE resulted in irregular pollen coat (exine) pattern and deformed pollen shape causing partial sterility. AtRBOHC also express in the tapetum but at much lower abundance than RBOHE. Consequently, rbohC/ E double knockout showed severe phenotype of pollen development, with a delay in tapetum degradation that was greater than in the rbohE single mutant, revealing the redundant role of RBOHC in this process. In contrast, over expressing RBOHE caused early degradation tapetal cells and tapetal dysfunction [120]. This study showed the essential role of RBOHE-induced ROS in regulating the timing of programmed cell death of tapetum cells and pollen development. Reproduction in angiosperms occurs in flowers when the male gametophyte, the pollen, is carried onto the stigma (top part of the pistil) during pollination. The compatible pollen germinates and grow a pollen tube that penetrates the stigma,

454

elongates in the stylar transmitting tissue within the pistil and is guided to the ovule, entering the embryo sac to achieve double fertilization [118]. The rupture of the pollen tube is induced by the ovule upon entering to the ovule. Growing of the pollen tube and interaction with the female gametophyte requires constant ROS production [65, 120–122]. The two sperm cells released from the pollen tube fertilize the egg cell and the central cell to form the embryo and the endosperm, respectively [118]. Similar to root hair polarized cell growth, tip-localized formation of ROS by NOX activity is needed to sustain normal rate of pollen tube growth [123, 124]. Yet, pollen tubes grow 10-times faster than root hairs [125] and produce higher levels of ROS in the process by NOX activity, as it takes higher doses of DPI and Mn(III)tetrakis(1-methyl-4pyridyl)porphyrin, Tetratosylate, Hydroxide (Mn-TMPyP; a cell permeable SOD mimetic agent) to inhibit pollen tube growth compared to root hair growth. Treatment with H2O2 can stimulate pollen germination in several plant species, and rescue the phenotype of pollen tube growth in Arabidopsis antisense lines suppressed in RBOH expression [124]. In Arabidopsis, RBOHH and RBOHJ are preferentially expressed in pollen [10, 126]. Using the H2O2-sensitive HyPer and the Ca2+-sensitive YC3.60 sensors in NADPH oxidase-deficient mutants, it was shown that NADPH oxidases generate pollen tip-localized, pulsating H2O2 production that functions, possibly through Ca2+ channel activation, to maintain a steady tip-focused Ca2+ gradient during growth. Deficiency in RBOHH or RBOHJ did not inhibit pollen germination in vitro and seed set formation was comparable to the wild type [123]. While in rbohJ pollen tubes raptured at low frequency as in the wild type, rbohH mutants pollen tubes raptured at a high rate (57%), which increase to 80% in the rbohH/J double mutant due to impairment in tip-localized ROS production and Ca2+ homeostasis in the pollen tubes [65, 123]. Direct phosphorylation and Ca2+ binding were shown to synergistically increase the ROS producing activity of RBOHH and RBOHJ in HEK293T cells, indicating that Ca2+-induced ROS has a positive feedback regulation role in polarized cell growth, similar to root hair tip growth [65]. Interestingly, double mutants for calcium channel genes CNGC7 and CNGC8 (cyclic nucleotide gated cation channel; non-selective, Ca2+-permeable ion channels) have phenotypes almost identical to rbohH/J, suggesting that these channels played a key role in the feedback regulation of RBOHH and RBOHJ activities [127]. Pollen tube grows rapidly in a coordinated manner with growth rate that oscillates rhythmically with periodicity of a few minutes [128]. Growth dynamics in the rbohH/J double mutant is different from the wild type. Prior to tube rapture, pollen tubes of the rbohH/J mutant exhibit high frequency

G. Miller and R. Mittler

growth oscillations and strong fluctuation in amplitudes causing elevated growth rates, which suggested that NOX activity slows-down pollen tube growth to coordinate the rate of cell expansion with the rate of exocytosis of cell-wall material [121]. As a result, the rbohH/J double mutant suffers from severe male-sterility leading to dramatic reduction in seed set formation [123]. Seeds Germination Germination of seeds begins with release from dormancy, which is controlled by ABA and the activation of gibberellic acid (GA) that controls germination-promoting signals [129, 130]. Studies in Arabidopsis and barley seeds have shown that H2O2 mediates the regulation of ABA catabolism, antagonizes ABA signaling, promotes GA synthesis and the seed germination program [131–134]. Dormancy release also occurs during dry storage of seeds at room temperature, a process referred to as seeds ‘after ripening’. ROS accumulation during after-ripening plays a key function in protein oxidation, which promotes changes in the proteome and prepare the embryo for a pro-germination program [135, 136]. Seed’s after-ripening dormancy release mechanisms involve selective oxidation of proteins and mRNA facilitated by gradual accumulation of ROS in the embryo leading to their irreversible oxidation, primarily via carbonylation of Lys, Arg, Pro, and Thr, or 8-hydroxylation of guanine (8-OHG), respectively. Degradation of proteins and RNA enables these alternations in the proteome and transcriptome and the recycling of amino and nucleic acids [137–141]. Although, during dry storage in seeds, ROS are mainly produced via non-enzymatic autoxidative reactions occurring spontaneously [138, 140], the potential contribution of NOX activity to this process was also demonstrated [142]. In Arabidopsis, RBOHB transcripts significantly accumulate in embryos and whole seeds when dry seeds absorb water (i.e., imbibed seeds) [47]. RBOHB is post-transcriptionally regulated via splicing during seed after ripening and germination. There are two RBOHB splice variants; a smaller variant that is a fully spliced RBOHBα, and larger RBOHB-β splice variant retaining intron 1 that leads to premature termination and a nonfunctional protein. Expression of the two variants is developmentally and hormonally regulated in seeds, with after-ripened seeds containing the functional RBOHBα [142]. While freshly collected seeds accumulated only RBOHBβ, after a ripening period or imbibition initiated there is an increase in the abundance of the RBOHB-α transcripts and NOX-induced ROS accumulation in the embryo. Treatment with ABA blocked the accumulation of RBOHB-α transcripts

26

Plant NADPH Oxidases

during seed imbibition, suggesting the involvement of ABA in regulating of RBOHB activation in seeds. Fresh collected seeds of rbohB mutant showed relative insensitivity to inhibition of germination by ABA. In contrast, after ripened seeds of the mutant demonstrated increased sensitivity to ABA [142]. These results highlighted the role of RBOHB in seeds germination dynamics by slowing germination in fresh seeds and promoting dormancy release during after ripening.

6

Involvement in Stress Responses

Due to their sessile lifestyle, plants are highly attentive to changes in their environment, perhaps more so than other organisms. Accumulation of ROS is inherent to any biotic or abiotic stress response in plants, acting as essential signal messengers during different stages of the response, from perception of the stress to maintaining tolerance or acclimation to a particular threat. NOX-mediated ROS production is the main driver of ROS signaling required for biotic stress resistance and acclimation to abiotic stresses [10]. RBOHD and RBOHF are the main RBOHs playing a specific role in multiple responses to diverse stresses, including defense against pathogens attacks [40, 143, 144], and acclimation to salinity [145], hypoxia [146, 147], Ozone [148], and iron deficiency [149]. In fact, RBOHD is considered to play an important role in plant–pathogen interactions while RBOHF is thought to be more important in ABA signaling [72], although these two NADPH oxidases have functional redundancy under different environment conditions [40, 56]. For example, waterlogging-induced autophagy was greatly inhibited in Arabidopsis rbohD and rbohF roots, suggesting the involvement of both genes in hypoxia-triggered cell death [147] Other RBOHs may also play a role in plant responses to additional stresses (e.g., RBOHI in Arabidopsis seeds and roots under drought stress [150]). Alternatively, other RBOHs with low abundance or distinct tissue specificity may cooperate or contribute to redundantly in responses orchestrated by RBOHD and RBOHF, such as RBOHH in rice and maize that are involved in inducing aerenchyma cells (modification of the parenchyma to form a spongy tissue that creates spaces or air channels, which allows exchange of gases between the shoot and the root) formation under anoxia conditions caused by waterlogging [151, 152].

7

Role in Plant Immunity

Production of ROS in plant cells is a hallmark of successful recognition of plant pathogens and activation of plant defenses [153, 154].

455

NOX activity is important for the function of immune systems and the balance of the microbial community in both animal cells and plants [155, 156]. The apoplastic ROS burst generated in elicited plant cells are thought to sufficiently kill some invading pathogens, but also act as signaling molecules, triggering plant immune responses, and cell death [157]. In humans, genetic defects in the CYBB gene encoding NOX2 in phagocytes cause a primary immunodeficiency known as chronic granulomatous disease (CGD), a condition in which phagocytosis is normal but the respiratory burst is absent, exposing patients to persistent multiple infections by microbes [156]. Disease resistance in plants is sometimes associated with the HR death, the most obvious manifestation of the defense response occurring at the point of pathogen penetration [11]. The HR cell death requires the hormone salicylic acid (SA), which is an activator of the plant immune system [11, 158]. Production of ROS has long been recognized to orchestrate the HR and has been intimately associated with the early generation of the oxidative burst [11]. In the absence of RBOHD, opportunistic pathogens that colonize wild-type plants asymptomatically cause disease in Arabidopsis [155]. Although rbohD exhibited substantial reduction in the oxidative burst these mutants showed only modest decrease in cell death when challenged with avirulent bacteria. In contrast, enhanced cell death was observed in the rbohD and rbohD/F mutants after infection with parasitic oomycetes (downy mildew) [40]. Although the contribution of RBOHD to ROS production in leaves was greater than that of RBOHF, rbohF mutants display a stronger effect on cell death [40]. These results indicated that RBOHD and RBOHF share functions in ROS production in response to pathogens and although not required for triggering cell death, they have separate functions in regulating HR. A follow up study discovered that RBOHD and RBOHF suppress the spread of cell death beyond the local infected region in leaves, thereby taming the HR to prevent the plant from cell death getting out of control. Both RBOHs function together with lesion simulating disease resistance 1 (LSD1), a zinc-finger protein that negatively regulates the spread of cell death to the uninfected cells surrounding the HR site. Double mutants of rbohD or rbohF with lsd1 showed uncontrolled cell death after HR induction with SA. In the absence of all three proteins, the triple rbohD/F/lsd1 mutant showed uncontrolled death at young age under growth conditions that normally suppress lsd1 runaway cell death (i.e., HR cell death that spreads beyond the infected region to engulf the entire leaf) [96]. Unlike vertebrates, plants lack mobile immune cells and an adaptive immune system and mainly rely on two interconnected layers of the innate immune system to perceive and respond to pathogen infections [159]. The

456

perception of pathogen associated molecular patterns (PAMPs) by pattern recognition receptors (PRRs) at the apoplast serve as the first layer of the recognition response in plant immunity. This recognition leads to a strong, rapid, and specific production of ROS that is dependent on RBOH activity and PAMP-triggered immunity [54]. The involvement of RBOHD and its regulation in plant immunity mechanisms was characterized, mainly in Arabidopsis, but was also shown to play an important role in ROS production in other plant species in response to PAMPs (e.g., flg22) and pathogens [36, 44, 45, 73]. The second layer known as effector-triggered immunity (ETI) generates a stronger immune signaling involving predominantly nucleotide-binding site leucin-rich repeat (NBS-LRR) receptor proteins, which directly or indirectly recognize virulence effectors secreted to the host by the microbial cell [160–162]. These avirulence (Avr) effectors (e.g., AvrPto, AvrPtoB) have been evolved initially by adapted pathogens as virulence agents to reprogram the host’s physiology toward infectious compatibility. The perception of these effectors by NBS-LRR proteins, which are structurally related to mammalian NOD-like receptors (NLRs), is the result of a constant arms race between plants and their pathogens [163]. Plant PRRs (involved in PAMP-triggered immunity) are either receptor kinases (RKs) or receptor-like proteins (RLPs) containing various ligand-binding ectodomains, a single transmembrane domain, and an extracellular kinase domain that perceive PAMPs such as bacterial flagellin (flg) and elongation factor-Tu (EF-Tu) [163]. RKs and RLPs sense microbe- or selfderived molecular patterns to regulate pattern-triggered immunity (PTI), a robust form of antimicrobial immunity [164]. PRRs, such as FLAGELLIN SENSING 2 (FLS2) or EF-Tu receptor (EFR) bind immunogenic peptides of flagellin (e.g., flg22) or EF-Tu (e.g., elf18), respectively, thereby inducing their instant association with the co-receptor LRR-RK BAK1 (BRI1-ASSOCIATED RECEPTOR KINASE) [163]. The activated FLS2/EFR-BAK1 PRR complex rapidly phosphorylates the associated receptor-like cytoplasmic kinases (RLCKs) such as BIK1 or PBL1, releasing them to activate oxidative burst and MAPK cascades triggering immune responses [54, 157]. Phosphorylated BIK1 has a higher binding affinity for RBOHD and phosphorylates it on some specific sites, inducing a rapid and strong oxidative burst in a Ca2+-independent manner. Although BIK1 can activate RBOHD in the absence of Ca2+, it is not sufficient for inducing ROS production upon PAMP recognition. Ca2+ is also required for RBOHD activation during immunity responses [59, 60, 165, 166]. However, the involvement of Ca2+ binding and CDPKs-mediated activation of RBOHD requires further elucidation [54]. Integration between the two modes of RBOHD induction was suggested to occur in a two-step activation, in which the initial phosphorylation by BIK1 may induce conformation changes in RBOHD that lead to increased Ca2+ binding

G. Miller and R. Mittler

affinity by the EF-hand motifs and/or increased accessibility for CDPK-mediated phosphorylation. At the same time the activated BIK1 may also activate Ca2+ channels(s) that induce Ca2+ influx, increasing Ca2+ binding and CDPKs activity, reinforcing the direct activation by BIK1 phosphorylation [85]. To date the identity of the calcium channel (s) involved in PTI remains unclear [164]. Recognition of pathogen Avr effectors via direct or indirect interaction with NLR proteins, leads to a strong NOX-dependent ROS burst and the HR cell death response, the two key components of ETI [157]. Activation of the oxidative burst may be mediated by Rac GTPase [84] and other yet unknown regulators. Sustaining NOX activity during immunity responses requires constant supply of NADPH by NADP-malic enzyme (ME) that reduces NADP+ to NADPH [167]. Some pathogens have developed effectors that interact with proteins regulating NO activity, including NADP-ME, thereby circumventing the plant immune system [112].

8

Role in Stomata Closing

Stomata are pores surrounded by pairs of guard cells found on leaves and other aerial organs that enable gas exchange between the plant and the atmosphere, including influx of CO2 for photosynthetic carbon fixation and water loss via transpiration. In guard cells, a signal transduction network integrates the status of water light, CO2, hormone responses, and other environmental conditions to regulate stomatal apertures for optimal growth and survival of the plant under diverse conditions [168]. Stomata opening is achieved by activating inward K+ channels, increase of solutes in the cytosol and vacuole of the guard cells followed by water uptake causing turgor pressure to increase, resulting in curving the inner thick cell walls guarding the pore. Stomatal closure is mediated by turgor reduction caused by efflux of K+ out of the cells in response to drought stress, elevated CO2, and low humidity [168]. Water deficit and high salinity stress induce the accumulation of the hormone ABA that promotes the closing of stomata, which leads to reduction in transpiration rate and water loss [65]. The increase in ABA is perceived in cells by the PYR/PYL/RCAR (Pyrabactin resistant protein/PYR-like proteins/Regulatory Components of ABA-receptor) [169, 170] that binds to ABA, which facilitates interaction with group A protein phosphates type C (PP2C) family members. Under conditions of low ABA levels, PP2Cs such ABA-insensitive 1 (ABI1) and ABI2, function as negative regulators of ABA signalling. The binding of ABA by PYR/PYL/RCAR receptors triggers sequestration of the PP2C component that liberates class III ABA-activated Snf1-related kinases (SnRK2s) to phosphorylate various targets downstream [171]. ABA stimulates the activity of hyperpolarizationactivated Ca2+ channels in the plasma membrane of guard

26

Plant NADPH Oxidases

457

Fig. 26.5 ABA-induced ROS burst in guard cells mediates stomatal closure. (a) ABA-induced ROS burst precedes stomatal closure in wildtype tomato. Increases in DCF fluorescence were visualized in guard cells across a 45 min time course of ABA treatment in wild-type tomato leaves. (b) The Interaction of ROS, ABA, and Ca2+ in Stomatal Responses. ABA can also be synthesized in guard cells in response to signals such as elevated ROS. Accumulation of ABA could also result from decreased degradation or release of ABA from conjugated sources. ABA can cause the accumulation of ROS via interaction with the PYR/PYL receptor and the inhibition of PP2C that will result in activation of OST1 and phosphorylation and activation of NADPH oxidase (RBOH). ROS generated by RBOH or arriving at the guard cell with the ROS wave can enter guard cells from the apoplast via aquaporins and

result in the opening of ROS-regulated Ca2+ channels that in turn will cause the further activation of RBOH proteins by CIPK26. ROS can also directly inhibit PP2C, preventing the inhibition of OST1 and resulting in further biosynthesis of ROS, or accumulate due to enhanced photorespiration. (a) was adapted with permission from Chapman JM, Muhlemann JK, Gayomba SR, Muday GK: RBOH-dependent ROS synthesis and ROS scavenging by plant specialized metabolites to modulate plant development and stress responses. Chem Res Toxicol 2019, 32:370–396 [47]. Copyright (2019) American Chemical Society. (b) This figure was reprinted with permission from Mittler R, Blumwald E: The roles of ROS and ABA in systemic acquired acclimation. Plant Cell 2015, 27:64–70 [175]. Copyright (2015) Oxford University Press

cells [172], increasing cytosolic Ca2+ [173]. This Ca2+ influx causes an efflux of potassium and anions through K+ channels and slow anion channel 1 (SLAC1), respectively,

followed by water exiting from the guard cells, decrease in turgor pressure, cell shrinkage and closing of the pore (Fig. 26.5) [174, 175].

458

In addition to Ca2+, ROS, and nitric oxide are also important second messengers in ABA-induced stomatal closing [47, 172, 176]. Accumulation of apoplastic ROS was shown to be involved in the induction of stomatal closing [177, 178]. ABA-induced ROS accumulation in guard cells and stomatal closing was suppressed by DPI [179], denoting the importance of RBOH activity in the process. ABA activates RBOHD and RBOHF in Arabidopsis guard cells, which redundantly function in ABA signalling in these cells and their deficiency impairs stomatal closure. The single rbohF mutant showed partial reduction in ABA-triggered stomatal closing while rbohD was comparable to the wild-type [56]. Treatment of the rbohD/F double mutant with exogenous H2O2 rescued Ca2+ channel activation and stomatal closing [56]. OPEN STOMATA 1 (OST1; SnRK2.6), an ABA-activated protein kinase homologous to SnRK2.2/ SnRK2.3, is a positive regulator of stomatal closure (Fig. 26.5) [175, 180]. In the Arabidopsis ost1 knockout mutant the ABA-induced ROS generation in guard cells is abolished [181]. OST1 phosphorylates AtRBOHF at Ser-13 and the conserved Ser-174 residues at the N-terminal of RBOHF but not AtRBOHD [72], corresponding with the impaired stomata closing in the rbohF mutant [56]. This also suggested that the respective conserved serine residue in RBOHD (Ser168) that has Arg at the -3 position may be targeted by other SnRK2 family members in guard cells [72]. While the leaf water status of the snrk2.2/3 double mutant was comparable to the wild-type, the leaves of the triple snrk2.2/3/6 (ost1) mutant showed rapid water loss far exceeding that in the single ost1 mutant [182]. The additive phenotype of the triple snrk2 mutant is reminiscent of that in the rbohD/F double mutant, which support the idea that RBOHD may be a target for SnRK2.2 and SnRK2.3. In addition to osmotic stresses, other stress stimuli such as high light intensity, which cause photooxidative stress, induce ABA-dependent stomatal closing [183]. Yet, although the stomatal closing in response to ABA treatment in rbohD mutants was comparable to the wild-type [56], the mutant is completely insensitive to light stress-induced stomal closing stimulus, suggesting the involvement of other secondary messengers in this process. An interplay between jasmonic acid (JA), SA, ABA, and ROS were proposed to play a role in this response [183]. Genetic evidence using Arabidopsis mutants showed that the activation light-induced guard cell closing mediated by ABA and ROS depended on the receptor kinase-like protein hydrogen peroxide resistant 1 (GHR1) and SLAC1 [183]. Interestingly, the bacterial elicitor flg22 has been shown to induce stomatal closure in WT plants but not in the ost1 mutant [184] indicating that OST1 might integrate both ABA and pathogen signaling to induce stomatal closure through the activation of NADPH oxidases [72].

G. Miller and R. Mittler

9

Role in Cell-to-Cell Signaling and Systemic Responses

In the absence of a circulatory or nerve systems, plants evolved different mechanisms that employ numerous mobile signals generated in response to different stimuli affecting a particular tissue (termed local tissue). These signals then spread in a cell-to-cell fashion throughout the plant to reach all other tissues (termed systemic tissues). Once reaching their target tissues, the different signals trigger in them defense of acclimation mechanisms and induce a form of systemic acquired acclimation or defense that makes the plant more resilient to changes in environmental conditions. Cell-to-cell communication driving this processes can occur via three main routs: (1) symplastic, across cells via plasmodesmata within the same or adjacent tissues; (2) vascular, via the phloem or xylem systems (using cell-to-cell communication); and (3) apoplastic, via the extracellular space connecting cells [185]. Two of the best studied types of systemic responses are; (1) Systemic acquired resistance (SAR), a long-distance SA-dependent signaling mechanism that provides broad spectrum and long-lasting resistance to secondary infection by pathogens (e.g., viruses, bacteria, fungi) , and (2) systemic acquired acclimation (SAA) that prevents further damage by abiotic stresses (e.g., high light intensity, UV, heat, cold, salinity) [185]. Generation of ROS plays a key role in both SAR and SAA responses [52, 186, 187]. The discovery that NOX activity plays an important role in cell-to-cell communication and extracellular signaling occurred around the same time in mammalian systems, zebrafish, and Arabidopsis [97, 188, 189]. This included the involvement of NOX2 in chemotaxis, generating ROS to ensure proper guidance and recruitment of immune cells to the site of infection in humans and mice [188], generation of extracellular tissue-scale H2O2 gradient by dual-oxidase (Doux) to attract leukocytes to a wound site in Zebrafish [189], and rapid activation of gene expression in systemic tissues in response to diverse stress stimuli mediated by RBOHD in Arabidopsis [97]. ROS intercellular signals travel from cell-to-cell through diffusion [189], by exosomes, or other nano-vesicles containing high levels of H2O2 [190], or via oxidative burst proliferation at the apoplast [1]. This type of systemic ROS signaling was later demonstrated in other plants and even in mushrooms [187] and termed the ‘ROS wave’. The RBOHD-mediated ROS-wave is highly important for the ability of plants to rapidly induce an SAA program that is specific to the type of stress stimulus [191]. For example, ROS wave triggered in a single rosette leaf exposed to high light (HL) intensity that propagates long distance through the inflorescence stem of Arabidopsis triggered within minutes a distinct transcriptional program and

26

Plant NADPH Oxidases

Fig. 26.6 Amplification of ROS production by RBOH and the ROS wave. The process of RIRR that is used for cell-to-cell communication mediates a plant-wide systemic signaling pathway. In this pathway, a stress or a stimulus that is sensed by cells at one leaf triggers an autocatalytic cell-to-cell RIRR process that spreads throughout the entire plant and triggers acclimation and/or defense mechanisms. Local application of the hydrogen peroxide scavenger enzyme catalase (CATALASE), the calcium channel blocker lanthanum chloride (LaCl), or the RBOH inhibitor diphenyleneiodonium (DPI), as well as grafting experiments with rbohD mutants, can completely block the spread of the RIRR signal. This figure was adapted with permission from Suzuki N, Miller G, Morales J, Shulaev V, Torres MA, Mittler R: Respiratory burst oxidases: the engines of ROS signaling. Curr Opin Plant Biol 2011, 14:691–699 [10]. Copyright Elsevier (2011)

acclimation response in cauline leaf, protecting it from a secondary direct exposure to HL but not against wounding or heat stress [191]. The systemic long distance ROS signals travels rapidly, up to several centimeters per minutes, as wave of ROS that is dependent on RBOHD activity [1, 97]. Abrogation of the systemic signal by RBOHD deficiency, DPI, or catalase treatment suggested that the ROS wave develops by an autopropagation loop in which ROS generated by RBOHD activation are capable of triggering all adjacent cells to undergo the same process of RBOHD activation along the path of the signal [185, 191]. Thus, the autopropagation of the systemic ROS wave depends on ROS-induced ROS release (RIRR) that is mostly, if not entirely, dependent on RBOH activity at the PM (Fig. 26.6) [192]. In animal cells, RIRR is an intracellular process in which ROS released from one cellular compartment or organelle (e.g., mitochondria) triggers the enhanced production and/or release of ROS by another compartment [192–194]. Mitochondrial NOX4, and perhaps other NOXs having various

459

subcellular localizations (NOX2 and NOX5) are thought to play a role in RIRR in animal cardiovascular systems [192– 195]. In contrast, the nature of RIRR in plants seems to be exclusively dependent on extracellular ROS production by RBOHs. The initial activation of RBOHs is thought to trigger a positive amplification loop sustaining a high RBOH state of activity, thus maintaining higher levels of cellular ROS production for a long duration [192]. This RIRR process that mediates the ROS wave requires the activity of calcium channels followed by direct binding of Ca2+ and calcium activated kinases (e.g., CIPKs, and SnRK2) [78, 196, 197]. Moreover, because ROS can inhibit phosphatases activity, the RBOH-activation state may persist via ROS-triggered inhibition of different phosphatases [192]. The ROS wave is, therefore, dependent on, and interwoven, with a systemic calcium-induced calcium release (CICR) also termed the calcium wave [192, 196]. The calcium wave depends on the tonoplast localized ion channel two pore channel 1 (TCP1) [197, 198], and glutamate receptor-like channels GLR3.3 and GLR3.6 [199, 200]. The transmission speed of the calcium wave was significantly reduced by treatments with DPI or ascorbic acid, and in the rbohD mutant [198]. Thus, the systemic calcium wave is supported by RBOHD mediated RIRR.

10

Conclusions

ROS generated by RBOH proteins serve as signals regulating large number of processes in plants. They are unique among other ROS producing mechanisms in plants as they integrate ROS production with different signal transduction pathways such as calcium, NO, and lipid signaling. The ten members of the RBOH family have different spatial expression and play distinct roles in different programs with some degree of redundancy or cooperation between two or more RBOHs. Differential posttranslational modifications (e.g., phosphorylation) as well as different regulatory proteins acting on the variable N-terminal region of RBOHs distinctly regulate the activities of coexpressing RBOH proteins. ROS generated by RBOHs, mainly RBOHD, function as rapid systemic long-distance signals (i.e., The ROS wave) by autopropagation of RBOH activity, mainly RBOHD, via RIRR. The ROS wave is interwoven with the Ca2+ wave and electric signals in coordinating systemic responses in plants. Acknowledgements We thank Prof. Gloria Muday from Wake Forest University for providing us with original high-resolution images for reusing in some of the figures.

460

References 1. Mittler R, Vanderauwera S, Suzuki N, Miller G, Tognetti VB, Vandepoele K, Gollery M, Shulaev V, Van Breusegem F (2011) ROS signaling: the new wave? Trends Plant Sci 16:300–309 2. Mittler R (2017) ROS are good. Trends Plant Sci 22:11–19 3. Kmiecik P, Leonardelli M, Teige M (2016) Novel connections in plant organellar signalling link different stress responses and signalling pathways. J Exp Bot 67:3793–3807 4. Mullineaux PM, Baker NR (2010) Oxidative stress: antagonistic signaling for acclimation or cell death? Plant Physiol 154:521–525 5. Mittler R, Vanderauwera S, Gollery M, Van Breusegem F (2004) Reactive oxygen gene network of plants. Trends Plant Sci 9:490– 498 6. Foyer CH, Noctor G (2009) Redox regulation in photosynthetic organisms: signaling, acclimation, and practical implications. Antioxid Redox Signal 11:861–905 7. Halliwell B (2006) Reactive species and antioxidants. Redox biology is a fundamental theme of aerobic life. Plant Physiol 141:312– 322 8. Miller G, Suzuki N, Ciftci-Yilmaz S, Mittler R (2010) Reactive oxygen species homeostasis and signalling during drought and salinity stresses. Plant Cell Environ 33:453–467 9. Mittler R (2002) Oxidative stress, antioxidants and stress tolerance. Trends Plant Sci 7:405–410 10. Suzuki N, Miller G, Morales J, Shulaev V, Torres MA, Mittler R (2011) Respiratory burst oxidases: the engines of ROS signaling. Curr Opin Plant Biol 14:691–699 11. Mur LA, Kenton P, Lloyd AJ, Ougham H, Prats E (2008) The hypersensitive response; the centenary is upon us but how much do we know? J Exp Bot 59:501–520 12. Doke N (1983) Generation of superoxide anion by potato tuber protoplasts during the hypersensitive response to hyphal wall components of Phytophthora infestans and specific inhibition of the reaction by suppressors of hypersensitivity. Physiol Plant Pathol 23:359–367 13. Doke N (1983) Involvement of superoxide anion generation in the hypersensitive response of potato tuber tissues to infection with an incompatible race of Phytophthora infestans and to the hyphal wall components. Physiol Plant Pathol 23:345–357 14. Doke N (1985) NADPH-dependent O2- generation in membrane fractions isolated from wounded potato tubers inoculated with Phytophthora infestans. Physiol Plant Pathol 27:311–322 15. Wojtaszek P (1997) Oxidative burst: an early plant response to pathogen infection. Biochem J 322(Pt 3):681–692 16. Levine A, Tenhaken R, Dixon R, Lamb C (1994) H2O2 from the oxidative burst orchestrates the plant hypersensitive disease resistance response. Cell 79:583–593 17. Akazawa T, Conn EE (1958) The oxidation of reduced pyridine nucleotides by peroxidase. J Biol Chem 232(1):403–415 18. Elstner EF, Heupel A (1976) Formation of hydrogen peroxide by isolated cell walls from horseradish (Armoracia lapathifolia Gilib.). Planta 130:175–180 19. Gross GG, Janse C, Elstner EF (1977) Involvement of malate, monophenols, and the superoxide radical in hydrogen peroxide formation by isolated cell walls from horseradish (Armoracia lapathifolia Gilib.). Planta 136:271–276 20. Halliwell B (1978) Lignin synthesis: The generation of hydrogen peroxide and superoxide by horseradish peroxidase and its stimulation by manganese (II) and phenols. Planta 140:81–88 21. Mäder M, Amberg-Fisher V (1982) Role of peroxidase in lignification of tobacco cells 1: I. Oxidation of nicotinamide adenine dinucleotide and formation of hydrogen peroxide by cell wall peroxidases. Plant Physiol 70:1128–1131

G. Miller and R. Mittler 22. Pichorner H, Couperus A, Korori SAA, Ebermann R (1992) Plant peroxidase has a thiol oxidase function. Phytochemistry 31:3371– 3376 23. Bolwell GP, Blee KA, Butt VS, Davies DR, Gardner SL, Gerrish C, Minibayeva F, Rowntree EG, Wojtaszek P (1999) Recent advances in understanding the origin of the apoplastic oxidative burst in plant cells. Free Radic Res 31(Suppl):S137– S145 24. Bolwell GP (1999) Role of active oxygen species and NO in plant defence responses. Curr Opin Plant Biol 2:287–294 25. Lindner WA, Hoffmann C, Grisebach H (1988) Rapid elicitorinduced chemiluminescence in soybean cell suspension cultures. Phytochemistry 27:2501–2503 26. Frahry G, Schopfer P (1998) Inhibition of O2-reducing activity of horseradish peroxidase by diphenyleneiodonium. Phytochemistry 48:223–227 27. O’Donnell VB, Tew DG, Jones OTG, England PJ (1993) Studies on the inhibitory mechanism of iodonium compounds with special reference to neutrophil NADPH oxidase. Biochem J 290:41–49 28. Bolwell GP, Butt VS, Davies DR, Zimmerlin A (1995) The origin of the oxidative burst in plants. Free Radic Res 23:517–532 29. Legendre L, Heinstein PF, Low PS (1992) Evidence for participation of GTP-binding proteins in elicitation of the rapid oxidative burst in cultured soybean cells. J Biol Chem 267:20140–20147 30. Legendre L, Rueter S, Heinstein PF, Low PS (1993) Characterization of the oligogalacturonide-induced oxidative burst in cultured soybean (glycine max) cells. Plant Physiol 102:233–240 31. Groom QJ, Torres MA, Fordham-Skelton AP, Hammond-Kosack KE, Robinson NJ, Jones JD (1996) rbohA, a rice homologue of the mammalian gp91phox respiratory burst oxidase gene. Plant J 10: 515–522 32. Torres MA, Onouchi H, Hamada S, Machida C, Hammond-Kosack KE, Jones JD (1998) Six Arabidopsis thaliana homologues of the human respiratory burst oxidase (gp91phox). Plant J 14:365–370 33. Keller T, Damude HG, Werner D, Doerner P, Dixon RA, Lamb C (1998) A plant homolog of the neutrophil NADPH oxidase gp91phox subunit gene encodes a plasma membrane protein with Ca2+ binding motifs. Plant Cell 10:255–266 34. The Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408: 796–815 35. Day IS, Reddy VS, Shad Ali G, Reddy AS (2002) Analysis of EFhand-containing proteins in Arabidopsis. Genome Biol 3: Research0056 36. Wong HL, Pinontoan R, Hayashi K, Tabata R, Yaeno T, Hasegawa K, Kojima C, Yoshioka H, Iba K, Kawasaki T et al (2007) Regulation of rice NADPH oxidase by binding of Rac GTPase to its N-terminal extension. Plant Cell 19:4022–4034 37. Dwyer SC, Legendre L, Low PS, Leto TL (1996) Plant and human neutrophil oxidative burst complexes contain immunologically related proteins. Biochim Biophys Acta 1289:231–237 38. Tenhaken R, Rübel C (1998) Cloning of putative subunits of the soybean plasma membrane NADPH-oxidase involved in the oxidative burst by antibody expression screening. Protoplasma 21–28 39. Sagi M, Fluhr R (2001) Superoxide production by plant homologues of the gp91(phox) NADPH oxidase. Modulation of activity by calcium and by tobacco mosaic virus infection. Plant Physiol 126:1281–1290 40. Torres MA, Dangl JL, Jones JD (2002) Arabidopsis gp91phox homologues AtrbohD and AtrbohF are required for accumulation of reactive oxygen intermediates in the plant defense response. Proc Natl Acad Sci U S A 99:517–522 41. Sagi M, Davydov O, Orazova S, Yesbergenova Z, Ophir R, Stratmann JW, Fluhr R (2004) Plant respiratory burst oxidase homologs impinge on wound responsiveness and development in Lycopersicon esculentum. Plant Cell 16:616–628

26

Plant NADPH Oxidases

42. Marino D, Andrio E, Danchin EG, Oger E, Gucciardo S, Lambert A, Puppo A, Pauly N (2011) A Medicago truncatula NADPH oxidase is involved in symbiotic nodule functioning. New Phytol 189:580–592 43. Nestler J, Liu S, Wen TJ, Paschold A, Marcon C, Tang HM, Li D, Li L, Meeley RB, Sakai H et al (2014) Roothairless5, which functions in maize (Zea mays L.) root hair initiation and elongation encodes a monocot-specific NADPH oxidase. Plant J 79:729–740 44. Simon-Plas F, Elmayan T, Blein JP (2002) The plasma membrane oxidase NtrbohD is responsible for AOS production in elicited tobacco cells. Plant J 31:137–147 45. Yoshioka H, Numata N, Nakajima K, Katou S, Kawakita K, Rowland O, Jones JD, Doke N (2003) Nicotiana benthamiana gp91phox homologs NbrbohA and NbrbohB participate in H2O2 accumulation and resistance to Phytophthora infestans. Plant Cell 15:706–718 46. Kaur G, Guruprasad K, Temple BRS, Shirvanyants DG, Dokholyan NV, Pati PK (2018) Structural complexity and functional diversity of plant NADPH oxidases. Amino Acids 50:79–94 47. Chapman JM, Muhlemann JK, Gayomba SR, Muday GK (2019) RBOH-dependent ROS synthesis and ROS scavenging by plant specialized metabolites to modulate plant development and stress responses. Chem Res Toxicol 32:370–396 48. Kaur G, Sharma A, Guruprasad K, Pati PK (2014) Versatile roles of plant NADPH oxidases and emerging concepts. Biotechnol Adv 32:551–563 49. Sumimoto H (2008) Structure, regulation and evolution of Nox-family NADPH oxidases that produce reactive oxygen species. FEBS J 275:3249–3277 50. Finegold AA, Shatwell KP, Segal AW, Klausner RD, Dancis A (1996) Intramembrane bis-heme motif for transmembrane electron transport conserved in a yeast iron reductase and the human NADPH oxidase. J Biol Chem 271:31021–31024 51. Park J, Gu Y, Lee Y, Yang Z, Lee Y (2004) Phosphatidic acid induces leaf cell death in Arabidopsis by activating the Rho-related small G protein GTPase-mediated pathway of reactive oxygen species generation. Plant Physiol 134:129–136 52. Baxter A, Mittler R, Suzuki N (2014) ROS as key players in plant stress signalling. J Exp Bot 65:1229–1240 53. Belambri SA, Rolas L, Raad H, Hurtado-Nedelec M, Dang PM, El-Benna J (2018) NADPH oxidase activation in neutrophils: Role of the phosphorylation of its subunits. Eur J Clin Invest 48(Suppl 2):e12951 54. Kadota Y, Shirasu K, Zipfel C (2015) Regulation of the NADPH oxidase RBOHD during plant immunity. Plant Cell Physiol 56: 1472–1480 55. Oda T, Hashimoto H, Kuwabara N, Akashi S, Hayashi K, Kojima C, Wong HL, Kawasaki T, Shimamoto K, Sato M et al (2010) Structure of the N-terminal regulatory domain of a plant NADPH oxidase and its functional implications. J Biol Chem 285: 1435–1445 56. Kwak JM, Mori IC, Pei ZM, Leonhardt N, Torres MA, Dangl JL, Bloom RE, Bodde S, Jones JD, Schroeder JI (2003) NADPH oxidase AtrbohD and AtrbohF genes function in ROS-dependent ABA signaling in Arabidopsis. EMBO J 22:2623–2633 57. Foreman J, Demidchik V, Bothwell JH, Mylona P, Miedema H, Torres MA, Linstead P, Costa S, Brownlee C, Jones JD et al (2003) Reactive oxygen species produced by NADPH oxidase regulate plant cell growth. Nature 422:442–446 58. Takahashi S, Kimura S, Kaya H, Iizuka A, Wong HL, Shimamoto K, Kuchitsu K (2012) Reactive oxygen species production and activation mechanism of the rice NADPH oxidase OsRbohB. J Biochem 152:37–43 59. Kimura S, Kaya H, Kawarazaki T, Hiraoka G, Senzaki E, Michikawa M, Kuchitsu K (2012) Protein phosphorylation is a prerequisite for the Ca2+-dependent activation of Arabidopsis

461 NADPH oxidases and may function as a trigger for the positive feedback regulation of Ca2+ and reactive oxygen species. Biochim Biophys Acta 1823:398–405 60. Ogasawara Y, Kaya H, Hiraoka G, Yumoto F, Kimura S, Kadota Y, Hishinuma H, Senzaki E, Yamagoe S, Nagata K et al (2008) Synergistic activation of the Arabidopsis NADPH oxidase AtrbohD by Ca2+ and phosphorylation. J Biol Chem 283:8885– 8892 61. Han JP, Koster P, Drerup MM, Scholz M, Li S, Edel KH, Hashimoto K, Kuchitsu K, Hippler M, Kudla J (2019) Fine-tuning of RBOHF activity is achieved by differential phosphorylation and Ca(2+) binding. New Phytol 221:1935–1949 62. Drerup MM, Schlucking K, Hashimoto K, Manishankar P, Steinhorst L, Kuchitsu K, Kudla J (2013) The Calcineurin B-like calcium sensors CBL1 and CBL9 together with their interacting protein kinase CIPK26 regulate the Arabidopsis NADPH oxidase RBOHF. Mol Plant 6:559–569 63. Takeda S, Gapper C, Kaya H, Bell E, Kuchitsu K, Dolan L (2008) Local positive feedback regulation determines cell shape in root hair cells. Science 319:1241–1244 64. Kaya H, Takeda S, Kobayashi MJ, Kimura S, Iizuka A, Imai A, Hishinuma H, Kawarazaki T, Mori K, Yamamoto Y et al (2019) Comparative analysis of the reactive oxygen species-producing enzymatic activity of Arabidopsis NADPH oxidases. Plant J 98: 291–300 65. Kaya H, Nakajima R, Iwano M, Kanaoka MM, Kimura S, Takeda S, Kawarazaki T, Senzaki E, Hamamura Y, Higashiyama T et al (2014) Ca2+-activated reactive oxygen species production by Arabidopsis RbohH and RbohJ is essential for proper pollen tube tip growth. Plant Cell 26:1069–1080 66. Luan S (2009) The CBL-CIPK network in plant calcium signaling. Trends Plant Sci 14:37–42 67. Karkonen A, Kuchitsu K (2015) Reactive oxygen species in cell wall metabolism and development in plants. Phytochemistry 112: 22–32 68. Jammes F, Song C, Shin D, Munemasa S, Takeda K, Gu D, Cho D, Lee S, Giordo R, Sritubtim S et al (2009) MAP kinases MPK9 and MPK12 are preferentially expressed in guard cells and positively regulate ROS-mediated ABA signaling. Proc Natl Acad Sci U S A 106:20520–20525 69. Asai S, Ohta K, Yoshioka H (2008) MAPK signaling regulates nitric oxide and NADPH oxidase-dependent oxidative bursts in Nicotiana benthamiana. Plant Cell 20:1390–1406 70. Zhang J, Shao F, Li Y, Cui H, Chen L, Li H, Zou Y, Long C, Lan L, Chai J et al (2007) A Pseudomonas syringae effector inactivates MAPKs to suppress PAMP-induced immunity in plants. Cell Host Microbe 1:175–185 71. Nuhse TS, Bottrill AR, Jones AM, Peck SC (2007) Quantitative phosphoproteomic analysis of plasma membrane proteins reveals regulatory mechanisms of plant innate immune responses. Plant J 51:931–940 72. Sirichandra C, Gu D, Hu HC, Davanture M, Lee S, Djaoui M, Valot B, Zivy M, Leung J, Merlot S et al (2009) Phosphorylation of the Arabidopsis AtrbohF NADPH oxidase by OST1 protein kinase. FEBS Lett 583:2982–2986 73. Kobayashi M, Ohura I, Kawakita K, Yokota N, Fujiwara M, Shimamoto K, Doke N, Yoshioka H (2007) Calcium-dependent protein kinases regulate the production of reactive oxygen species by potato NADPH oxidase. Plant Cell 19:1065–1080 74. Benschop JJ, Mohammed S, O’Flaherty M, Heck AJ, Slijper M, Menke FL (2007) Quantitative phosphoproteomics of early elicitor signaling in Arabidopsis. Mol Cell Proteomics 6:1198–1214 75. Takahashi F, Mizoguchi T, Yoshida R, Ichimura K, Shinozaki K (2011) Calmodulin-dependent activation of MAP kinase for ROS homeostasis in Arabidopsis. Mol Cell 41:649–660

462 76. Zhang Y, Zhu H, Zhang Q, Li M, Yan M, Wang R, Wang L, Welti R, Zhang W, Wang X (2009) Phospholipase dalpha1 and phosphatidic acid regulate NADPH oxidase activity and production of reactive oxygen species in ABA-mediated stomatal closure in Arabidopsis. Plant Cell 21:2357–2377 77. Monaghan J, Matschi S, Shorinola O, Rovenich H, Matei A, Segonzac C, Malinovsky FG, Rathjen JP, MacLean D, Romeis T et al (2014) The calcium-dependent protein kinase CPK28 buffers plant immunity and regulates BIK1 turnover. Cell Host Microbe 16:605–615 78. Dubiella U, Seybold H, Durian G, Komander E, Lassig R, Witte CP, Schulze WX, Romeis T (2013) Calcium-dependent protein kinase/NADPH oxidase activation circuit is required for rapid defense signal propagation. Proc Natl Acad Sci U S A 110:8744– 8749 79. Gao X, Chen X, Lin W, Chen S, Lu D, Niu Y, Li L, Cheng C, McCormack M, Sheen J et al (2013) Bifurcation of Arabidopsis NLR immune signaling via Ca(2)(+)-dependent protein kinases. PLoS Pathog 9:e1003127 80. Yun BW, Feechan A, Yin M, Saidi NB, Le Bihan T, Yu M, Moore JW, Kang JG, Kwon E, Spoel SH et al (2011) S-nitrosylation of NADPH oxidase regulates cell death in plant immunity. Nature 478:264–268 81. Chu-Puga A, Gonzalez-Gordo S, Rodriguez-Ruiz M, Palma JM, Corpas FJ (2019) NADPH oxidase (Rboh) activity is up regulated during sweet pepper (Capsicum annuum L.) fruit ripening. Antioxidants (Basel) 8 82. Luthje S, Moller B, Perrineau FC, Woltje K (2013) Plasma membrane electron pathways and oxidative stress. Antioxid Redox Signal 18:2163–2183 83. Fujiwara M, Hamada S, Hiratsuka M, Fukao Y, Kawasaki T, Shimamoto K (2009) Proteome analysis of detergent-resistant membranes (DRMs) associated with OsRac1-mediated innate immunity in rice. Plant Cell Physiol 50:1191–1200 84. Morel J, Claverol S, Mongrand S, Furt F, Fromentin J, Bessoule JJ, Blein JP, Simon-Plas F (2006) Proteomics of plant detergentresistant membranes. Mol Cell Proteomics 5:1396–1411 85. Mongrand S, Morel J, Laroche J, Claverol S, Carde JP, Hartmann MA, Bonneu M, Simon-Plas F, Lessire R, Bessoule JJ (2004) Lipid rafts in higher plant cells: purification and characterization of Triton X-100-insoluble microdomains from tobacco plasma membrane. J Biol Chem 279:36277–36286 86. Liu P, Li RL, Zhang L, Wang QL, Niehaus K, Baluska F, Samaj J, Lin JX (2009) Lipid microdomain polarization is required for NADPH oxidase-dependent ROS signaling in Picea meyeri pollen tube tip growth. Plant J 60:303–313 87. Noirot E, Der C, Lherminier J, Robert F, Moricova P, Kieu K, Leborgne-Castel N, Simon-Plas F, Bouhidel K (2014) Dynamic changes in the subcellular distribution of the tobacco ROS-producing enzyme RBOHD in response to the oomycete elicitor cryptogein. J Exp Bot 65:5011–5022 88. Leshem Y, Seri L, Levine A (2007) Induction of phosphatidylinositol 3-kinase-mediated endocytosis by salt stress leads to intracellular production of reactive oxygen species and salt tolerance. Plant J 51:185–197 89. Leshem Y, Golani Y, Kaye Y, Levine A (2010) Reduced expression of the v-SNAREs AtVAMP71/AtVAMP7C gene family in Arabidopsis reduces drought tolerance by suppression of abscisic acid-dependent stomatal closure. J Exp Bot 61:2615–2622 90. Touyz RM, Anagnostopoulou A, Camargo LL, Rios FJ, Montezano AC (2019) Vascular biology of superoxide-generating NADPH oxidase 5-implications in hypertension and cardiovascular disease. Antioxid Redox Signal 30:1027–1040 91. Block K, Gorin Y (2012) Aiding and abetting roles of NOX oxidases in cellular transformation. Nat Rev Cancer 12:627–637

G. Miller and R. Mittler 92. Graham KA, Kulawiec M, Owens KM, Li X, Desouki MM, Chandra D, Singh KK (2010) NADPH oxidase 4 is an oncoprotein localized to mitochondria. Cancer Biol Ther 10:223–231 93. Hilenski LL, Clempus RE, Quinn MT, Lambeth JD, Griendling KK (2004) Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 24: 677–683 94. Sirokmany G, Donko A, Geiszt M (2016) Nox/Duox family of NADPH oxidases: lessons from knockout mouse models. Trends Pharmacol Sci 37:318–327 95. Van Buul JD, Fernandez-Borja M, Anthony EC, Hordijk PL (2005) Expression and localization of NOX2 and NOX4 in primary human endothelial cells. Antioxid Redox Signal 7:308–317 96. Torres MA, Dangl JL (2005) Functions of the respiratory burst oxidase in biotic interactions, abiotic stress and development. Curr Opin Plant Biol 8:397–403 97. Miller G, Schlauch K, Tam R, Cortes D, Torres MA, Shulaev V, Dangl JL, Mittler R (2009) The plant NADPH oxidase RBOHD mediates rapid systemic signaling in response to diverse stimuli. Sci Signal 2:ra45 98. Davletova S, Rizhsky L, Liang H, Shengqiang Z, Oliver DJ, Coutu J, Shulaev V, Schlauch K, Mittler R (2005) Cytosolic ascorbate peroxidase 1 is a central component of the reactive oxygen gene network of Arabidopsis. Plant Cell 17:268–281 99. Mase K, Tsukagoshi H (2021) Reactive oxygen species link gene regulatory networks during Arabidopsis root development. Front Plant Sci 12:660274 100. Krieger G, Shkolnik D, Miller G, Fromm H (2016) Reactive oxygen species tune root tropic responses. Plant Physiol 172: 1209–1220 101. Barbosa ICR, Rojas-Murcia N, Geldner N (2019) The Casparian strip-one ring to bring cell biology to lignification? Curr Opin Biotechnol 56:121–129 102. Doblas VG, Geldner N, Barberon M (2017) The endodermis, a tightly controlled barrier for nutrients. Curr Opin Plant Biol 39: 136–143 103. Dixon RA, Barros J (2019) Lignin biosynthesis: old roads revisited and new roads explored. Open Biol 9:190215 104. Lee Y, Rubio MC, Alassimone J, Geldner N (2013) A mechanism for localized lignin deposition in the endodermis. Cell 153:402– 412 105. Orman-Ligeza B, Parizot B, de Rycke R, Fernandez A, Himschoot E, Van Breusegem F, Bennett MJ, Perilleux C, Beeckman T, Draye X (2016) RBOH-mediated ROS production facilitates lateral root emergence in Arabidopsis. Development 143:3328–3339 106. Manzano C, Pallero-Baena M, Casimiro I, De Rybel B, OrmanLigeza B, Van Isterdael G, Beeckman T, Draye X, Casero P, Del Pozo JC (2014) The emerging role of reactive oxygen species signaling during lateral root development. Plant Physiol 165: 1105–1119 107. Li N, Sun L, Zhang L, Song Y, Hu P, Li C, Hao FS (2015) AtrbohD and AtrbohF negatively regulate lateral root development by changing the localized accumulation of superoxide in primary roots of Arabidopsis. Planta 241:591–602 108. Grierson C, Schiefelbein J (2002) Root hairs. Arabidopsis Book 1: e0060 109. Schiefelbein JW, Somerville C (1990) Genetic control of root hair development in Arabidopsis thaliana. Plant Cell 2:235–243 110. Schopfer P (2001) Hydroxyl radical-induced cell-wall loosening in vitro and in vivo: implications for the control of elongation growth. Plant J 28:679–688 111. Pitts RJ, Cernac A, Estelle M (1998) Auxin and ethylene promote root hair elongation in Arabidopsis. Plant J 16:553–560 112. Rahman A, Hosokawa S, Oono Y, Amakawa T, Goto N, Tsurumi S (2002) Auxin and ethylene response interactions during

26

Plant NADPH Oxidases

Arabidopsis root hair development dissected by auxin influx modulators. Plant Physiol 130:1908–1917 113. Mangano S, Denita-Juarez SP, Marzol E, Borassi C, Estevez JM (2018) High auxin and high phosphate impact on RSL2 expression and ROS-homeostasis linked to root hair growth in Arabidopsis thaliana. Front Plant Sci 9:1164 114. Masucci JD, Schiefelbein JW (1994) The rhd6 mutation of Arabidopsis thaliana alters root-hair initiation through an auxinand ethylene-associated process. Plant Physiol 106:1335–1346 115. Yi K, Menand B, Bell E, Dolan L (2010) A basic helix-loop-helix transcription factor controls cell growth and size in root hairs. Nat Genet 42:264–267 116. Menand B, Yi K, Jouannic S, Hoffmann L, Ryan E, Linstead P, Schaefer DG, Dolan L (2007) An ancient mechanism controls the development of cells with a rooting function in land plants. Science 316:1477–1480 117. Mangano S, Denita-Juarez SP, Choi HS, Marzol E, Hwang Y, Ranocha P, Velasquez SM, Borassi C, Barberini ML, Aptekmann AA et al (2017) Molecular link between auxin and ROS-mediated polar growth. Proc Natl Acad Sci U S A 114:5289–5294 118. Suzuki G (2009) Recent progress in plant reproduction research: the story of the male gametophyte through to successful fertilization. Plant Cell Physiol 50:1857–1864 119. Flores-Renteria L, Orozco-Arroyo G, Cruz-Garcia F, GarciaCampusano F, Alfaro I, Vazquez-Santana S (2013) Programmed cell death promotes male sterility in the functional dioecious Opuntia stenopetala (Cactaceae). Ann Bot 112:789–800 120. Xie HT, Wan ZY, Li S, Zhang Y (2014) Spatiotemporal production of reactive oxygen species by NADPH oxidase is critical for tapetal programmed cell death and pollen development in Arabidopsis. Plant Cell 26:2007–2023 121. Lassig R, Gutermuth T, Bey TD, Konrad KR, Romeis T (2014) Pollen tube NAD(P)H oxidases act as a speed control to dampen growth rate oscillations during polarized cell growth. Plant J 78: 94–106 122. Duan Q, Kita D, Johnson EA, Aggarwal M, Gates L, Wu HM, Cheung AY (2014) Reactive oxygen species mediate pollen tube rupture to release sperm for fertilization in Arabidopsis. Nat Commun 5:3129 123. Boisson-Dernier A, Lituiev DS, Nestorova A, Franck CM, Thirugnanarajah S, Grossniklaus U (2013) ANXUR receptor-like kinases coordinate cell wall integrity with growth at the pollen tube tip via NADPH oxidases. PLoS Biol 11:e1001719 124. Potocky M, Jones MA, Bezvoda R, Smirnoff N, Zarsky V (2007) Reactive oxygen species produced by NADPH oxidase are involved in pollen tube growth. New Phytol 174:742–751 125. Hepler PK, Vidali L, Cheung AY (2001) Polarized cell growth in higher plants. Annu Rev Cell Dev Biol 17:159–187 126. Marino D, Dunand C, Puppo A, Pauly N (2012) A burst of plant NADPH oxidases. Trends Plant Sci 17:9–15 127. Breygina M, Klimenko E (2020) ROS and ions in cell signaling during sexual plant reproduction. Int J Mol Sci 21 128. Hayashi M, Palmgren M (2021) The quest for the central players governing pollen tube growth and guidance. Plant Physiol 185: 682–693 129. Finch-Savage WE, Leubner-Metzger G (2006) Seed dormancy and the control of germination. New Phytol 171:501–523 130. Finkelstein R, Reeves W, Ariizumi T, Steber C (2008) Molecular aspects of seed dormancy. Annu Rev Plant Biol 59:387–415 131. Ishibashi Y, Tawaratsumida T, Kondo K, Kasa S, Sakamoto M, Aoki N, Zheng SH, Yuasa T, Iwaya-Inoue M (2012) Reactive oxygen species are involved in gibberellin/abscisic acid signaling in barley aleurone cells. Plant Physiol 158:1705–1714 132. Bahin E, Bailly C, Sotta B, Kranner I, Corbineau F, Leymarie J (2011) Crosstalk between reactive oxygen species and hormonal

463 signalling pathways regulates grain dormancy in barley. Plant Cell Environ 34:980–993 133. Liu Y, Ye N, Liu R, Chen M, Zhang J (2010) H2O2 mediates the regulation of ABA catabolism and GA biosynthesis in Arabidopsis seed dormancy and germination. J Exp Bot 61:2979–2990 134. Krishnamurthy A, Rathinasabapathi B (2013) Oxidative stress tolerance in plants: novel interplay between auxin and reactive oxygen species signaling. Plant Signal Behav 8. https://doi.org/ 10.4161/psb.25761 135. Job C, Rajjou L, Lovigny Y, Belghazi M, Job D (2005) Patterns of protein oxidation in Arabidopsis seeds and during germination. Plant Physiol 138:790–802 136. Oracz K, El-Maarouf-Bouteau H, Kranner I, Bogatek R, Corbineau F, Bailly C (2009) The mechanisms involved in seed dormancy alleviation by hydrogen cyanide unravel the role of reactive oxygen species as key factors of cellular signaling during germination. Plant Physiol 150:494–505 137. Bazin J, Langlade N, Vincourt P, Arribat S, Balzergue S, ElMaarouf-Bouteau H, Bailly C (2011) Targeted mRNA oxidation regulates sunflower seed dormancy alleviation during dry afterripening. Plant Cell 23:2196–2208 138. El-Maarouf-Bouteau H, Meimoun P, Job C, Job D, Bailly C (2013) Role of protein and mRNA oxidation in seed dormancy and germination. Front Plant Sci 4:77 139. Oracz K, Stawska M (2016) Cellular recycling of proteins in seed dormancy alleviation and germination. Front Plant Sci 7:1128 140. Katsuya-Gaviria K, Caro E, Carrillo-Barral N, Iglesias-Fernandez R (2020) Reactive oxygen species (ROS) and nucleic acid modifications during seed dormancy. Plants (Basel) 9 141. Oracz K, El-Maarouf Bouteau H, Farrant JM, Cooper K, Belghazi M, Job C, Job D, Corbineau F, Bailly C (2007) ROS production and protein oxidation as a novel mechanism for seed dormancy alleviation. Plant J 50:452–465 142. Muller K, Carstens AC, Linkies A, Torres MA, Leubner-Metzger G (2009) The NADPH-oxidase AtrbohB plays a role in Arabidopsis seed after-ripening. New Phytol 184:885–897 143. Morales J, Kadota Y, Zipfel C, Molina A, Torres MA (2016) The Arabidopsis NADPH oxidases RbohD and RbohF display differential expression patterns and contributions during plant immunity. J Exp Bot 67:1663–1676 144. Otulak-Koziel K, Koziel E, Bujarski JJ, Frankowska-Lukawska J, Torres MA (2020) Respiratory burst oxidase homologs RBOHD and RBOHF as key modulating components of response in turnip mosaic virus-Arabidopsis thaliana (L.) Heyhn system. Int J Mol Sci 21 145. Ma L, Zhang H, Sun L, Jiao Y, Zhang G, Miao C, Hao F (2012) NADPH oxidase AtrbohD and AtrbohF function in ROS-dependent regulation of Na(+)/K(+)homeostasis in Arabidopsis under salt stress. J Exp Bot 63:305–317 146. Liu B, Sun L, Ma L, Hao FS (2017) Both AtrbohD and AtrbohF are essential for mediating responses to oxygen deficiency in Arabidopsis. Plant Cell Rep 36:947–957 147. Guan B, Lin Z, Liu D, Li C, Zhou Z, Mei F, Li J, Deng X (2019) Effect of waterlogging-induced autophagy on programmed cell death in Arabidopsis roots. Front Plant Sci 10:468 148. Joo JH, Wang S, Chen JG, Jones AM, Fedoroff NV (2005) Different signaling and cell death roles of heterotrimeric G protein alpha and beta subunits in the Arabidopsis oxidative stress response to ozone. Plant Cell 17:957–970 149. Zhai L, Sun C, Feng Y, Li D, Chai X, Wang L, Sun Q, Zhang G, Li Y, Wu T et al (2018) AtROP6 is involved in reactive oxygen species signaling in response to iron-deficiency stress in Arabidopsis thaliana. FEBS Lett 592:3446–3459 150. He H, Yan J, Yu X, Liang Y, Fang L, Scheller HV, Zhang A (2017) The NADPH-oxidase AtRbohI plays a positive role in drought-

464 stress response in Arabidopsis thaliana. Biochem Biophys Res Commun 491:834–839 151. Yamauchi T, Yoshioka M, Fukazawa A, Mori H, Nishizawa NK, Tsutsumi N, Yoshioka H, Nakazono M (2017) An NADPH oxidase RBOH functions in rice roots during lysigenous Aerenchyma formation under oxygen-deficient conditions. Plant Cell 29:775– 790 152. Yamauchi T, Rajhi I, Nakazono M (2011) Lysigenous aerenchyma formation in maize root is confined to cortical cells by regulation of genes related to generation and scavenging of reactive oxygen species. Plant Signal Behav 6:759–761 153. Torres MA (2010) ROS in biotic interactions. Physiol Plant 138: 414–429 154. Lamb C, Dixon RA (1997) The oxidative burst in plant disease resistance. Annu Rev Plant Physiol Plant Mol Biol 48:251–275 155. Pfeilmeier S, Petti GC, Bortfeld-Miller M, Daniel B, Field CM, Sunagawa S, Vorholt JA (2021) The plant NADPH oxidase RBOHD is required for microbiota homeostasis in leaves. Nat Microbiol 6:852–864 156. Yu HH, Yang YH, Chiang BL (2021) Chronic granulomatous disease: a comprehensive review. Clin Rev Allergy Immunol 61: 101–113 157. Jwa NS, Hwang BK (2017) Convergent evolution of pathogen effectors toward reactive oxygen species signaling networks in plants. Front Plant Sci 8:1687 158. Balint-Kurti P (2019) The plant hypersensitive response: concepts, control and consequences. Mol Plant Pathol 20:1163–1178 159. Han GZ (2019) Origin and evolution of the plant immune system. New Phytol 222:70–83 160. Dangl JL, Horvath DM, Staskawicz BJ (2013) Pivoting the plant immune system from dissection to deployment. Science 341:746– 751 161. Jones JD, Dangl JL (2006) The plant immune system. Nature 444: 323–329 162. Friesen TL, Faris JD (2021) Characterization of effector-target interactions in necrotrophic pathosystems reveals trends and variation in host manipulation. Annu Rev Phytopathol 59:77–98 163. Zipfel C (2014) Plant pattern-recognition receptors. Trends Immunol 35:345–351 164. DeFalco TA, Zipfel C (2021) Molecular mechanisms of early plant pattern-triggered immune signaling. Mol Cell 81:3449–3467 165. Segonzac C, Feike D, Gimenez-Ibanez S, Hann DR, Zipfel C, Rathjen JP (2011) Hierarchy and roles of pathogen-associated molecular pattern-induced responses in Nicotiana benthamiana. Plant Physiol 156:687–699 166. Kadota Y, Sklenar J, Derbyshire P, Stransfeld L, Asai S, Ntoukakis V, Jones JD, Shirasu K, Menke F, Jones A et al (2014) Direct regulation of the NADPH oxidase RBOHD by the PRR-associated kinase BIK1 during plant immunity. Mol Cell 54: 43–55 167. Singh R, Dangol S, Chen Y, Choi J, Cho YS, Lee JE, Choi MO, Jwa NS (2016) Magnaporthe oryzae effector AVR-Pii helps to establish compatibility by inhibition of the rice NADP-malic enzyme resulting in disruption of oxidative burst and host innate immunity. Mol Cells 39:426–438 168. Kwak JM, Maser P, Schroeder JI (2008) The clickable guard cell, version II: interactive model of guard cell signal transduction mechanisms and pathways. Arabidopsis Book 6:e0114 169. Ma Y, Szostkiewicz I, Korte A, Moes D, Yang Y, Christmann A, Grill E (2009) Regulators of PP2C phosphatase activity function as abscisic acid sensors. Science 324:1064–1068 170. Park SY, Fung P, Nishimura N, Jensen DR, Fujii H, Zhao Y, Lumba S, Santiago J, Rodrigues A, Chow TF et al (2009) Abscisic acid inhibits type 2C protein phosphatases via the PYR/PYL family of START proteins. Science 324:1068–1071

G. Miller and R. Mittler 171. Joshi-Saha A, Valon C, Leung J (2011) A brand new START: abscisic acid perception and transduction in the guard cell. Sci Signal 4:re4 172. Pei ZM, Murata Y, Benning G, Thomine S, Klusener B, Allen GJ, Grill E, Schroeder JI (2000) Calcium channels activated by hydrogen peroxide mediate abscisic acid signalling in guard cells. Nature 406:731–734 173. Schroeder JI, Hagiwara S (1990) Repetitive increases in cytosolic Ca2+ of guard cells by abscisic acid activation of nonselective Ca2+ permeable channels. Proc Natl Acad Sci U S A 87:9305– 9309 174. Hedrich R, Geiger D (2017) Biology of SLAC1-type anion channels—from nutrient uptake to stomatal closure. New Phytol 216:46–61 175. Mittler R, Blumwald E (2015) The roles of ROS and ABA in systemic acquired acclimation. Plant Cell 27:64–70 176. Sokolovski S, Hills A, Gay R, Garcia-Mata C, Lamattina L, Blatt MR (2005) Protein phosphorylation is a prerequisite for intracellular Ca2+ release and ion channel control by nitric oxide and abscisic acid in guard cells. Plant J 43:520–529 177. An Z, Jing W, Liu Y, Zhang W (2008) Hydrogen peroxide generated by copper amine oxidase is involved in abscisic acidinduced stomatal closure in Vicia faba. J Exp Bot 59:815–825 178. Khokon AR, Okuma E, Hossain MA, Munemasa S, Uraji M, Nakamura Y, Mori IC, Murata Y (2011) Involvement of extracellular oxidative burst in salicylic acid-induced stomatal closure in Arabidopsis. Plant Cell Environ 34:434–443 179. Watkins JM, Chapman JM, Muday GK (2017) Abscisic acidinduced reactive oxygen species are modulated by flavonols to control stomata aperture. Plant Physiol 175:1807–1825 180. Wang P, Song CP (2008) Guard-cell signalling for hydrogen peroxide and abscisic acid. New Phytol 178:703–718 181. Mustilli AC, Merlot S, Vavasseur A, Fenzi F, Giraudat J (2002) Arabidopsis OST1 protein kinase mediates the regulation of stomatal aperture by abscisic acid and acts upstream of reactive oxygen species production. Plant Cell 14:3089–3099 182. Fujii H, Zhu JK (2009) Arabidopsis mutant deficient in 3 abscisic acid-activated protein kinases reveals critical roles in growth, reproduction, and stress. Proc Natl Acad Sci U S A 106:8380–8385 183. Devireddy AR, Zandalinas SI, Gomez-Cadenas A, Blumwald E, Mittler R (2018) Coordinating the overall stomatal response of plants: Rapid leaf-to-leaf communication during light stress. Sci Signal 11 184. Melotto M, Underwood W, Koczan J, Nomura K, He SY (2006) Plant stomata function in innate immunity against bacterial invasion. Cell 126:969–980 185. Gilroy S, Suzuki N, Miller G, Choi WG, Toyota M, Devireddy AR, Mittler R (2014) A tidal wave of signals: calcium and ROS at the forefront of rapid systemic signaling. Trends Plant Sci 19:623–630 186. Karpinski S, Szechynska-Hebda M, Wituszynska W, Burdiak P (2013) Light acclimation, retrograde signalling, cell death and immune defences in plants. Plant Cell Environ 36:736–744 187. Fichman Y, Miller G, Mittler R (2019) Whole-plant live imaging of reactive oxygen species. Mol Plant 12:1203–1210 188. Hattori H, Subramanian KK, Sakai J, Jia Y, Li Y, Porter TF, Loison F, Sarraj B, Kasorn A, Jo H et al (2010) Small-molecule screen identifies reactive oxygen species as key regulators of neutrophil chemotaxis. Proc Natl Acad Sci U S A 107:3546–3551 189. Niethammer P, Grabher C, Look AT, Mitchison TJ (2009) A tissue-scale gradient of hydrogen peroxide mediates rapid wound detection in zebrafish. Nature 459:996–999 190. Soderberg A, Barral AM, Soderstrom M, Sander B, Rosen A (2007) Redox-signaling transmitted in trans to neighboring cells by melanoma-derived TNF-containing exosomes. Free Radic Biol Med 43:90–99

26

Plant NADPH Oxidases

191. Suzuki N, Miller G, Salazar C, Mondal HA, Shulaev E, Cortes DF, Shuman JL, Luo X, Shah J, Schlauch K et al (2013) Temporalspatial interaction between reactive oxygen species and abscisic acid regulates rapid systemic acclimation in plants. Plant Cell 25: 3553–3569 192. Zandalinas SI, Mittler R (2018) ROS-induced ROS release in plant and animal cells. Free Radic Biol Med 122:21–27 193. Zorov DB, Filburn CR, Klotz LO, Zweier JL, Sollott SJ (2000) Reactive oxygen species (ROS)-induced ROS release: a new phenomenon accompanying induction of the mitochondrial permeability transition in cardiac myocytes. J Exp Med 192:1001–1014 194. Zorov DB, Juhaszova M, Sollott SJ (2014) Mitochondrial reactive oxygen species (ROS) and ROS-induced ROS release. Physiol Rev 94:909–950 195. Dikalov SI, Li W, Doughan AK, Blanco RR, Zafari AM (2012) Mitochondrial reactive oxygen species and calcium uptake regulate activation of phagocytic NADPH oxidase. Am J Physiol Regul Integr Comp Physiol 302:R1134–R1142 196. Choi WG, Miller G, Wallace I, Harper J, Mittler R, Gilroy S (2017) Orchestrating rapid long-distance signaling in plants with Ca(2+), ROS and electrical signals. Plant J 90:698–707

465 197. Choi WG, Toyota M, Kim SH, Hilleary R, Gilroy S (2014) Salt stress-induced Ca2+ waves are associated with rapid, long-distance root-to-shoot signaling in plants. Proc Natl Acad Sci U S A 111: 6497–6502 198. Evans MJ, Choi WG, Gilroy S, Morris RJ (2016) A ROS-assisted calcium wave dependent on the AtRBOHD NADPH oxidase and TPC1 cation channel propagates the systemic response to salt stress. Plant Physiol 171:1771–1784 199. Mousavi SA, Chauvin A, Pascaud F, Kellenberger S, Farmer EE (2013) Glutamate receptor-like genes mediate leaf-to-leaf wound signalling. Nature 500:422–426 200. Toyota M, Spencer D, Sawai-Toyota S, Jiaqi W, Zhang T, Koo AJ, Howe GA, Gilroy S (2018) Glutamate triggers long-distance, calcium-based plant defense signaling. Science 361:1112–1115 201. Julca I, Ferrari C, Flores-Tornero M, Proost S, Lindner AC, Hackenberg D, Steinbachova L, Michaelidis C, Gomes Pereira S, Misra CS et al (2021) Comparative transcriptomic analysis reveals conserved programmes underpinning organogenesis and reproduction in land plants. Nat Plants 7:1143–1159

Nematode Noxes: The DUOXes of Caenorhabditis elegans

27

Danielle A. Garsin

Abstract

Caenorhabditis elegans (C. elegans) is a key model organism that has been used to elucidate conserved mechanisms and components involved in development, immune response, and aging. Relevant to this chapter, C. elegans has been used to study the roles and functions of NADPH oxidases and has contributed to the overall understanding of these enzymes. Unlike mammals that have 5 NOX and 2 DUOX NADPH oxidases, the C. elegans genome only encodes two DUOX enzymes. One of these, BLI-3, is critical for viability and has ascribed biological roles in development, immune response, and aging. The second, DUOX-2, arises from a weakly expressed gene and only one report describes a phenotype associated with its loss. Therefore, this chapter will focus mostly on BLI-3, examining its sequence and structural features, its expression with respect to developmental stage and tissue location, its biological roles in development, immune response, aging, and stress response, and its regulators and functional interactors. BLI-3 structure, and function will be compared to mammalian DUOX enzymes to highlight and explore relevant similarities and differences. Finally, important gaps in knowledge and avenues for future investigation will be discussed. Keywords

Caenorhabditis elegans · BLI-3 · DUOX · NADPH oxidase · Blistered · Cuticle development · Immunity · Aging

D. A. Garsin (✉) Department of Microbiology and Molecular Genetics, The University of Texas Health Science Center at Houston, Houston, TX, USA e-mail: [email protected]

1

Introduction

The seeds to the discovery of C. elegans DUOX homologs dates to 1974 and the first description of the use of C. elegans as a model organism by Sydney Brenner. Among the many mutants described in this seminal work were those called “blistered,” which were given the genetic nomenclature of bli [1]. These mutants have fluid-filled blisters along their cuticles, the collagenous material that constitutes the outer, flexible, exoskeleton of C. elegans [2] (Fig. 27.1). Four independent genes were mapped, bli-1-4. Among these bli mutants were two alleles of bli-3. Fast forward to 2001. Edens et al. described the BLAST search, cloning and sequencing of two C. elegans genes homologous to gp91phox, the catalytic subunit of the mammalian phagosomal NADPH oxidase [4]. However, unlike gp91phox, the proteins encoded by these two genes with the sequence names F56C11.1 and F53G12.3, had an additional N-terminal peroxidase domain and other structural features, described in more detail below, that classified them as dual oxidases. The genes were predicted to encode highly homologous (>90% identity) DUOXes, likely resulting from a duplication event [4]. When the expression of the genes was disrupted using RNAi, a range of phenotypes emerged, the most prevalent being large superficial blisters [4]. However, the connection back to Sydney Brenner’s bli mutants was not elucidated until 2003 when a genome-wide RNAi screen of C. elegans discovered RNAi clones, including one with the F56C11.1 sequence, that cause bli phenotypes [5]. Sequencing of the genes corresponding to the RNAi clones in mutants containing bli alleles was undertaken. The two alleles of bli-3 had point mutations in the F56C11.1 sequence; F56C11.1 is now known as bli-3. The second gene with the sequence name F53G12.3 is now called duox-2. It has low levels of expression [6]. A mutant containing a deletion of the gene was first characterized in 2009 and found to be normal in appearance; loss of the gene caused no blisters or other visible phenotypes

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_27

467

468

D. A. Garsin

100µM

100µM

control

bli-3 RNAi

Fig. 27.1 The bli phenotype. Pictures of C. elegans exposed to 1/10 bli-3 RNAi compared to control animals using previously described methodology [3]. Blisters, marked with black arrows, are apparent. (Photo courtesy of Melissa R. Cruz)

characteristic of cuticle changes [3]. However, a recent report documented a role for duox-2 in vulva formation that will be discussed in the development section [7]. Otherwise, this chapter will focus exclusively on BLI-3, which has documented roles in development of the cuticle, pathogen defense, and aging.

2

Structural Characteristics and Expression

2.1

Structural Features

Like DUOXes in other species including humans, BLI-3 has an N-terminal peroxidase domain and a C-terminal NADPH oxidase domain (Fig. 27.2). The heme peroxidase domain is preceded by a 21 amino acid signal sequence and followed by a transmembrane domain. These features result in a structural organization typical of DUOXes. The protein is predicted to localize to the cytoplasmic membrane with the peroxidase domain of BLI-3 in the extracellular space. Following the first transmembrane domain are calcium-binding EF-hands predicted to be on the cytoplasmic side of the membrane (Fig. 27.2). These EF-hands are poorly conserved in BLI-3 compared to human DUOX1/2 [4]. Whether they are capable of binding calcium is currently unknown; in a heterologous expression system, BLI-3 did not require calcium to be

activated [8]. Next, another hydrophobic region predicted to consist of six transmembrane domains is followed by the cytoplasmic NADPH binding/FAD-binding/catalytic site of the oxidase domain [4]. Looking more closely at the NADPH oxidase domain, considerable homology with human DUOXes in the flavin adenine dinucleotide (FAD)-binding regions and NADPHbinding regions (60–90%), including the canonical dinucleotide-binding helix GXGXXP was noted [4]. Characterized mutations in the NADPH oxidase include a mutation in a highly conserved proline (P1311L) which resulted in less hydrogen peroxide production in C. elegans carrying this mutation under inducing conditions [9] and in a heterologous mammalian cell expression system in which this mutant protein was produced [8]. The peroxidase domain of BLI-3 was able to covalently bind heme and exhibited intrinsic peroxidase activity when tested in vitro, unlike the peroxidase domain of human DUOX1 [10]. The domain also was shown to associate with calcium presumably via a calcium binding site that is well conserved when compared to classic heme peroxidases such as myeloperoxidase (MPO) and lactoperoxidase (LPO) [10]. Both BLI-3 and human DUOX1 contain the critical catalytic arginine residue, but interestingly both lack the catalytic distal histidine residue and the aspartate residue that covalently binds the heme co-factor, features present in the classic MPO and LPO heme peroxidases (Fig. 27.3).

27

Nematode Noxes: The DUOXes of Caenorhabditis elegans

Fig. 27.2 Structure and function of BLI-3. Like other DUOXes, BLI-3 is predicted to reside in the cytoplasmic membrane. The C-terminal half, depicted in blue, consists of the NADPH oxidase domain. The beginning is hydrophobic and contains six transmembrane domains. The second half is homologous to FAD domains with an NADPH binding site. The N-terminal half of BLI-3, depicted in purple, contains the peroxidase domain and one hydrophobic transmembrane domain. Between the peroxidase and NADPH oxidase domains is a cytoplasmic region containing two EF hands predicted to bind calcium. The topology

hMPO hDUOX1 BLI-3/DUOX-2 MLT-7 SKPO-1 HPX-2

258 105 102 268 219 237

LLDHDL HVLSDL VVAYEI FMSHDM FVSHDI FIAHDV

469

of the protein is such that the peroxidase domain is extracellular. The oxidation of NADPH generates electrons that are passed through heme groups located in the transmembrane regions to generate H2O2 extracellularly by a yet unknown mechanism. One established purpose of the H2O2 is the generation of tyrosyl radicals from tyrosine residues in the collagen proteins of the nematode cuticle (yellow). The unstable radicals react with one another to form di and tri-tyrosine cross-links that stabilize this extracellular matrix (ECM) (Figure adapted from [4] and created with BioRender.com)

// // // // // //

403 236 233 394 330 370

DTRSSE AERGNR DSRVNE DIRANL DGRAIL DFRNSL

// // // // // //

500 329 328 491 426 474

YGHTL FLSTM FPHSI FGHGM RLHGM FGHSQ

Fig. 27.3 Comparison of key residues for peroxidase activity. The residues required for hMPO peroxidase activity are highlighted in red (catalysis residues) and blue (heme-binding residues). hDUOX1, BLI-3, MLT-7, SKPO-1, and HPX-2 lack one or more of these residues, but

evidence for some activity exists for all with the exception hDUOX1. See main text for details. Note that the sequences shown for BLI-3 are identical to those in C. elegans DUOX-2

However, BLI-3 retains the glutamate that forms the second covalent bond to the heme and the proximal histidine residue, which may explain why it retains heme binding and peroxidase activity, unlike human DUOX1 [10]. In addition to the peroxidase domain residues important for activity mentioned above, G246 and D392 were discovered to be changed to a D and a N, respectively, in the two alleles of bli-3 originally discovered by Sydney Brenner as described in the introduction [1, 5]. Biochemical characterization of the mutant peroxidase domains revealed that both

point mutations inhibit heme binding, with G246D having a modest effect compared to the complete loss of binding observed for D392N. Loss of heme binding caused a reduction in peroxidase activity, specifically the ability to crosslink tyrosine [11], the reaction necessary for cross-linking of the cuticle as part of its proper formation (Fig. 27.2) [4]. Surprisingly, in a heterologous mammalian expression system, fulllength BLI-3 with a G46D mutation abrogated hydrogen peroxide production, suggesting that this change in the peroxidase domain somehow affects the NADPH oxidase

470

D. A. Garsin

domain [8]. However, this was not observed in the native system; C. elegans mutants naturally containing this allele and the D392N mutation produced about the same amount of ROS under inducing conditions [3, 9]. Additionally, these mutants were not more sensitive to pathogen, unlike animals containing the P1331L mutation in the NADPH oxidase domain [3, 9].

2.2

Expression and Tissue Distribution

In transcriptional studies, expression of the bli-3 gene was seen in all the stages of the worm life cycle, including the embryo, larval and adult [12] (https://wormbase.org/db/get? name=WBGene00000253;class=Gene). Genes involved in cuticle development are often upregulated during the molts between the larval stages, and one study mentions bli-3 having this cyclical pattern of expression consistent with its role in cuticular biosynthesis [13]. A very low level of duox2 expression was also observed at all stages [12] (https:// wormbase.org/db/get?name=WBGene00018771;class= Gene). Using antibody staining, BLI-3 was found to be localized to the hypodermal cells [4]. However, tissue-specific RNAi of bli-3 followed by phenotypic analysis suggested that it was required in the intestine as well as the hypodermis [3]. Using a strain in which the promoter region and the first two exons of bli-3 were fused to mCherry, localization to the hypodermis, the pharynx and the apical membrane of the intestinal cells were observed [9]. Looking at the evidence in total, all three studies support hypodermal localization. Only Edens et al. did not find evidence of intestinal localization [9]. It may not have been observed in this study due to inadequate disruption of the worms during the fixation process preventing α-BLI-3 antibody binding to the internal tissues. Only the fusion construct supports pharyngeal localization, and it is possible that it is artifactual due to an incomplete promoter or overexpression [9]. Unfortunately, the localization cannot be investigated by RNAi as it is not possible to build a tissue-specific RNAi strain that targets just the pharynx [14]. However, the pharynx contains cuticle material [15, 16], and considering BLI-3’s critical role in crosslinking this extracellular matrix it is logical that BLI-3 would be found in this tissue.

3

Biological Roles

3.1

Roles in Development

3.1.1 Cuticle The most obvious visible phenotype caused by knock-down or mutation of bli-3 is “blistering” in which the cuticle peels away from the hypodermis forming a bump (Fig. 27.1)

[1, 4]. The blister has been shown to be filled with cellular debris, unlike some other types of characterized blisters [8]. Many, but not all animals, also appear short or “dumpy” and some have trouble moving in a serpentine fashion because of the unevenness of their cuticles [4]. Similar phenotypes are associated with mutations in the collagen biosynthetic pathway [2, 17]. A critical process to forming a properly structured cuticle is the cross-linking of collagen and other cuticle proteins by di- and tri-tyrosine linkages [2, 17]. Animals exposed to bli-3 RNAi almost completely lack these linkages as measured by HPLC analysis [4], and recombinant BLI-3 was able to crosslink tyrosine, measured by using tyrosine ethyl ester as a substrate [10]. Based on these data, a major functional role of BLI-3 is to cross-link tyrosine in formation of the cuticle during development with the NADPH oxidase domain generating the hydrogen peroxide utilized by the peroxidase domain to form the cross-link.

3.1.2 Vulva One recent study observed that loss of either bli-3 or duox2 inhibits the multi-vulval (muv) phenotype associated with the let-60gf background [7]. As its name suggests, animals with this phenotype develop multiple invaginations associated with vulva development. The study is notable as it is the first to attribute a phenotype associated with the duox2 mutation. Interpretation of how the phenotype occurs was muddled by confusion over the function of the peroxidase domain; it was wrongly stated to have superoxide dismutase activity [7]. Recall that biochemical studies of this domain indicate no superoxide dismutase activity, only peroxidative polymerization of tyrosine [10]. It is noteworthy that the vulva is also lined with cuticular material, and this material must be remodeled during vulva formation [2, 18]. Furthermore, one phenotype that was originally associated with loss of bli-3 was retaining eggs and/or larvae [4], a consequence associated with improper vulva development. Since cuticle remodeling is a known function of BLI-3, I propose that loss of this activity inhibits vulva remodeling and could explain the inhibition of the muv phenotype in the let60gf background. Since loss of duox-2 resulted in the same phenotype, the study suggests that DUOX-2 may also participate in some cuticle remodeling. Perhaps DUOX-2 plays a more specialized, auxiliary, and/or redundant role to BLI-3, since no changes or defects in gross cuticle morphology were observed in a duox-2 deletion mutant [3, 7]. Further studies are needed to investigate the potential roles of BLI-3 and DUOX-2 in the context of vulva development.

3.2

Roles in Immunity

C. elegans produces ROS in response to pathogen, and the ROS is generally protective [3, 9, 19, 20]. Using an Amplex Red assay with live C. elegans, it was observed that a

27

Nematode Noxes: The DUOXes of Caenorhabditis elegans

significant amount of hydrogen peroxide (H2O2) was released when animals were infected with pathogen. The H2O2 release was determined to be arising from the worm, not the microbe, and it appeared to be a protective response as loss of enzymes that breakdown H2O2, such as catalase, rendered the animals significantly more sensitive to pathogen. Because diphenyleneiodonium chloride (DPI) inhibited the ROS release, it was hypothesized that an NADPH oxidase was involved, but the data was not conclusive due to the non-specificity of DPI [20]. In follow-up work, the protective response of ROS production during infection of C. elegans with various bacterial and fungal pathogens was shown to depend on BLI-3 [3, 9, 19]. Because of the severe developmental defects caused by knock-down of bli-3 by RNAi [4], the initial study utilized a partial knock-down of bli-3, resulting in adult animals that could be assayed on pathogen. These animals were shown to produce less ROS when exposed to pathogen, and they succumbed more quickly to infection [3]. Tissue specific RNAi determined that BLI-3 is required in both the intestine and the hypodermis for full protection [3]. Because the RNAi utilized could also target duox-2, animals containing a deletion allele in duox-2 were tested and found to have resistance equal to wild type, demonstrating that the observed phenotypes were due solely to bli-3 [3]. To determine if the NADPH oxidase and/or the peroxidase domain were important for resistance, mutants specific to each domain were tested. Interestingly, animals containing loss of function mutations in the NADPH oxidase domain were more sensitive to pathogen, but those with mutations in the peroxidase domain were not [3, 9]. These data suggest that the oxidase, but not the peroxidase, function of BLI-3 is important for defense against the two pathogens tested, Enterococcus faecalis and Candida albicans [3, 9]. One explanation for the sensitivity of the blistered bli-3 mutants to pathogen could be breaches and/or changes in barrier tissue that physically increase susceptibility. However, blisters form on animals with peroxidase domain mutants as well as NADPH oxidase domain mutants, but only the NADPH oxidase mutants are more sensitive to pathogen [3, 9]. Additionally, as discussed in detail in the functional interactors section, MLT-7, contributes to crosslinking of the cuticle in partnership with BLI-3. mlt-7 RNAi also causes blisters but does not increase pathogen sensitivity [13, 21]. The possibility that another peroxidase partners with BLI-3 to form a more potent oxidant was considered [21, 22], as happens with mammalian NADPH oxidase NOX2, partnering with myeloperoxidase (MPO) to form the potent oxidant hypochlorous acid (HOCl) from H2O2 and Cl- to kill invading microbes (reviewed by [23]). So far, there is no evidence that such reactions occur (see below section on “Functional Interactors). Finally, ROS generated by the NADPH oxidase domain could act as a signaling molecule

471

to generate protective responses. There is significant evidence that this occurs by BLI-3 activating SKN-1, a key transcription factor controlling protective responses involved in immunity, aging and xenobiotic exposure [24–26]. SKN-1 is the C. elegans ortholog of the human Nrf1/ 2 transcription factors and was first studied for its role in embryonic development and its initiation of the mesendodermal fate [27, 28]. However, it was later found to be important in adult animals for optimizing life span and defending against oxidative and xenobiotic stress [29]. Since oxidants induce SKN-1 nuclear localization and transcriptional activity, whether the ROS produced during pathogen exposure would also activate this transcription factor was examined. We and others showed that SKN-1 was necessary for pathogen defense. Loss of the transcription factor rendered the animals more sensitive to pathogen and overexpression increased resistance [24–26]. Nuclear localization and transcriptional activity of SKN-1 on pathogen was shown to require BLI-3 [24].

3.3

Roles in Aging

BLI-3 was also found to affect lifespan in a manner dependent on SKN-1. More BLI-3 activity, caused by overexpression, [30] or by the redox co-factor pyrroloquinoline [31], increased lifespan. When bli-3 was partially knocked-down by RNAi starting at the larval stage, lifespan was reduced [3]. However when it was knocked down only during adulthood, lifespan was not affected [30]. Analysis of hypomorphic alleles revealed that animals containing mutations in the peroxidase domain or the NADPH oxidase domain resulted in a slight and dramatic decreases in lifespan, respectively [9]. Overall, the data suggest that increases in BLI-3 activity increase longevity through activation of SKN-1. However, the negative effects of loss of BLI-3 activity on lifespan are difficult to separate from the negative effects on development. While performing RNAi only on adult animals can seemingly circumvent this problem, controls are needed to ensure that loss of BLI-3 occurs. Since the protein is required during cuticle development in the larval stages it is not clear if RNAi in adulthood results in a significant reduction in the amount of protein present. A major question related to BLI-3’s effects on both immunity and aging is how the ROS produced by this NADPH oxidase feeds into SKN-1 activation. It is known that the p38 MAPK pathway is required for SKN-1 activation; the p38 MAPK, PMK-1, directly phosphorylates SKN-1 on activating residues [32]. A variety of analyses support this pathway being downstream of BLI-3 and other sources of oxidants [24, 26, 32]. Hourihan et al. presented some evidence that ROS generated by BLI-3, the ER, or the

472

D. A. Garsin

mitochondria, sulfenylate a critical cysteine in IRE-1, an endoplasmic reticulum (ER) stress sensing protein. The modification promotes physical association of IRE-1 with p38 MAPKKK, NSY-1, which initiates the p38 MAPK phosphorylation cascade [33]. Whether BLI-3 influences SKN-1 activity by additional mechanisms is not known.

3.4

Roles in Stress Resistance

SKN-1 is important for resistance to harmful agents, many of which generate problematic levels of ROS, including manganese, sodium arsenite, paraquat, and iodide [24, 30, 33– 35]. SKN-1 activation was shown to require BLI-3 in the case of sodium arsenite, but not paraquat [24, 30, 33]. On the other hand, stress resistance to iodide was enhanced by loss of BLI-3, as this chemical caused a dramatic increase in the generation of ROS. In this case, losing another cellular contributor to ROS was beneficial [36]. Manganese neurotoxicity was also enhanced, rather than alleviated by the presence of BLI-3, and it was postulated that BLI-3 promotes the conversion of extracellular dopamine into toxic reactive species in the presence of excessive manganese [35]. Overall, the role of BLI-3 in stress resistance to various chemicals appears to depend very much on the agent. BLI-3 could be required as an amplifying signal with some compounds, as observed with sodium arsenite [33]. However, ROS generation in response to some chemicals occurs at such high levels that the presence of BLI-3 is either no longer necessary to amplify the signal and may in fact be harmful, as was seen with paraquat and iodide [24, 36]. Finally, BLI-3 might be involved in the conversion of a chemical into an even more harmful form, as hypothesized with dopamine in relation to manganese toxicity [35]. An alternative explanation is that there are significant differences in the sub-cellular locales and types of ROS generated by exposure to these chemicals causing SKN-1 activation by different pathways, some BLI-3 dependent, some not, depending on the compound.

4

Regulators

4.1

Tetraspanin/DUOXA Activating Complex

A screen for mutant alleles resulting in the same blistered morphotype characteristic of a bli-3 hypormorphic allele discovered mutations in four genes—bli-3, mlt-7, doxa-1 and tsp-15 [8]. MLT-7 is a heme peroxidase that is discussed in a later section. Doxa-1 encodes a dual oxidase maturation factor, DUOXA1; these factors were shown to be necessary for the correct intracellular trafficking and activation of mammalian DUOXes (See Chap. 14 by F. Miot and X. De Deken) [37]. Tsp-15 encodes a tetraspanin. Tetraspanins are

integral membrane proteins with four transmembrane regions flanked by small and large extracellular loops containing conserved cysteine residues. They are thought to orchestrate association of other proteins into specialized membrane microdomains to carry out a variety of biological functions such as signaling, cell adhesion, membrane fusion and antigen presentation [38, 39]. The similar nature of the tsp-15, doxa-1, and bli-3 phenotypes coupled to co-IP experiments showing that each can interact with the other two suggest they form a complex necessary for BLI-3 activity. Stable transfectants were produced in human cells. When all three proteins were expressed, a high level of hydrogen peroxide was produced, which was inhibited by the commonly accepted NADPH oxidase inhibitor, DPI [8]. Expression of one or two of the proteins had no effect. Interestingly, hydrogen peroxide production was not further stimulated by the addition of known activators of human DUOXes including ionomycin, forskolin and phorbol 12-myristate 13-acetate (PMA), which activate by calcium influx, protein kinase A, and protein kinase C, respectively [8, 40]. Additionally, BLI-3 was still able to traffic to the cytoplasmic membrane in the absence of DUOXA1. Whether this is because BLI-3 is regulated differently compared to the mammalian DUOXes, or the heterologous system was missing negative regulators or otherwise poorly recapitulates the native system remains unclear.

4.2

Small G-Proteins

Mammalian DUOXes are not reported to be regulated by Rac, unlike mammalian NOXes. In general, the mechanisms elucidated for the activation of mammalian DUOXes compared to mammalian NOXes are very different, and details of the factors and pathways involved are described in other chapters of this book (see Chap. 18 by Y. Lin and Y. Zheng). Here, I will focus on reports of small G-proteins activating the C. elegans DUOXes.

4.2.1 CED-10 As mentioned above, one C. elegans developmental phenotype that has been investigated is called muti-vulva (muv), and both BLI-3 and DUOX-2 were shown to impact the muv phenotype in a particular gain-of-function background [7]. Specifically, a mutation of CED-10, resulting in a critical redox-sensing cysteine being changed to a serine, caused an increase in the multi-vulva incidence that was abrogated by the concurrent loss of bli-3 or duox-2. The results are consistent with CED-10 being in a genetic pathway with BLI-3 and DUOX-2, though biochemical and/or structural studies are needed to elucidate whether the genetic interactions are direct. Because CED-10 is a Rac protein, and Rac proteins are reported in the literature to activate some NOX enzymes,

27

Nematode Noxes: The DUOXes of Caenorhabditis elegans

the authors suggested that CED-10 directly activates BLI-3 and DUOX-2. However, in mammalian systems Rac proteins do not activate DUOXes, only certain NOXes, providing an interesting contrast that deserves further investigation [41, 42].

4.2.2 MEMO-1/RHO-1 Another small G-protein in the Rho family, which is not linked to NADPH oxidase activation in mammalian systems, was linked to BLI-3 activation in C. elegans by the study of MEMO-1, a mammalian mediator of ErbB2-driven cell motility homolog that inhibits BLI-3 [30]. Loss of MEMO1 by RNAi or mutation resulted in more ROS production and greater longevity. These effects were found to be mediated by BLI-3 activation of SKN-1 (see below). To understand the mechanism by which MEMO-1 inhibited BLI-3, the authors looked at known mammalian activators of NADPH oxidases, a mix of factors that included those that activate mammalian NOXes and DUOXes. Loss of known DUOX activating proteins such as tetraspanin and DUOXA abrogated the effect of loss of MEMO-1, consistent with the authors’ model. Surprisingly, loss of some small G-proteins, also abrogated the effect. Specifically, the authors demonstrate that RHO-1 directly interacts with MEMO-1 and BLI-3. They postulate that RHO-1 activates BLI-3 but is normally inhibited by MEMO-1 [30]. The strength of this study is that it presents biochemical as well as genetic data to back up this assertion. However, it does not consider how BLI-3, a protein 60–70% identical to the human DUOXes is regulated in such a different manner. It will be of interest for future studies to investigate the differences between NOX/DUOX regulation in model systems like C. elegans, compared to mammalian systems, to better understand the evolution of these NADPH oxidases as well as what they can and cannot tell us about our own biology.

5

Functional Interactors: Peroxidases

While BLI-3 has an enzymatically active heme peroxidase domain, unlike the human DUOXes [10], studies have shown that exogenous peroxidases still functionally interact with BLI-3 and contribute to its function similar to lactoperoxidase (LPO) and thyroperoxidase (TPO) that functionally interact with human DUOXes (reviewed by [43, 44]). In this section, the functionally interacting peroxidases of C. elegans BLI-3 are considered.

5.1

MLT-7

Thein et al. carried out a study to identify components involved in cuticle biosynthesis and discovered that knock-

473

down or mutation of mlt-7 resulted in blistered and dumpy phenotypes very similar to those found following knockdown or mutation of bli-3 [13]. mlt-7 encodes a heme peroxidase that contains the core conserved catalytic and Ca2 + binding residues necessary for activity, including the active site distal and proximal histidines, and the distal arginine (Fig. 27.3). Biochemical analysis of mlt-7 animals indicated a paucity of di-tyrosine linkages [13] similar to bli-3 animals [4]. Utilizing recombinant protein, MLT-7 was demonstrated to have peroxidase activity. Like bli-3, mlt-7 is expressed in the hypodermis. Combining mutant alleles of both bli-3 and mlt-7 increased the severity of the individual phenotypes, in some cases, completely abrogating post-embryonic development. In total, the data support a model in which both MLT-7 and the peroxidase domain of BLI-3 contribute to cuticle morphogenesis by forming di-tyrosine cross-links in collagen utilizing the hydrogen peroxide provided by the BLI-3 NADPH oxidase domain [4, 13].

5.2

SKPO-1

If MLT-7 contributes to BLI-3’s function in cuticular morphogenesis, does it also contribute to BLI-3’s function in innate immunity? To test MLT-7 for effects on immune function, plus 10 other putative peroxidases encoded by the C. elegans genome, an RNAi screen was conducted [21]. Following knock-down of each gene, the animals were exposed to E. faecalis to detect changes in susceptibility. While reduction in expression of mlt-7 had no effect on pathogen resistance, loss of three other genes caused significant changes - hpx-2 and hpx-3, for heme peroxidase, and skpo1, for ShkT-containing peroxidase [21, 45, 46]. ShkT-like domains are associated with extracellular matrix proteins and thought to be involved in protein-protein interactions; MLT-7 and two other putative peroxidases, SKPO-2, and SKPO-3, contain ShkT-like domains. In addition to pathogen sensitivity, the skpo-1 mutant had a slight longevity defect and exhibited morphological changes to the cuticle [21]. While the mutant animals did not have the blistered phenotype associated with bli-3 and mlt-7 mutants, they did have a dumpy phenotype of incomplete penetrance. About half of the worms looked normal, but the rest were slightly to extremely dumpy [21]. The cuticle of skpo-1 animals was more penetrant to dye and the regular structure of ridges (annuli) and furrows that comprise the cuticle was disrupted with these features being irregularly spaced when examined by atomic force microscopy [47]. Consistent with a role in cuticle development, SKPO1 protein was localized to the hypodermis [21]. Transcriptome analysis of skpo-1 animals revealed an upregulation of genes encoding for collagens and proteins related to cuticle and extracellular matrix development

474

D. A. Garsin

[47]. One interpretation is that the increased expression of these genes is a compensatory mechanism for the morphological deficiencies caused by lack of SKPO-1 function. Interestingly, recent studies have found that exposure to some pathogens, including Enterococcus faecalis and Pseudomonas aeruginosa, alters the morphology as well as the barrier function of the cuticle [47, 48]. Concomitant with these morphological changes are alterations in cuticle-related gene expression. Animals crawling through the lawns of these bacteria could experience damage to this tissue and undergo gene expression changes consistent with it being remodeled and/or repaired [47, 48]. The data support a model in which SKPO-1 contributes to this process on E. faecalis [21, 47]. As a putative peroxidase, SKPO-1 was observed to contain the necessary catalytic residues for peroxidase activity (Fig. 27.3). While no successful effort to purify and study the protein in vitro has been reported, one investigation observed that loss of skpo-1 greatly increased the amount of hydrogen peroxide released when animals were exposed to pathogen and that this increase was lost when an NADPH oxidase inhibitor was applied. The results are consistent with a model in which the peroxidase activity of SKPO-1 utilizes much of the hydrogen peroxide produced by BLI-3 during pathogen exposure [21].

5.3

HPX-2

While RNAi of hpx-2 and hpx-3 caused increased susceptibility to pathogens, only deletion mutants in the hpx-2 gene retained this phenotype [21, 45]. Unlike mlt-7 and skpo-1 mutants, hpx-2 animals did not have any visible changes to their cuticle structure by microscopy. However, the animals were more permeable to a DNA dye, consistent with some change to the cuticle [45]. Additionally, genes involved in cuticle synthesis were mis-regulated, but to a lesser degree than in skpo-1 animals [45, 47]. Finally, tissue specific RNAi revealed loss of pathogen resistance when hpx-2 was knocked down in the hypodermis. A partial translational fusion of HPX-2 to green fluorescent protein (GFP) was used for localization. Expression was not visually observed in the hypodermis, which could be due to instability of the fusion construct or the gene having a low expression level as indicated by the transcriptomic data. However, signal was occasionally observed in the pharynx in a pattern consistent with gland cell localization [45]. Interestingly, gland cells excrete material into the pharyngeal lumen [49]. As part of the GI tract, the pharynx is responsible for disrupting incoming microbial food and has teeth-like structures formed from cuticle material [15, 16]. Consistent with a functional role in this process, the hpx-2 mutant was colonized with pathogen to a greater extent than wild type when fed E. faecalis.

Additionally, increased upregulation of pathogen defense genes was observed, which could be related to the heavier bacterial burden [45]. A point mutation was generated that changed the active site arginine to an alanine in the catalytic core of the peroxidase domain (Fig. 27.3), and animals harboring this allele were more sensitive to pathogen indicative of peroxidase function being important to HPX-2’s biological role [45]. Based on HPX-2’s localization pattern and function in cuticle biosynthesis, BLI-3 is postulated to be the main source of hydrogen peroxide for HPX-2 peroxidase activity, but this has not yet been addressed experimentally.

6

Conclusions

The study of nematode DUOXes, primarily of C. elegans BLI-3, has yielded many insights into worm and NOX/DUOX biology. The structural features, expression, biological roles, regulators, and functional interactors of BLI-3 have been reviewed and compared to mammalian/ human DUOXes throughout this chapter. In this final section, open questions in the field worthy of future investigation will be highlighted. Based on the first report of a phenotype associated with loss of the C. elegans duox-2 gene, further study of DUOX2 is warranted [7]. Previously, little work had been done on this gene due to lack of a notable phenotype; the animals had normal morphology by light microscopy and wild type levels of resistance to pathogen [3]. However, it is possible that DUOX-2 contributes to cuticle morphogenesis in a subtle manner, like what was observed with HPX-2 [45]. Careful investigation of cuticle permeability and morphology using more advanced techniques might reveal differences. Investigating genetic interactions with mutations that affect cuticle development could uncover a role for the gene not discernable in the wild type background. Localization studies to observe the timing and pattern of tissue localization may provide further insight into DUOX-2’s biological function, though low expression of its gene could make such studies technically challenging [6]. Of special interest is the emerging evidence that C. elegans DUOXes are regulated differently than mammalian DUOXes. In common, both BLI-3 and the human DUOXes require the maturation factor, DUOXA, for proper activation [8, 37]. However, DUOXA being necessary for DUOX trafficking to the membrane was only observed in the human study. Moreover, human DUOX1 was reported to be regulated by protein kinase A, human DUOX2 by protein kinase C, and both by calcium, requiring functional EF-hands motifs [40, 50]. In contrast, a report using a heterologous system to study BLI-3 reported that neither calcium, forskolin (Fsk), which stimulates protein kinase A, or phorbol 12-myristate 13-acetate (PMA), which stimulates

27

Nematode Noxes: The DUOXes of Caenorhabditis elegans

protein kinase C, were activating [8]. Further supporting fundamental differences between how human and nematode DUOXes are regulated was a screen for regulators of BLI-3 [30]. The small G-protein, RHO-1, was identified in the screen and shown to directly interact with BLI-3. Additionally, small G-proteins RAC-2 and PAK-2 were identified, but whether their effects were also through direct interaction with BLI-3 was not investigated [30]. These data, along with the recent study showing that CED-10 genetically interacts with DUOX-2 [7], suggest that small G-proteins are involved in the activation of the C. elegans DUOXes, unlike their mammalian counterparts [42]. It would be of interest to the field for the regulators identified by Ewald et al. to be further investigated and additional screens for BLI-3 and DUOX2 regulators to be carried out [30]. The field would also benefit from more in vivo studies, as some of the data in support of both human and C. elegans DUOX regulatory mechanisms relied heavily on heterologous expression systems that may not perfectly recapitulate the natural contexts of these enzymes [8, 40]. In conclusion, DUOX regulation in the worm appears quite different from that of humans/mammals. Further investigation is warranted to fully dissect these differences as well as determine if there are commonalities that have been overlooked. Comparisons from diverse systems will continue to provide valuable insights into NOX/DUOX biology. Acknowledgements Melissa R. Cruz is acknowledged for contributing the picture in Fig. 27.1. This publication was supported by the National Institute of Allergy and Infectious Diseases of the National Institutes of Health under award number R01AI150045 to DAG. The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health.

References 1. Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics 77:71–94 2. Page AP, Johnstone IL (2007) The cuticle. WormBook 1–15. https:// doi.org/10.1895/wormbook.1.138.1 3. Chavez V, Mohri-Shiomi A, Garsin DA (2009) Ce-Duox1/BLI-3 generates reactive oxygen species as a protective innate immune mechanism in Caenorhabditis elegans. Infect Immun 77:4983–4989. https://doi.org/10.1128/IAI.00627-09 4. Edens WA, Sharling L, Cheng G et al (2001) Tyrosine cross-linking of extracellular matrix is catalyzed by Duox, a multidomain oxidase/ peroxidase with homology to the phagocyte oxidase subunit gp91phox. J Cell Biol 154:879–891 5. Simmer F, Moorman C, van der Linden AM et al (2003) Genomewide RNAi of C. elegans using the hypersensitive rrf-3 strain reveals novel gene functions. PLoS Biol 1:E12 6. Hill AA, Hunter CP, Tsung BT, et al (2000) Genomic analysis of gene expression in C. elegans. Science (80–) 290:809–812. 7. Kramer-Drauberg M, Liu J-L, Desjardins D et al (2020) ROS regulation of RAS and vulva development in Caenorhabditis elegans. PLoS Genet 16:e1008838. https://doi.org/10.1371/journal.pgen. 1008838

475 8. Moribe H, Konakawa R, Koga D et al (2012) Tetraspanin is required for generation of reactive oxygen species by the dual oxidase system in Caenorhabditis elegans. PLoS Genet 8:e1002957. https://doi.org/ 10.1371/journal.pgen.1002957 9. van der Hoeven R, Cruz MR, Chavez V, Garsin DA (2015) Localization of the dual oxidase BLI-3 and characterization of its NADPH oxidase domain during infection of Caenorhabditis elegans. PLoS One 10:e0124091. https://doi.org/10.1371/journal.pone.0124091 10. Meitzler JL, Ortiz de Montellano PR (2009) Caenorhabditis elegans and human dual oxidase 1 (DUOX1) “peroxidase” domains: insights into heme binding and catalytic activity. J Biol Chem 284:18634– 18643. https://doi.org/10.1074/jbc.M109.013581 11. Meitzler JL, Brandman R, Ortiz de Montellano PR (2010) Perturbed heme binding is responsible for the blistering phenotype associated with mutations in the Caenorhabditis elegans dual oxidase 1 (DUOX1) peroxidase domain. J Biol Chem 285:40991–41000. https://doi.org/10.1074/jbc.M110.170902 12. Harris TW, Arnaboldi V, Cain S et al (2019) WormBase: a modern model organism information resource. Nucleic Acids Res 48:D762– D767. https://doi.org/10.1093/nar/gkz920 13. Thein MC, Winter AD, Stepek G et al (2009) Combined extracellular matrix cross-linking activity of the peroxidase MLT-7 and the dual oxidase BLI-3 is critical for post-embryonic viability in Caenorhabditis elegans. J Biol Chem 284:17549–17563. https:// doi.org/10.1074/jbc.M900831200 14. Shiu PK, Hunter CP (2017) Early developmental exposure to dsRNA is critical for initiating efficient nuclear RNAi in C. elegans. Cell Rep 18:2969–2978. https://doi.org/10.1016/j. celrep.2017.03.002 15. Mango SE (2007) The C. elegans pharynx: a model for organogenesis. WormBook 1–26. https://doi.org/10.1895/wormbook.1.129.1 16. Albertson DG, Thomson JN (1976) The pharynx of Caenorhabditis elegans. Philos Trans R Soc Lond B Biol Sci 275:299–325 17. Lažetić V, Fay DS (2017) Molting in C. elegans. Worm 6: e1330246. https://doi.org/10.1080/21624054.2017.1330246 18. Cohen JD, Sparacio AP, Belfi AC et al (2020) A multi-layered and dynamic apical extracellular matrix shapes the vulva lumen in Caenorhabditis elegans. elife 9. https://doi.org/10.7554/eLife. 57874 19. Jain C, Yun M, Politz SM, Rao RP (2009) A pathogenesis assay using Saccharomyces cerevisiae and Caenorhabditis elegans reveals novel roles for yeast AP-1, Yap1, and host dual oxidase BLI-3 in fungal pathogenesis. Eukaryot Cell 8:1218–1227. https:// doi.org/10.1128/EC.00367-08 20. Chavez V, Mohri-Shiomi A, Maadani A et al (2007) Oxidative stress enzymes are required for DAF-16-mediated immunity due to generation of reactive oxygen species by Caenorhabditis elegans. Genetics 176:1567–1577. https://doi.org/10.1534/genetics.107. 072587 21. Tiller GR, Garsin DA (2014) The SKPO-1 peroxidase functions in the hypodermis to protect Caenorhabditis elegans from bacterial infection. Genetics 197:515–526. https://doi.org/10.1534/genetics. 113.160606 22. McCallum KC, Garsin DA (2016) The role of reactive oxygen species in modulating the Caenorhabditis elegans immune response. PLoS Pathog 12:e1005923. https://doi.org/10.1371/ journal.ppat.1005923 23. Klebanoff SJ (2005) Myeloperoxidase: friend and foe. J Leukoc Biol 77:598–625. https://doi.org/10.1189/jlb.1204697 24. Hoeven R, McCallum KC, Cruz MR, Garsin DA (2011) Ce-Duox1/ BLI-3 generated reactive oxygen species trigger protective SKN-1 activity via p38 MAPK signaling during infection in C. elegans. PLoS Pathog 7:e1002453. https://doi.org/10.1371/journal.ppat. 1002453

476 25. van der Hoeven R, McCallum KC, Garsin DA (2012) Speculations on the activation of ROS generation in C. elegans innate immune signaling. Worm 1:160–163. https://doi.org/10.4161/worm.19767 26. Papp D, Csermely P, Soti C (2012) A role for SKN-1/Nrf in pathogen resistance and immunosenescence in Caenorhabditis elegans. PLoS Pathog 8:e1002673. https://doi.org/10.1371/journal.ppat. 1002673 27. Bowerman B, Eaton BA, Priess JR (1992) Skn-1, a maternally expressed gene required to specify the fate of ventral blastomeres in the early C. elegans embryo. Cell 68:1061–1075. https://doi.org/ 10.1016/0092-8674(92)90078-Q 28. Bowerman B, Draper BW, Mello CC, Priess JR (1993) The maternal gene skn-1 encodes a protein that is distributed unequally in early C. elegans embryos. Cell 74:443–452. https://doi.org/10.1016/ 0092-8674(93)80046-H 29. An JH, Blackwell TK (2003) SKN-1 links C. elegans mesendodermal specification to a conserved oxidative stress response. Genes Dev 17:1882–1893. https://doi.org/10.1101/gad. 11078031107803 30. Ewald CY, Hourihan JM, Bland MS et al (2017) NADPH oxidasemediated redox signaling promotes oxidative stress resistance and longevity through memo-1 in C. elegans. elife 6. https://doi.org/10. 7554/eLife.19493 31. Sasakura H, Moribe H, Nakano M et al (2017) Lifespan extension by peroxidase and dual oxidase-mediated ROS signaling through pyrroloquinoline quinone in C. elegans. J Cell Sci 130:2631–2643. https://doi.org/10.1242/jcs.202119 32. Inoue H, Hisamoto N, An JH et al (2005) The C. elegans p38 MAPK pathway regulates nuclear localization of the transcription factor SKN-1 in oxidative stress response. Genes Dev 19:2278–2283. https://doi.org/10.1101/gad.1324805 33. Hourihan JM, Moronetti Mazzeo LE, Fernandez-Cardenas LP, Blackwell TK (2016) Cysteine sulfenylation directs IRE-1 to activate the SKN-1/Nrf2 antioxidant response. Mol Cell 63:553–566. https://doi.org/10.1016/j.molcel.2016.07.019 34. Xu Z, Hu Y, Deng Y et al (2018) WDR-23 and SKN-1/Nrf2 coordinate with the BLI-3 dual oxidase in response to iodidetriggered oxidative stress. G3 (Bethesda) 8:3515–3527. https://doi. org/10.1534/g3.118.200586 35. Benedetto A, Au C, Avila DS et al (2010) Extracellular dopamine potentiates mn-induced oxidative stress, lifespan reduction, and dopaminergic neurodegeneration in a BLI-3-dependent manner in Caenorhabditis elegans. PLoS Genet 6. https://doi.org/10.1371/ journal.pgen.1001084 36. Xu Z, Luo J, Li Y, Ma L (2014) The BLI-3/TSP-15/DOXA-1 dual oxidase complex is required for iodide toxicity in Caenorhabditis elegans. G3 (Bethesda) 5:195–203. https://doi.org/10.1534/g3.114. 015982

D. A. Garsin 37. Grasberger H, Refetoff S (2006) Identification of the maturation factor for dual oxidase: evolution of an eukaryotic operon equivalent. J Biol Chem 281:18269–18272 38. van Deventer SJ, Dunlock V-ME, van Spriel AB (2017) Molecular interactions shaping the tetraspanin web. Biochem Soc Trans 45: 741–750. https://doi.org/10.1042/BST20160284 39. Hemler ME (2005) Tetraspanin functions and associated microdomains. Nat Rev Mol Cell Biol 6:801–811. https://doi.org/ 10.1038/nrm1736 40. Rigutto S, Hoste C, Grasberger H et al (2009) Activation of dual oxidases Duox1 and Duox2: differential regulation mediated by camp-dependent protein kinase and protein kinase C-dependent phosphorylation. J Biol Chem 284:6725–6734. https://doi.org/10. 1074/jbc.M806893200 41. Buvelot H, Jaquet V, Krause K-H (2019) Mammalian NADPH oxidases. In: Knaus UG, Leto TL (eds) NADPH oxidases methods and protocol. Springer Science+Business Media, New York, pp 17–36 42. Fortemaison N, Miot F, Dumont JE, Dremier S (2005) Regulation of H2O2 generation in thyroid cells does not involve Rac1 activation. Eur J Endocrinol 152:127–133. https://doi.org/10.1530/eje.1.01815 43. Sarr D, Tóth E, Gingerich A, Rada B (2018) Antimicrobial actions of dual oxidases and lactoperoxidase. J Microbiol 56:373–386. https://doi.org/10.1007/s12275-018-7545-1 44. de Faria CC, Fortunato RS (2020) The role of dual oxidases in physiology and cancer. Genet Mol Biol 43:e20190096. https://doi. org/10.1590/1678-4685/GMB-2019-0096 45. Liu Y, Kaval KG, van Hoof A, Garsin DA (2019) Heme peroxidase HPX-2 protects Caenorhabditis elegans from pathogens. PLoS Genet 15:e1007944. https://doi.org/10.1371/journal.pgen.1007944 46. Liu Y (2018) Heme peroxidase HPX-2 protects caenorhabditis elegans from pathogens. The University of Texas MD Anderson Cancer Center UTHealth Graduate School of Biomedical Sciences Dissertations and Theses 47. Liu Y, Martinez-Martinez D, Essmann CL, et al (2021) Transcriptome analysis of Caenorhabditis elegans lacking heme peroxidase SKPO-1 reveals an altered response to Enterococcus faecalis. G3 (Bethesda) 11. https://doi.org/10.1093/g3journal/ jkaa055 48. Sellegounder D, Liu Y, Wibisono P, et al (2019) Neuronal GPCR NPR-8 regulates C. elegans defense against pathogen infection. Sci Adv https://doi.org/10.1126/sciadv.aaw4717 49. Hall DH, Hedgecock EM (1991) Kinesin-related gene unc-104 is required for axonal transport of synaptic vesicles in C. elegans. Cell 65:837–847 50. Ameziane-El-Hassani R, Morand S, Boucher JL et al (2005) Dual oxidase-2 has an intrinsic Ca2+-dependent H2O2-generating activity. J Biol Chem 280:30046–30054

NADPH Oxidases in Arthropods

28

Ana Caroline P. Gandara and Pedro L. Oliveira

Abstract

Arthropods, especially insects, have been used as physiological models for over a century. Once the microbicidal action of reactive oxygen species produced by mammalian phagocytes was discovered, many arthropods were studied and showed adaptative ROS production in immune defense against pathogenic microorganisms, but also in different physiological situations such as cuticle stabilization, muscle contraction, tissue regeneration and fertility, among other biological roles. Among eukaryotes, arthropods comprise the most diverse group, one that is central to the ecology of all types of environments and has three NADPH oxidase (NOX) isoforms: NOX5, dual oxidase (DUOX) and the arthropod-specific NOX4-art. However, studies are still limited to a few model organisms and biological contexts. In this chapter, we will provide an overview of NOXes found in different groups of arthropods, discuss their roles in physiology, and show how this group contributes to elucidating the evolution of NOX. Keywords

NADPH oxidase · Arthropods · Insects · NOX4-art · NOX evolution · NOX5 · DUOX

A. C. P. Gandara (✉) Department of Biochemistry and Molecular Biology, School of Public Health, Johns Hopkins University, Baltimore, MD, USA Present address: Department of Genetics, University of WisconsinMadison, Madison, WI, USA Present address: Morgridge Institute for Research, Madison, WI, USA P. L. Oliveira Instituto de Bioquímica Médica Leopoldo de Meis, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil e-mail: [email protected]

1

Introduction

NADPH oxidase (NOX) enzymes made their initial appearance in the cell biology arena due to the seminal work of Bernard Babior [1], which revealed the microbicidal action of reactive oxygen species (ROS) produced by mammalian phagocytes. This work was followed by the discovery of enzymes with similar activity in a wide range of cell types in vertebrates, but also in invertebrates, fungi, plants, and unicellular eukaryotes [2]. NOXes were soon established as a major source of ROS in eukaryotes, along with mitochondria. Most of the work with mitochondria described production of ROS by this organelle as a potentially costly “side effect” of aerobic metabolism. However, studies with NOXes showed a highly regulated production of ROS, implying that these sophisticated mechanisms might be a consequence of adaptive roles of formation of these molecules—as illustrated in innate immunity studies. In parallel, the conceptual framework for the field of free radical biology shifted from a focus on oxidative damage/ stress counteracted by “good” antioxidant defenses to a more complex picture of the signaling role played by reactive species [3]. The progress made on NOXes was an important part of this large shift in this field. In addition to killing microbes, NOX-derived ROS were implicated by a lot of reports as major players in a myriad of cell functions (discussed in other chapters of this book). This versatility in cell physiology and signaling was complemented by data acquisition regarding the evolution of the NOX family of proteins, showing that virtually all eukaryotic groups have NOX members, a finding that requires a very early origin of the NOX ancestor. The precise storyboard of the evolutionary history of NOXes is not yet clear, with alternative hypotheses suggesting either a EF hand-dependent NOX [4, 5] or an original NOX similar to NOX2 [6] (discussed in this chapter). The phylum Arthropoda (including primarily arachnids, crustaceans, insects and myriapods) accounts for more than two thirds of known eukaryotic species and has undergone

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_28

477

478

A. C. P. Gandara and P. L. Oliveira

incomparable radiation [7]. They were the first animals to leave the sea and develop a terrestrial way of life and currently occupy more ecological niches than any other group, as well as accounting for the largest share of animal biomass. Interestingly, the evolution of the NOX family in this group was marked by a reduction in the NOX paralog repertoire and only NOX5 and DUOX are recognized as ubiquitous among arthropods [5, 8]. A specific NOX4, referred to as NOX4-art, is present in several but not all arthropod species, suggesting the presence of a NOX4 homolog in the arthropod common ancestor, followed by several independent loss events [8]. By no means this streamlined repertoire of NOXes has been associated with reduced relevance to the physiology of these organisms. DUOX is the source of hydrogen peroxide (H2O2) that fuels protein cross-linkage, which is an essential part of cuticle formation in the Ecdysozoa (a superphylum that includes nematodes and arthropods), where it contributes to morphogenesis, cuticle stabilization and waterproofing [9– 12]. The cuticle of terrestrial arthropods is essential to allow respiration in atmospheric air while limiting water loss. Cuticle formation is regulated in a way that provides the astonishing diversity of morphology and mechanical properties that underlies the adaptive plasticity of the arthropod exoskeleton. NOXes also act on microbial killing in hemocytes (arthropod defense cells) and at the gut epithelium-microbiota/ pathogen interface [13–15]. In insect vectors, they modulate the mosquito response to the malaria parasite, in different but complementary ways [16, 17]. Arthropods can be beneficial within a given habitat or act as crop pests and disease vectors. However, they are central to the ecology of all types of environments and, as mentioned above, represent a large fraction of the planet’s animal diversity. We will try to provide a comprehensive view of the NOXes found in different groups of arthropods, discussing their roles in physiology and in the evolution of this unique and essential gene family.

2

DUOX

DUOX was characterized first in porcine and human thyroid plasma membranes [18]. Within a decade, it was described in mosquitos [19], fruit flies [14] and other invertebrates [5]. In arthropods, DUOX is normally found as a single-copy gene [5, 8] but it is duplicated in chelicerates and crustaceans [20, 21]. It is by far the most studied NOX isoform in arthropods, especially in the context of their innate immune system.

2.1

Immune System

DUOX is part of the first line of defense in arthropods. Few years after the publication of Babior’s work, there was

histochemical evidence for a NADH or NADPH oxidase activity, assessed by nitroblue tetrazolium (NBT) and increased by phagocytosis in blood cells of three species of different insect orders [22]. However, this activity could not be attributed to DUOX because of the absence of molecular data.

2.1.1 Arachnida The class Arachnida has examples of all of the arthropod NOX isoforms [8, 20]. As far as we know, the first identification of a NOX in an arachnid was obtained with a cDNA library from a cell lineage derived from embryos of the tick Rhipicephalus microplus, where DUOX transcripts involved in immune responses were identified [23]. Once challenged with Anaplasma marginale bacteria, these cells were downregulated for DUOX and upregulated for antioxidant enzymes, favoring Anaplasma colonization in the host [24]. Interestingly, infection with Rickettsia rickettsii bacteria and heat-killed microorganisms had the opposite effect on expression of these genes. DUOX also participates in the formation of a dityrosine network in the tick Ixodes scapularis that protects it from the bacteria Borrelia burgdorferi [25]. 2.1.2 Crustacea The first reports that characterized a DUOX in this subphylum involved the shrimp Marsupenaeus japonicus, where both DUOX isoforms are highly expressed in the gills and midgut [21, 26]. When they are knocked down, shrimp survival rates decrease when challenged with the bacteria Vibrio spp. or the white spot syndrome virus. The two DUOXes were studied in the crab Scylla paramamosain [27], where the authors found increased expression in hemocytes and hepatopancreas (an organ that plays a key immune role) after challenge with the bacteria Vibrio parahemolyticus or lipopolysaccharides (LPS). Furthermore, DUOX silencing affected the composition of the hemolymph microbiota [27]. 2.1.3 Lepidoptera The first molecular characterization of a NOX in this insect order was carried out with the moth Bombyx mori, where DUOX transcript levels and ROS production were increased after oral infection [28–30]. A DUOX was also studied in other moths [8, 31–35]. DUOX expression increases in the larval gut after oral ingestion of the bacteria Bacillus thuringiensis kurstaki [31, 35], Escherichia coli [32, 35] or the nematode Steinernema feltiae [34], and its knockdown suppresses ROS levels, increasing bacterial pathogenicity [31, 32]. DUOX expression and ROS production are affected by eicosanoid biosynthesis inhibitors in response to bacterial challenge, whereas addition of the prostaglandins PGE2 or PGD2 rescues the inhibition through the cAMP signaling pathway [31, 32]. A detailed characterization of a

28

NADPH Oxidases in Arthropods

lepidopteran DUOX activation after infection was recently done in Manduca sexta [35].

2.1.4 Diptera DUOX has been extensively studied in this insect order due to its role in the microbicidal defense and the control of gut microbiome homeostasis. Mosquitos—In the last two decades, it has been shown that the intestinal microbiota has profound effects on the physiology of most metazoans, including shaping of their immune response, and mosquitos are not exceptions. The mosquito midgut microbiota can influence human pathogens because it modulates vector competence (the capacity of the insect vector to be infected by the pathogen), in large part due to effects on the insect immune responses [36]. The first dipteran in which DUOX was characterized was Anopheles gambiae, where it was shown to modulate the immune response to the malaria parasite Plasmodium berghei. Invasion by this parasite results in Anopheles midgut tyrosine nitration [16, 19]. Proteins of the gut extracellular matrix are cross-linked by dityrosine covalent bonds catalyzed by a secreted peroxidase and DUOX, whose disruption results in increased permeability and activation of epithelial immunity and impairs microbiota balance [16]. While DUOX silencing in P. berghei-infected mosquitos suppresses parasite development through a nitric oxide synthase (NOS)-dependent mechanism [16], this antiplasmodial effect in Anopheles stephensi is mediated also by induction of the TEP1 (a mosquito complement-like protein) pathway [37]. Infection of the vector for yellow fever and dengue viruses Aedes aegypti by the naturally occurring trypanosomatid monoxenic mosquito parasite Strigomonas culicis stimulates superoxide (O2
-) production by midgut mitochondria while infection by an oxidation-resistant parasite strain exacerbates H2O2 production by DUOX activity [38]. Infection of A. stephensi with the fungus Beauveria bassiana causes dysbiosis of gut microbiota, increasing gut bacterial load and decreasing bacterial diversity [39]. It also downregulates antimicrobial peptides (AMPs) and DUOX expression in the midgut, the latter being mediated by secretion of the B. bassiana oosporein toxin. A similar scenario was seen in A. aegypti infected with different species of fungi which increased bacterial loads, and DUOX and DuoxA (mistakenly called DUOX2) gene expressions in the fat body [40]. Eicosanoids mediate the activation of a DUOX in the midgut of Aedes albopictus, after oral infection with Serratia marcensens, initiated by the release of a damageassociated pattern molecule (DSP1) from the gut epithelium into the hemocoel, activating PLA2 [41]. In A. aegypti, ROS are present in the enterocyte plasmatic membrane of sugar-fed mosquitos and these levels decrease (through activation of protein kinase C) once a blood-meal is taken, in parallel with an expansion of gut bacteria

479

[42]. Many factors regulating DUOX activity were already identified in this system: Mesh, a gut membrane-associated protein, regulates DUOX expression and ROS production through an arrestin-mediated MAPK JNK/ERK phosphorylation cascade [43]; serotonin modulates gut microbiome via the regulation of DUOX gene expression and ROS levels [44]; the silencing of a seryl-tRNA synthetase leads to increased ROS production by a DUOX in the midgut due to an expansion of the Bacteroidetes load [45]; and the silencing of heme peroxidase 1 (affecting the peritrophic matrix) exposes the gut epithelia to microbiota-derived immune elicitors, which activates the DUOX oxidative burst [46]. DUOX silencing allows the proliferation of gut microbiota, decreasing mosquito resistance to oral bacterial infection and increasing transcriptional modulation of immune-related genes, epithelial damage, and mosquito mortality [42, 43]. Thus, ROS production by DUOX at the midgut/lumen interface is a ubiquitous feature of mosquito physiology, contributing to defense against both human parasites and mosquito-specific pathogens and playing a pivotal role in the control of mosquito host-microbiota interplay and gut homeostasis. Flies—The first demonstration that gut microbiota was controlled by a DUOX was obtained with the fruit fly Drosophila melanogaster, which is used as a model for insect immune system. G-protein αq-Phospholipase C-β signaling modulates DUOX activity through Ca2+ mobilization and when it is silenced in the midgut of adult flies, there is an increase in the mortality rate after a minor oral bacterial infection [14, 47]. Mesh and serotonin also regulate DUOX activity in flies [43, 44]. During oral infection by the pathogenic bacteria Erwinia carotovora, bacterial peptidoglycan increases DUOX expression through activation of p38-activating transcription factor 2 pathway [48], also inducing peroxiredoxin V through JNK/FOXO signaling, which can control residual ROS to protect the gut against oxidative damage [49]. Later, it was shown that uracil derived from opportunistic pathogenic bacteria, involving uridine catabolism, induces Hedgehog signaling to control the formation of Cadherin 99C-dependent signaling endosomes in enterocytes, activating DUOXdependent ROS production in the gut, causing inflammation and then triggering epithelial cell renewal [50–53]. Also, some nematodes elicit DUOX responses in D. melanogaster mutants for TGF-ß extracellular ligands [54, 55]. Enteric infection induces the reprogramming of gut lipid metabolism, leading to a metabolic shift from an energy-storage to an energy-consumption state, which involves integration of the several different pathways that collaborate to promote DUOX activation [56, 57]. Different insects can have their gut microbiota composition regulated by DUOX-mediated responses to various sources of stress, such as exposure to pathogenic bacteria or

480

A. C. P. Gandara and P. L. Oliveira

insecticides when DUOX is silenced [58–61]. The tsetse fly Glossina morsitans is normally resistant to infection with trypanosomes due to age, maternal-derived trypanolytic molecules, and microbiome status. However, mature refractory flies have high levels of AMP, NOS and DUOX transcripts when challenged with Trypanosoma b. rhodesiense [62]. Collectively, these results indicate that DUOX, together with other components of the innate immune response is a determinant of the so-called vectorial competence.

source of H2O2 for dityrosine-mediated protein cross-linking and eggshell hardening. DUOX knockdown decreased egg numbers and eclosion rates in the hemipteran insect, a Chagas disease vector, Rhodnius prolixus [11]. This diminished egg viability was due to reduced resistance to water loss, adding a new task to the list of NOX functions. Although no assays were done to test it, the predominant expression of DUOX was found in the ovary transcriptome of the tick R. microplus, suggesting a similar role as a source of H2O2 for the cross-linking of the chorion proteins [77].

2.2

2.4

Tissue Regeneration

Dying epithelial cells produce H2O2 that propagates to neighboring cells to activate wound-healing responses, recruiting the first hemocytes to the lesion [63–65]. The caspase Dronc is translocated to the basal side of membrane by Myosin1D to activate DUOX [66]. This extracellular H2O2 recruits hemocytes, activating JNK through the Eiger/Grindelwald TNF system to induce cell proliferation [67]. Diffusion of H2O2 via aquaporin1 in hemocytes is essential for the induction of cytokine upd3 (dependent on the Src42A/Shark/ Draper pathway) and activation of the Toll receptor at the injury site; one of the consequences is protection against systemic infection with the Enterococcus faecalis bacteria [68]. Benzene exposure and sterile injury elicit DUOXdependent ROS production in D. melanogaster larvae, which can lead to lamellocyte differentiation [64, 69, 70]. Ingestion of capsaicin and intestinal expression of α-synuclein induces gut dysplasia through the DUOX–ROS–JNK pathway [71, 72]. In addition, nip (homolog of human DuoxA1, a DUOX maturation factor) is essential for D. melanogaster embryogenesis, affecting larval growth and wing shape [73, 74].

2.3

Cuticle Stabilization

If NOX2 activity in phagocytes can be metaphorically described as a burst, the participation of ROS in the building of the delicate structures of the insect cuticle would be better described as fine welding work, fueled by NOX. Although cuticle morphogenesis is part of the essence of being an arthropod, there is still very little work done on this fascinating subject. The early observations of redox reactions involved in the stabilization of arthropod cuticles via tyrosine crosslinking date from the 1960s [75, 76]. A few decades later, DUOX proved to be the provider of the H2O2 necessary for this process, acting during pupal development so that the adult can develop the correct wing shape [9, 10, 12]. The observation that this mechanism is shared with nematodes suggests a key role for NOXes in the evolution of the molting process. In addition to molting, DUOX was shown to be the

Other Functions

Bacterially-derived uracil helps eliminate pathogens through defecation in D. melanogaster, which requires TrpA1 (present in a subset of midgut enteroendocrine cells) acting as hypochlorous acid (HOCl) receptor and DUOX, involved in the production of HOCl in response to a bacterial infection [78]. In neurons, DUOX mediates sensitivity to heat and mechanical force following UV irradiation [79] and its knockdown extends lifespan [80]. In a Drosophila model of Alzheimer’s disease, DUOX and AMPs gene expressions are increased in aged flies, which can be reverted with the ingestion of β-D Mannuronic acid [81]. DUOX-derived ROS has a role in the resistance to the insecticide clothianidin [82] and are pivotal to dityrosine network formation during tracheal cuticle development in different insects [83, 84]. As mentioned, DUOX in arthropods has many different functions, including in the immune system, modulation of cell proliferation during tissue repair and development, cuticle stabilization and morphogenesis, eggshell hardening, defecation, and nociception. This multipurpose functioning is attributed to what appears to be a unique set of DUOX partners being expressed in each of the different tissues that show DUOX expression.

3

NOX5

NOX5 was first characterized in human spermatocytes and lymphocytes [85], but in insects only few years later [5], although NADPH-dependent superoxide dismutase (SOD)inhibitable respiratory bursts were described in insects in the late 90’s [86, 87]. Superoxide production by salivary gland homogenates (analyzed by luminol and reduced cytochrome c activity assays) from the female mosquito Anopheles albimanus (but not A. aegypti) was thought to be involved with the peroxidation of vasoconstrictor amines released by aggregating platelets, helping the mosquito to probe and suck blood [86]. Using in vitro reduction of NBT, O2
- production that was inhibitable by SOD and diphenyleneiodonium (DPI) was measured in the cockroach Blaberus discoidalis

28

NADPH Oxidases in Arthropods

hemolymph in response to various immune elicitors such as heat-killed E. coli, laminarin, and E. coli LPS, implying a role for hemocytes in antibacterial defense [87]. Even though these seminal papers did not present any genetic characterization, it is plausible to suggest that a NOX5 activity was studied in insects before its characterization in humans.

3.1

Immune System

In addition to the antibacterial defense role that may plausibly be attributed to a NOX5 [87], there are other reports that published similar roles in other insects prior to the molecular characterization of NOX5 [88–91]. However, they should be carefully considered as possible clues to NOX5 activity because we know now that insects do not possess NOX1-3, which depend on phox subunits [8, 92].

3.1.1 Arachnida As far as we know, the only report of a modulation of NOX5 gene and protein expressions in this group was carried out in I. scapularis when infected with the bacteria Anaplasma phagocytophilum, where ROS production by the tick is also inhibited [93]. 3.1.2 Crustacea The first crustacean NOX5 to receive a careful characterization is found in a report on the shrimp M. japonicus [94], where expression was elevated when hemocytes were stimulated with the bacteria Vibrio penaeicida or polyinosinic-polycytidylic acid in vitro, although a demonstration of actual O2
- production was not performed. Other papers showed a NADPH-dependent production of O2
- by isolated hemocytes or an increase in NOX5 expression, activated by microorganism-derived molecules or parasites [95–98]. 3.1.3 Lepidoptera A NOX5 is primarily expressed in the larval P. xylostella gut, especially in response to bacterial challenge from B. thuringiensis kurstaki [8, 31]. An NADPH oxidoreductase was detected in the gut fluid of B. mori and erroneously identified as NOX [99]. 3.1.4 Diptera NOX5 expression increases in the midgut of A. gambiae mosquitos fed with P. berghei-infected blood and its silencing decreases protein nitration in the midgut, allowing the establishment of a greater number of parasite oocysts [17]. The JNK pathway is a key regulator in this signaling, limiting P. berghei infection via heme peroxidase 2 and NOX5, which work together to promote nitration of the

481

malaria parasite; this reaction, in turn, activates the mosquito complement-like protein TEP1 [100, 101]. Ingestion of commensal bacteria, particularly Lactobacillus plantarum, by germ-free D. melanogaster induces ROS-dependent cellular proliferation by NOX5 (and not DUOX) in the midgut, in addition to the activation of the cap’n’collar (CncC) pathway (homolog of Nrf2 in humans), which will upregulate cytoprotective genes against oxidative damage [102, 103]. However, overgrowth of L. plantarum due to a null mutation in PGRP-SD (a pattern-recognition receptor) in the fly gut shortens its lifespan. Excess ROS are produced by NOX5, which is triggered by excess L. plantarum-derived lactic acid, promoting intestinal damage, intestinal stem cell proliferation, and premature intestinal aging [104]. Infection of D. melanogaster S2 cells with DNA virus IIV-6 presumably triggers ROS production by NOX5, activating p38b-unpaireds-JAK/STAT pathway and expression of Turandot genes, a group of poorly characterized peptides that are secreted in response to stress or infection [105].

3.2

Contraction

A distinctive function of NOX5 is to participate in the pro-contractile molecular machinery. First described in D. melanogaster, NOX5 silencing or treatment with DPI causes retention of eggs in the ovaries due to inhibition of smooth muscle contractions that are normally induced in the oviduct by the neuropeptide proctolin [106]. A decade later, ubiquitous NOX5 silencing by RNAi revealed an important role for NOX5 in midgut peristalsis, in the hematophagous R. prolixus [107, 108]. In this insect, peristalsis is also under the control of xanthine dehydrogenase, which works together with NOX5 in the redox regulation of blood digestion [108]. Silencing either of these two genes impairs hemoglobin digestion and, eventually, egg production.

3.3

Tissue Regeneration

In a wounding model, it was shown that injecting actin (mimicking actin released from dying cells) into D. melanogaster hemolymph triggers the JAK/STAT pathway in the fat body, activating NOX5 in this tissue [109]. The hypothesis proposed was that O2
- causes oxidationdependent activation of the kinase Src42A and, eventually, of the tyrosine kinase Shark. In epithelial and photoreceptor cells, Crumbs (a homolog of human CRB1–3) inhibits NOX5-derived ROS and PI3K activation to preserve tissue integrity [110].

482

A. C. P. Gandara and P. L. Oliveira

Various stresses, such as infectious, chemical, and mechanical, can activate two kinases, Ask1 and Licorne (MKK3) to signal enterocytes to promote activation and regeneration of intestinal stem cells. Activation of p38 requires ROS produced by NOX5 in adult D. melanogaster enterocytes [111]. ROS-Ask1-MKK3-p38 signaling is proposed as the pathway involved in gut regeneration but it remains unknown how NOX5 is activated by each stress. Through genetic manipulation of redox genes, it was shown that NOX5-derived O2
- was able to cause cell delamination of the pupal notum of D. melanogaster with activation of caspase-3 [112]. However, H2O2 protects delaminating cells from apoptotic nuclear fragmentation. This led the authors to suggest that extracellular O2
- before and intracellular H2O2 after caspase-3 activation is the fine control that leads to delamination of live cells. Dihydroceramide desaturase (ifc) is the key gene regulating de novo ceramide synthesis in neuronal cells and, when removed from D. melanogaster photoreceptors, promotes dihydroceramide accumulation. Dihydroceramide alters the association of active small GTPase Rac1 to the endolysossome marker Rab7, leading to the ROS production due to an association of Rac1 with NOX5. Morphological and functional degeneration of photoreceptors in ifc-knockout flies can be rescued by apocynin, a NOX inhibitor [113].

3.4

Other Functions

Selenium-dependent glutathione peroxidase (SeGPx) detoxifies organic and hydrogen peroxides, and it was characterized in R. prolixus, being mainly expressed in the fat body, and more closely related to extracellular mammalian SeGPx03. RNAi for SeGPx in first-instar nymphs delays molting and reduces the levels of DUOX and NOX5 transcripts—a reduction in ROS production may compensate for the lack of an antioxidant enzyme such as SeGPx [114]. NOX5 is also essential to molting and oviposition in the hemipteran Nilaparvata lugens [115]. A detailed follow-up of the work indicating that NOX5 controls ovulation through muscle contraction in D. melanogaster [106] showed that NOX5 is highly enriched in follicle cells and is controlled by the neuronal hormone octopamine/OAMB-Ca2+ signaling pathway in the brain [116]. NOX5 silencing in follicle cells leads to defective ovulation due to a reduction in O2
- levels that will be converted to H2O2 by SOD3. The authors suggested that H2O2 is the key signaling molecule for follicle rupture and consequent ovulation also in mammals, in addition to the control of muscular contraction. Knockdown of urate oxidase will shorten adult D. melanogaster lifespan if they are on a high-protein diet,

Fig. 28.1 Phylogenetic tree of NOX proteins. NOX5, DUOX and NOX4-art are the NOX isoforms found in arthropods. NOX4-art is a branch that originates within a non-arthropod NOX4 group (light blue). NOX1-3 are not found in arthropods. Numbers on branches are bootstrap values from 1000 replicates (adapted from [8])

because it will increase uric acid levels, forming crystals in the gut [117]. Interestingly, NOX5 inhibition rescued lifespan, uric acid levels and crystals in urate oxidase silenced flies [117] and NOX5 knockdown was able to reduce uric acid levels in the hemolymph of R. prolixus, reducing their lifespan [108]. Taken together these two reports suggest a crosstalk between NOX5 and purine metabolism/nitrogen excretion by a yet unknown mechanism. Nevertheless, it is clear that NOX5 in arthropods is a pleiotropic signaling protein that has several different roles, including in the immune system, muscle contraction, tissue regeneration, molting, ovulation, and purine metabolism.

4

NOX4-art

This isoform was first described as NOXm [5], where the “m” stands for “mosquito”. The authors concluded that this unique NOX was probably related to hematophagy because it was only found in mosquitos. However, this conclusion was due to the paucity of insect genomic information available at that time (two mosquitos, a fruit fly, and a honeybee). A decade later, a wider search of 95 genomes and 55 transcriptomes made it possible to analyze the evolution of this gene family among arthropods [8]. Both studies found that NOX5 and DUOX subfamilies are present in all groups, but the latter showed that NOXm can also be found in other arthropods besides mosquitos. NOXm was first placed in a branch coming from a root common to all chordate NOX4,

28

NADPH Oxidases in Arthropods

483

Fig. 28.2 Schematic representation of different NOXes present in arthropods. Models showing important loops and domains, based on their primary structures and modeling predictions. All NOXes have the ferric reductase domain with six hydrophobic helices (dark gray) with the four conserved histidine residues. The NOX domain is colored in light blue (FAD1-2) and light yellow (NADPH1-4). Segments and loops are identified by capital letters. Rhodnius prolixus DUOX (RPRC004883) has an extracellular peroxidaselike domain (light green) and two EF hand domains (light orange). Rhodnius prolixus NOX5 (RPRC008329) has four cytoplasmic EF hands (light orange). The asterisk in NADPH1 in Aedes aegypti NOX4-art (AAEL010179) shows where the VXGPFG-motif is located. Homo sapiens NOX4 (AAF68973.1) needs the p22phox subunit (pink) to function properly and it is not present in arthropods

with a weak bootstrap value, making it unclear whether NOXm belonged to the NOX1-4 or fungi NOXA/NOXB subgroups [5]. A larger sample size, showing that NOXm originated from a NOX4-like ancestor (with a >80% bootstrap) (Fig. 28.1) and a detailed analysis of these arthropod homologs justified renaming NOXm as NOX4-art [8]. NOX4-art has all the domains, signature residues and motifs necessary for H2O2 production [8] (Fig. 28.2). A major difference between NOX4 and NOX4-art is the lack of a gene for p22phox, which has not been found in any of the arthropods studied so far [8, 92]. This accessory protein is essential for NOX4 activation and H2O2 production [118]. However, gene silencing reduces H2O2 production, showing that the NOX4-art gene encodes an active enzyme [8]. Even though NOX4/p22phox complex is required for the constitutive production of ROS, activity and stability of NOX4-art may be different, having a unique mode of activation independent of accessory proteins, or one that is dependent on unknown activators. Interestingly, arthropods have homologs for Poldip2 (personal observation, PLO), a protein that also regulates NOX4 activity, but Poldip2 must associate with p22phox [119].

Among arthropods, NOX4-art is the least studied member of the NOX family. There are a few reports on the role of NOX4-art in vivo, all of them indicating a role in immunity. In A. aegypti it was shown that Wolbachia infection increased NOX4-art and DuoxA (mistakenly called DUOX2) transcript levels in midgut and carcass [120]. NOX4-art silencing by RNAi disabled the Toll pathway, suppressing expression of some essential molecules such as Spn27A and SPZ1 that mediate the expression of antioxidants and activate some AMPs. In S. exigua, NOX4-art is expressed in all life stages except eggs, and its expression is increased in larval hemocytes upon injection of E. coli or eicosanoids, followed by ROS production [13] and 6 h after infection with Heliothis virescens ascovirus [121].

5

NOX Evolution

To reach robust conclusions about the evolution of any gene, one of the first steps is to obtain large databases. Projects such as i5k (https://data.nal.usda.gov/project/insects%2D%2Di5k) or public domain databases like VectorBase (https:// vectorbase.org/) and Transcriptome Shotgun Assembly

484

A. C. P. Gandara and P. L. Oliveira

Sequence (https://www.ncbi.nlm.nih.gov/genbank/tsa/) are valuable sources for phylogenetic analyses. Although prokaryotes were believed to have no NOX gene [5, 6], a recent report using a large dataset (162 putative NOX sequences), showed that NOX is also present in Bacteria [122]. The first broad analysis of the evolution of the NOX gene concluded that a calcium-regulated NOX containing an EF hand appeared very early during the evolution of eukaryotes, with the specific regulatory subunits co-evolving with their respective NOXes [5]. Analyzing plants and protists, it was proposed that there was a functional change from a metalloreductase to a ROS-generating protein, linked to the occurrence of EF hands in the N-terminus and referred to as preNOXes [4]. However, a NOX2-like ortholog was found in Choanoflagellida (direct ancestral group of Metazoa), represented by Monosiga brevicollis, as well as in Porifera and Cnidaria [4, 6, 8]. More recently, it was shown that NOX2, NOX5 and DUOX were present in 6 sponge genomes, indicating their likely presence in the last common animal ancestor [123]. As an alternative hypothesis, the phylum Ctenophora is considered the sister-group of Metazoa and only NOX5 was found in the genome of Mnemiopsis leidyi [123]. Whichever the NOX origin is, we cannot state there is a universal NOX isoform in all metazoans. Additional studies using datasets containing more basal groups of animals would be useful to clarify the origin of NOX in Metazoa. Whatever the case, NOX evolution clearly involved several gene duplication events that led to the current collection of families comprising both EF hand-bearing members with calcium-binding capacity (NOX5, NOXC, DUOX, and the plant RBOH families) and those without EF hand domains (NOX1-4 members, NOXA, NOXB, NOXD), where calcium-dependent regulation has been replaced by other regulatory mechanisms such as phosphorylation and regulatory protein partners [5, 92]. However, a remarkable conclusion from phylogenetic analysis was that gene loss was a common follow-up for gene duplication [8, 123]. In summary, gene loss seems to be the main feature in the evolution of the NOX family: (a) NOX5 was lost in the phyla Choanozoa, Cnidaria and Nematoda, and the mammalian order Rodentia; (b) NOX2-like was lost in the Ecdysozoa; and (c) NOX4-art was lost many times within the phylum Arthropoda, including a whole insect order [4–6, 8].

6

Conclusion

The extensive diversification of this ancient group of animals, while providing an immense field for scientific discovery, also limits our ability to reach generalized conclusions, as experiments often focus on a few insect orders, namely

Lepidoptera, Hemiptera and Diptera (especially D. melanogaster, despite relevant work also done in mosquitos). Research in NOX biology will profit from the development of alternative models, made possible by the rapid increase in available genomic information. The simplified repertoire of NOXes in arthropods, evidenced by the lack of several isoforms, contrasts with the diversity of functions that were central to the evolutionary success of this phylum, especially related to molting and cuticle formation, but also involving a variety of other biological functions – possibly with others yet to be discovered. This intriguing fact claims for the discovery of regulatory mechanisms that would add the flexibility needed to explain the fine regulation of development. We anticipate that research in the field will reveal novel partners for the oxygen-reducing catalytic core that will highlight the controlled production of ROS, which is NOX’s unique role in eukaryotic cell biology. These will be either arthropod evolutionary novelties, or they may reveal unknown roles in redox mechanisms also shared with other organisms. Acknowledgments We would like to thank Dr. José Henrique Oliveira for critical review of this chapter and Dr. Martha Sorenson for reviewing the English.

References 1. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanisms. The production by leucocytes of superoxide, a potential bactericidal agent. J Clin Invest 52:741–744. https://doi.org/10. 1172/JCI107236 2. Aguirre J, Lambeth JD (2010) Nox enzymes from fungus to fly to fish and what they tell us about Nox function in mammals. Free Radic Biol Med 49:1342–1353. https://doi.org/10.1016/j. freeradbiomed.2010.07.027 3. Sies H, Berndt C, Jones DP (2017) Oxidative stress. Annu Rev Biochem 86:715–748. https://doi.org/10.1146/annurev-biochem061516-045037 4. Zhang X, Krause K-H, Xenarios I et al (2013) Evolution of the Ferric Reductase Domain (FRD) superfamily: modularity, functional diversification, and signature motifs. PLoS One 8:e58126. https://doi.org/10.1371/journal.pone.0058126 5. Kawahara T, Quinn MT, Lambeth JD (2007) Molecular evolution of the reactive oxygen-generating NADPH oxidase (Nox/Duox) family of enzymes. BMC Evol Biol 7:109. https://doi.org/10.1186/ 1471-2148-7-109 6. Sumimoto H (2008) Structure, regulation and evolution of Nox-family NADPH oxidases that produce reactive oxygen species. FEBS J 275:3249–3277. https://doi.org/10.1111/j.1742-4658. 2008.06488.x 7. Grimaldi DA, Engel MS (2005) Evolution of the insects. Cambridge University Press, Cambridge 8. Gandara ACP, Torres A, Bahia AC et al (2017) Evolutionary origin and function of NOX4-art, an arthropod specific NADPH oxidase. BMC Evol Biol 17:92. https://doi.org/10.1186/s12862-017-0940-0 9. Edens W, Sharling L, Cheng G et al (2001) Tyrosine cross-linking of extracellular matrix is catalyzed by Duox, a multidomain oxidase/peroxidase with homology to the phagocyte oxidase subunit

28

NADPH Oxidases in Arthropods

gp91phox. J Cell Biol 154:879–892. https://doi.org/10.1083/jcb. 200103132 10. Anh NTT, Nishitani M, Harada S et al (2011) Essential role of Duox in stabilization of Drosophila Wing. J Biol Chem 286: 33244–33251. https://doi.org/10.1074/jbc.M111.263178 11. Dias FA, Gandara ACP, Queiroz-Barros FG et al (2013) Ovarian Dual Oxidase (Duox) activity is essential for insect eggshell hardening and waterproofing. J Biol Chem 288:35058–35067. https:// doi.org/10.1074/jbc.M113.522201 12. Hurd TR, Liang F-X, Lehmann R (2015) Curly encodes dual oxidase, which acts with heme peroxidase Curly Su to shape the adult Drosophila Wing. PLoS Genet 11:e1005625. https://doi.org/ 10.1371/journal.pgen.1005625 13. Park Y, Stanley DW, Kim Y (2015) Eicosanoids up-regulate production of reactive oxygen species by NADPH-dependent oxidase in Spodoptera exigua phagocytic hemocytes. J Insect Physiol 79: 63–72. https://doi.org/10.1016/j.jinsphys.2015.06.005 14. Ha E-M, Oh C-T, Bae YS, Lee W-J (2005) A direct role for dual oxidase in Drosophila gut immunity. Science (80–)310:847–850. https://doi.org/10.1126/science.1117311 15. Maria, Shaka Aranzazu, Arias-Rojas Alexandra, Hrdina Dagmar, Frahm Igor, Iatsenko Eric, Oswald (2022) Lipopolysaccharide -mediated resistance to host antimicrobial peptides and hemocyte-derived reactive-oxygen species are the major Providencia alcalifaciens virulence factors in Drosophila melanogaster. PLOS Pathogens 18(9) e1010825-10.1371/journal.ppat.1010825 10.1371/journal.ppat.1010825 16. Kumar S, Molina-Cruz A, Gupta L et al (2010) A peroxidase/dual oxidase system modulates midgut epithelial immunity in Anopheles gambiae. Science (80–)327:1644–1648. https://doi.org/10. 1126/science.1184008 17. Oliveira G d A, Lieberman J, Barillas-Mury C (2012) Epithelial nitration by a peroxidase/NOX5 system mediates mosquito antiplasmodial immunity. Science (80–)335:856–859. https://doi. org/10.1126/science.1209678 18. Dupuy C, Ohayon R, Valent A et al (1999) Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. J Biol Chem 274:37265–37269. https://doi.org/10.1074/jbc.274.52.37265 19. Kumar S, Gupta L, Han YS, Barillas-Mury C (2004) Inducible peroxidases mediate nitration of anopheles midgut cells undergoing apoptosis in response to plasmodium invasion. J Biol Chem 279:53475–53482. https://doi.org/10.1074/jbc. M409905200 20. Palmer WJ, Jiggins FM (2015) Comparative genomics reveals the origins and diversity of arthropod immune systems. Mol Biol Evol 32:2111–2129. https://doi.org/10.1093/molbev/msv093 21. Yang H-T, Yang M-C, Sun J-J et al (2016) Dual oxidases participate in the regulation of intestinal microbiotic homeostasis in the kuruma shrimp Marsupenaeus japonicus. Dev Comp Immunol 59: 153–163. https://doi.org/10.1016/j.dci.2016.01.024 22. Messner B (1976) NADH-bzw. NADPH-oxidase-Aktivität in den Blutzellen hemi- und holometaboler Insekten. Acta Histochem 56: 261–269. https://doi.org/10.1016/S0065-1281(76)80114-6 23. Esteves E, Lara FA, Lorenzini DM et al (2008) Cellular and molecular characterization of an embryonic cell line (BME26) from the tick Rhipicephalus (Boophilus) microplus. Insect Biochem Mol Biol 38:568–580. https://doi.org/10.1016/j.ibmb. 2008.01.006 24. Kalil SP, Da Rosa RD, Capelli-Peixoto J et al (2017) Immunerelated redox metabolism of embryonic cells of the tick Rhipicephalus microplus (BME26) in response to infection with Anaplasma marginale. Parasit Vectors 10:613. https://doi.org/10. 1186/s13071-017-2575-9 25. Yang X, Smith AA, Williams MS, Pal U (2014) A dityrosine network mediated by dual oxidase and peroxidase influences the persistence of lyme disease pathogens within the vector. J Biol

485 Chem 289:12813–12822. https://doi.org/10.1074/jbc.M113. 538272 26. Inada M, Kihara K, Kono T et al (2013) Deciphering of the Dual oxidase (Nox family) gene from kuruma shrimp, Marsupenaeus japonicus: full-length cDNA cloning and characterization. Fish Shellfish Immunol 34:471–485. https://doi.org/10.1016/j.fsi.2012. 11.026 27. Sun Z, Hao S, Gong Y et al (2018) Dual oxidases participate in the regulation of hemolymph microbiota homeostasis in mud crab Scylla paramamosain. Dev Comp Immunol 89:111–121. https:// doi.org/10.1016/j.dci.2018.08.009 28. Hu X, Yang R, Zhang X et al (2013) Molecular Cloning and Functional Characterization of the Dual Oxidase (BmDuox) Gene from the Silkworm Bombyx mori. PLoS One 8:e70118. https://doi. org/10.1371/journal.pone.0070118 29. Zhang L, Wang Y, Lu Z (2015) Midgut immune responses induced by bacterial infection in the silkworm, Bombyx mori. J Zhejiang Univ B 16:875–882. https://doi.org/10.1631/jzus.B1500060 30. Zhang X, Feng H, He J, Liang X, Zhang N, Shao Y, Zhang F, Lu X (2022) Pest Manag Sci 78(6):2215–2227. 10.1002/ps.6846 31. Sajjadian SM, Kim Y (2020) Dual oxidase-derived reactive oxygen species against Bacillus thuringiensis and its suppression by eicosanoid biosynthesis inhibitors. Front Microbiol 11:1–16. https://doi. org/10.3389/fmicb.2020.00528 32. Sajjadian SM, Kim Y (2020) PGE 2 upregulates gene expression of dual oxidase in a lepidopteran insect midgut via cAMP signalling pathway. Open Biol 10:200197. https://doi.org/10.1098/rsob. 200197 33. Marzieh, Attarianfar Azam, Mikani Mohammad, Mehrabadi (2023) The endocrine disruptor fenoxycarb modulates gut immunity and gut bacteria titer in the cotton bollworm Helicoverpa armigera. Comparative Biochemistry and Physiology Part C: Toxicology & Pharmacology 264109507-S1532045622002423 109507 10.1016/j.cbpc.2022.109507 34. Roy MC, Ahmed S, Kim Y (2022) Dorsal switch protein 1 as a damage signal in insect gut immunity to activate dual oxidase via an eicosanoid PGE2. Front Immunol 13:994626. 10.3389/ fimmu.2022.994626 35. Windfelder AG, Müller FH, Mc Larney B, Hentschel M, Böhringer AC, Von Bredow CR, Leinberger FH, Kampschulte M, Maier L, von Bredow YM, Flocke V, Merzendorfer H, Krombach GA, Vilcinskas A, Grimm J, Trenczek TE, Flögel U (2022) Highthroughput screening of caterpillars as a platform to study host– microbe interactions and enteric immunity. Nat Commun 13 (1):7216. 10.1038/s41467-022-34865-7 36. Dennison NJ, Jupatanakul N, Dimopoulos G (2014) The mosquito microbiota influences vector competence for human pathogens. Curr Opin Insect Sci 3:6–13. https://doi.org/10.1016/j.cois.2014. 07.004 37. Kakani P, Kajla M, Choudhury TP et al (2019) Anopheles stephensi dual oxidase silencing activates the thioester-containing protein 1 pathway to suppress plasmodium development. J Innate Immun 11:496–505. https://doi.org/10.1159/000497417 38. Bombaça ACS, Gandara ACP, Ennes-Vidal V et al (2021) Aedes aegypti infection with trypanosomatid Strigomonas culicis alters midgut redox metabolism and reduces mosquito reproductive fitness. Front Cell Infect Microbiol 11:1–17. https://doi.org/10.3389/ fcimb.2021.732925 39. Wei G, Lai Y, Wang G et al (2017) Insect pathogenic fungus interacts with the gut microbiota to accelerate mosquito mortality. Proc Natl Acad Sci 114:5994–5999. https://doi.org/10.1073/pnas. 1703546114 40. José L., Ramirez Christopher A., Dunlap Ephantus J., Muturi Ana B. F., Barletta Alejandro P., Rooney Roberto, Barrera (2018) Entomopathogenic fungal infection leads to temporospatial modulation of the mosquito immune system. PLOS Neglected Tropical

486 Diseases 12(4) e0006433-10.1371/journal.pntd.0006433 10.1371/ journal.pntd.0006433 41. Ahmed S, Sajjadian SM, Kim Y (2022) HMGB1-like dorsal switch protein 1 triggers a damage signal in mosquito gut to activate dual oxidase via eicosanoids. J Innate Immun 14(6):657-672. 10.1159/ 000524561 42. Oliveira JHM, Gonçalves RLS, Lara FA et al (2011) Blood mealderived heme decreases ROS levels in the midgut of Aedes aegypti and allows proliferation of intestinal microbiota. PLoS Pathog 7: e1001320. https://doi.org/10.1371/journal.ppat.1001320 43. Xiao X, Yang L, Pang X et al (2017) A Mesh–Duox pathway regulates homeostasis in the insect gut. Nat Microbiol 2:17020. https://doi.org/10.1038/nmicrobiol.2017.20 44. Zeng T, Su HA, Liu YL, Li JF, Jiang DX, Lu YY, Qi YX (2022) Serotonin modulates insect gut bacterial community homeostasis. BMC Biol 20(1):105. 10.1186/s12915-022-01319-x 45. Silveira GO, Talyuli OAC, Walter-Nuno AB, Crnković A, Gandara ACP, Gaviraghi A, Bottino-Rojas V, Söll D, Polycarpo C (2022) An Aedes aegypti seryl-tRNA synthetase paralog controls bacteroidetes growth in the midgut. bioRxiv 2022.08.25.505225. 10.1101/2022.08.25.505225 46. Talyuli OAC, Oliveira JHM, Bottino-Rojas V, Silveira GO, Alvarenga PH, Barletta ABF, Kantor AM, Paiva-Silva GO, Barillas-Mury C, Oliveira PL (2023) The Aedes aegypti peritrophic matrix controls arbovirus vector competence through HPx1, a heme–induced peroxidase. PLoS Pathogens 10.1371/journal. ppat.1011149 47. Ha E-M, Lee K-A, Park SH et al (2009) Regulation of DUOX by the Gαq-Phospholipase Cβ-Ca2+ pathway in Drosophila gut immunity. Dev Cell 16:386–397. https://doi.org/10.1016/j.devcel. 2008.12.015 48. Ha E-M, Lee K-A, Seo YY et al (2009) Coordination of multiple dual oxidase-regulatory pathways in responses to commensal and infectious microbes in Drosophila gut. Nat Immunol 10:949–957. https://doi.org/10.1038/ni.1765 49. Ahn H-M, Lee K-S, Lee D-S, Yu K (2012) JNK/FOXO mediated PeroxiredoxinV expression regulates redox homeostasis during Drosophila melanogaster gut infection. Dev Comp Immunol 38: 466–473. https://doi.org/10.1016/j.dci.2012.07.002 50. Lee K-A, Kim S-H, Kim E-K et al (2013) Bacterial-derived uracil as a modulator of mucosal immunity and gut-microbe homeostasis in Drosophila. Cell 153:797–811. https://doi.org/10.1016/j.cell. 2013.04.009 51. Lee K-A, Kim B, Bhin J et al (2015) Bacterial uracil modulates Drosophila DUOX-dependent gut immunity via hedgehog-induced signaling endosomes. Cell Host Microbe 17:191–204. https://doi. org/10.1016/j.chom.2014.12.012 52. Lee K-A, Kim B, You H, Lee W-J (2015) Uracil-induced signaling pathways for DUOX-dependent gut immunity. Fly (Austin) 9:115– 120. https://doi.org/10.1080/19336934.2015.1126011 53. Kim E-K, Lee K-A, Hyeon DY et al (2020) Bacterial nucleoside catabolism controls quorum sensing and commensal-to-pathogen transition in the Drosophila gut. Cell Host Microbe 27:345–357.e6. https://doi.org/10.1016/j.chom.2020.01.025 54. Ozakman Y, Raval D, Eleftherianos I (2021) Activin and BMP signaling activity affects different aspects of host anti-nematode immunity in Drosophila melanogaster. Front Immunol 12:795331. 10.3389/fimmu.2021.795331 55. Ozakman Y, Eleftherianos I (2019) TGF-β signaling interferes with the Drosophila innate immune and metabolic response to parasitic nematode infection. Front Physiol 10:716. 10.3389/ fphys.2019.00716 56. Chakrabarti S, Poidevin M, Lemaitre B (2014) The Drosophila MAPK p38c regulates oxidative stress and lipid homeostasis in the intestine. PLoS Genet 10:e1004659. https://doi.org/10.1371/ journal.pgen.1004659

A. C. P. Gandara and P. L. Oliveira 57. Lee K-A, Cho K-C, Kim B et al (2018) Inflammation-modulated metabolic reprogramming is required for DUOX-dependent gut immunity in Drosophila. Cell Host Microbe 23:338–352.e5. https://doi.org/10.1016/j.chom.2018.01.011 58. Yao Z, Wang A, Li Y et al (2016) The dual oxidase gene BdDuox regulates the intestinal bacterial community homeostasis of Bactrocera dorsalis. ISME J 10:1037–1050. https://doi.org/10. 1038/ismej.2015.202 59. Chmiel JA, Daisley BA, Burton JP, Reid G (2019) Deleterious effects of neonicotinoid pesticides on drosophila melanogaster immune pathways. MBio 10. https://doi.org/10.1128/mBio. 01395-19 60. Huang Y, Yu Y, Zhan S et al (2020) Dual oxidase Duox and Tolllike receptor 3 TLR3 in the Toll pathway suppress zoonotic pathogens through regulating the intestinal bacterial community homeostasis in Hermetia illucens L. PLoS One 15:e0225873. https://doi.org/10.1371/journal.pone.0225873 61. Li F, Li M, Zhu Q et al (2021) Imbalance of intestinal microbial homeostasis caused by acetamiprid is detrimental to resistance to pathogenic bacteria in Bombyx mori. Environ Pollut 289:117866. https://doi.org/10.1016/j.envpol.2021.117866 62. Weiss BL, Wang J, Maltz MA et al (2013) Trypanosome infection establishment in the Tsetse fly gut is influenced by microbiomeregulated host immune barriers. PLoS Pathog 9:e1003318. https:// doi.org/10.1371/journal.ppat.1003318 63. Razzell W, Evans IR, Martin P, Wood W (2013) Calcium flashes orchestrate the wound inflammatory response through DUOX activation and hydrogen peroxide release. Curr Biol 23:424–429. https://doi.org/10.1016/j.cub.2013.01.058 64. Evans CJ, Liu T, Girard JR, Banerjee U (2022) Proc Natl Acad Sci 119(12):e2119109119. 10.1073/pnas.2119109119 65. Moreira S, Stramer B, Evans I, Wood W, Martin P (2010) Prioritization of competing damage and developmental signals by migrating macrophages in the Drosophila embryo. Curr Biol 20(5):464470. 10.1016/j.cub.2010.01.047 66. Amcheslavsky A, Wang S, Fogarty CE et al (2018) Plasma membrane localization of apoptotic caspases for non-apoptotic functions. Dev Cell 45:450–464.e3. https://doi.org/10.1016/j. devcel.2018.04.020 67. Fogarty CE, Diwanji N, Lindblad JL et al (2016) Extracellular reactive oxygen species drive apoptosis-induced proliferation via drosophila macrophages. Curr Biol 26:575–584. https://doi.org/10. 1016/j.cub.2015.12.064 68. Chakrabarti S, Visweswariah SS (2020) Intramacrophage ROS primes the innate immune system via JAK/STAT and toll activation. Cell Rep 33:108368. https://doi.org/10.1016/j.celrep.2020. 108368 69. D’Souza LC, Kuriakose N, Raghu SV, Kabekkodu SP, Sharma A (2022) ROS-directed activation of Toll/NF-κB in the hematopoietic niche triggers benzene-induced emergency hematopoiesis. Free Radic Biol Med 193:190-201. 10.1016/j. freeradbiomed.2022.10.002 70. Juarez MT, Patterson RA, Sandoval-Guillen E, McGinnis W (2011) Duox Flotillin-2 and Src42A are required to activate or delimit the spread of the transcriptional response to epidermal wounds in Drosophila. PLoS Genet 7(12): e1002424. 10.1371/ journal.pgen.1002424 71. Liu W, Lim KL, Tan EK (2022) Intestine-derived α-synuclein initiates and aggravates pathogenesis of Parkinson’s disease in Drosophila. Transl Neurodegener 11(1):44. 10.1186/s40035-02200318-w 72. Li Y, Bai P, Wei L, Kang R, Chen L, Zhang M, Tan EK, Liu W (2020) Capsaicin functions as drosophila ovipositional repellent and causes intestinal dysplasia. Sci Rep 10(1):9963. 10.1038/ s41598-020-66900-2

28

NADPH Oxidases in Arthropods

73. Xie X, Hu J, Liu X et al (2010) NIP/DuoxA is essential for Drosophila embryonic development and regulates oxidative stress response. Int J Biol Sci 6:252–267. https://doi.org/10.7150/ijbs. 6.252 74. Khan SJ, Abidi SNF, Skinner A et al (2017) The Drosophila Duox maturation factor is a key component of a positive feedback loop that sustains regeneration signaling. PLoS Genet 13:e1006937. https://doi.org/10.1371/journal.pgen.1006937 75. Andersen SO (1963) Characterization of a new type of crosslinkage in resilin, a rubber-like protein. Biochim Biophys Acta 69:249–262. https://doi.org/10.1016/0006-3002(63)91258-7 76. Locke M (1969) The localization of a peroxidase associated with hard cuticle formation in an insect, Calpodes ethlius stoll, lepidoptera, hesperiidae. Tissue Cell 1:555–574. https://doi.org/10.1016/ S0040-8166(69)80021-2 77. Tirloni L, Braz G, Nunes RD et al (2020) A physiologic overview of the organ-specific transcriptome of the cattle tick Rhipicephalus microplus. Sci Rep 10:18296. https://doi.org/10.1038/s41598-02075341-w 78. Du EJ, Ahn TJ, Kwon I et al (2016) TrpA1 regulates defecation of food-borne pathogens under the control of the Duox pathway. PLoS Genet 12:e1005773. https://doi.org/10.1371/journal.pgen. 1005773 79. Jang W, Baek M, Han YS, Kim C (2018) Duox mediates ultraviolet injury-induced nociceptive sensitization in Drosophila larvae. Mol Brain 11:16. https://doi.org/10.1186/s13041-018-0358-7 80. Baek M, Jang W, Kim C (2022) Dual oxidase, a hydrogenperoxide-producing enzyme, regulates neuronal oxidative damage and animal lifespan in Drosophila melanogaster. Cell 11:2059. https://doi.org/10.3390/cells11132059 81. Barati A, Masoudi R, Yousefi R, Monsefi M, Mirshafiey A (2022) Tau and amyloid beta differentially affect the innate immune genes expression in Drosophila models of Alzheimer’s disease and β- D Mannuronic acid (M2000) modulates the dysregulation. Gene 808:145972. 10.1016/j.gene.2021.145972 82. Zhang Y, Liu C, Jin R et al (2021) Dual oxidase-dependent reactive oxygen species are involved in the regulation of UGT overexpression-mediated clothianidin resistance in the brown planthopper, Nilaparvata lugens. Pest Manag Sci 77:4159–4167. https://doi.org/10.1002/ps.6453 83. Kizhedathu A, Chhajed P, Yeramala L et al (2021) Duox-generated reactive oxygen species activate ATR/Chk1 to induce G2 arrest in Drosophila tracheoblasts. elife 10:1–18. https://doi.org/10.7554/ eLife.68636 84. Jang S, Mergaert P, Ohbayashi T et al (2021) Dual oxidase enables insect gut symbiosis by mediating respiratory network formation. Proc Natl Acad Sci 118:e2020922118. https://doi.org/10.1073/ pnas.2020922118 85. Bánfi B, Molnár G, Maturana A et al (2001) A Ca2+-activated NADPH oxidase in testis, spleen, and lymph nodes. J Biol Chem 276:37594–37601. https://doi.org/10.1074/jbc.M103034200 86. Ribeiro JC (1996) NAD(P)H-dependent production of oxygen reactive species by the salivary glands of the mosquito Anopheles albimanus. Insect Biochem Mol Biol 26:715–720. https://doi.org/ 10.1016/S0965-1748(96)00040-9 87. Whitten MMA, Ratcliffe NA (1999) In vitro superoxide activity in the haemolymph of the West Indian leaf cockroach, Blaberus discoidalis. J Insect Physiol 45:667–675. https://doi.org/10.1016/ S0022-1910(99)00039-6 88. Whitten MMA, Mello CB, Gomes SAO et al (2001) Role of superoxide and reactive nitrogen intermediates in Rhodnius prolixus (Reduviidae)/Trypanosoma rangeli interactions. Exp Parasitol 98:44–57. https://doi.org/10.1006/expr.2001.4615 89. Bergin D, Reeves EP, Renwick J et al (2005) Superoxide production in Galleria mellonella hemocytes: identification of proteins homologous to the NADPH oxidase complex of human

487 neutrophils. Infect Immun 73:4161–4170. https://doi.org/10.1128/ IAI.73.7.4161-4170.2005 90. Renwick J, Reeves EP, Wientjes FB, Kavanagh K (2007) Translocation of proteins homologous to human neutrophil p47phox and p67phox to the cell membrane in activated hemocytes of Galleria mellonella. Dev Comp Immunol 31:347–359. https://doi.org/10. 1016/j.dci.2006.06.007 91. Fallon JP, Reeves EP, Kavanagh K (2011) The Aspergillus fumigatus toxin fumagillin suppresses the immune response of Galleria mellonella larvae by inhibiting the action of haemocytes. Microbiology 157:1481–1488. https://doi.org/10.1099/mic.0. 043786-0 92. Kawahara T, Lambeth JD (2007) Molecular evolution of Phoxrelated regulatory subunits for NADPH oxidase enzymes. BMC Evol Biol 7:178. https://doi.org/10.1186/1471-2148-7-178 93. Alberdi P, Cabezas-Cruz A, Prados PE et al (2019) The redox metabolic pathways function to limit Anaplasma phagocytophilum infection and multiplication while preserving fitness in tick vector cells. Sci Rep 9:13236. https://doi.org/10.1038/s41598-01949766-x 94. Inada M, Sudhakaran R, Kihara K et al (2012) Molecular cloning and characterization of the NADPH oxidase from the kuruma shrimp, Marsupenaeus japonicus: Early gene up-regulation after Vibrio penaeicida and poly(I:C) stimulations in vitro. Mol Cell Probes 26:29–41. https://doi.org/10.1016/j.mcp.2011.11.002 95. Vidya N, Thiagarajan R, Arumugam M (2007) In vitro generation of superoxide anion by the hemocytes of Macrobrachium rosenbergii: possible mechanism and pathways. J Exp Zool Part A Ecol Genet Physiol 307A:383–396. https://doi.org/10.1002/ jez.393 96. Li M, Wang J, Song S, Li C (2016) Molecular characterization of a novel nitric oxide synthase gene from Portunus trituberculatus and the roles of NO/O2--generating and antioxidant systems in host immune responses to Hematodinium. Fish Shellfish Immunol 52: 263–277. https://doi.org/10.1016/j.fsi.2016.03.042 97. Ren X, Lv J, Gao B et al (2017) Immune response and antioxidant status of Portunus trituberculatus inoculated with pathogens. Fish Shellfish Immunol 63:322–333. https://doi.org/10.1016/j.fsi.2017. 02.034 98. Xian J-A, Zhang X-X, Wang A-L et al (2018) Oxidative burst activity in haemocytes of the freshwater prawn Macrobrachium rosenbergii. Fish Shellfish Immunol 73:272–278. https://doi.org/ 10.1016/j.fsi.2017.12.028 99. Selot R, Kumar V, Shukla S et al (2007) Identification of a Soluble NADPH Oxidoreductase (BmNOX) with antiviral activites in the gut juice of Bombyx mori. Biosci Biotechnol Biochem 71:200– 205. https://doi.org/10.1271/bbb.60450 100. Garver LS, de Almeida Oliveira G, Barillas-Mury C (2013) The JNK pathway is a key mediator of Anopheles gambiae antiplasmodial immunity. PLoS Pathog 9:e1003622. https://doi. org/10.1371/journal.ppat.1003622 101. Zhu F, Zheng H, Chen S et al (2022) Malaria oocysts require circumsporozoite protein to evade mosquito immunity. Nat Commun 13:3208. https://doi.org/10.1038/s41467-022-30988-z 102. Jones RM, Luo L, Ardita CS et al (2013) Symbiotic lactobacilli stimulate gut epithelial proliferation via Nox-mediated generation of reactive oxygen species. EMBO J 32:3017–3028. https://doi. org/10.1038/emboj.2013.224 103. Jones RM, Desai C, Darby TM et al (2015) Lactobacilli modulate epithelial cytoprotection through the Nrf2 pathway. Cell Rep 12: 1217–1225. https://doi.org/10.1016/j.celrep.2015.07.042 104. Iatsenko I, Boquete J-P, Lemaitre B (2018) Microbiota-derived lactate activates production of reactive oxygen species by the intestinal NADPH oxidase nox and shortens Drosophila lifespan. Immunity 49:929–942.e5. https://doi.org/10.1016/j.immuni.2018. 09.017

488 105. West C, Silverman N (2018) p38b and JAK-STAT signaling protect against Invertebrate iridescent virus 6 infection in Drosophila. PLoS Pathog 14:e1007020. https://doi.org/10.1371/journal.ppat. 1007020 106. Ritsick DR, Edens W, Finnerty V, Lambeth JD (2007) Nox regulation of smooth muscle contraction. Free Radic Biol Med 43:31– 38. https://doi.org/10.1016/j.freeradbiomed.2007.03.006 107. Montezano AC, De Lucca Camargo L, Persson P et al (2018) NADPH oxidase 5 is a pro-contractile nox isoform and a point of cross-talk for calcium and redox signaling-implications in vascular function. J Am Heart Assoc 7. https://doi.org/10.1161/JAHA.118. 009388 108. Gandara ACP, Dias FA, de Lemos PC et al (2021) Urate and NOX5 control blood digestion in the hematophagous insect Rhodnius prolixus. Front Physiol 12:1–11. https://doi.org/10. 3389/fphys.2021.633093 109. Srinivasan N, Gordon O, Ahrens S et al (2016) Actin is an evolutionarily-conserved damage-associated molecular pattern that signals tissue injury in Drosophila melanogaster. elife 5:1– 25. https://doi.org/10.7554/eLife.19662 110. Chartier FJ-M, Hardy ÉJ-L, Laprise P (2012) Crumbs limits oxidase-dependent signaling to maintain epithelial integrity and prevent photoreceptor cell death. J Cell Biol 198:991–998. https://doi.org/10.1083/jcb.201203083 111. Patel PH, Pénalva C, Kardorff M et al (2019) Damage sensing by a Nox-Ask1-MKK3-p38 signaling pathway mediates regeneration in the adult Drosophila midgut. Nat Commun 10:4365. https://doi. org/10.1038/s41467-019-12336-w 112. Fujisawa Y, Shinoda N, Chihara T, Miura M (2020) ROS regulate caspase-dependent cell delamination without Apoptosis in the Drosophila Pupal Notum. iScience 23:101413. https://doi.org/10. 1016/j.isci.2020.101413 113. Tzou F-Y, Su T-Y, Lin W-S et al (2021) Dihydroceramide desaturase regulates the compartmentalization of Rac1 for neuronal oxidative stress. Cell Rep 35:108972. https://doi.org/10.1016/j. celrep.2021.108972

A. C. P. Gandara and P. L. Oliveira 114. Dias FA, Gandara ACP, Perdomo HD et al (2016) Identification of a selenium-dependent glutathione peroxidase in the blood-sucking insect Rhodnius prolixus. Insect Biochem Mol Biol 69:105–114. https://doi.org/10.1016/j.ibmb.2015.08.007 115. Peng L-Y, Dai Z-W, Yang R-R et al (2020) NADPH oxidase 5 is essential for molting and oviposition in a rice Planthopper Nilaparvata lugens. Insects 11:642. https://doi.org/10.3390/ insects11090642 116. Li W, Young JF, Sun J (2018) NADPH oxidase-generated reactive oxygen species in mature follicles are essential for Drosophila ovulation. Proc Natl Acad Sci USA 115:776–7770. https://doi. org/10.1073/pnas.1800115115 117. Lang S, Hilsabeck TA, Wilson KA et al (2019) A conserved role of the insulin-like signaling pathway in diet-dependent uric acid pathologies in Drosophila melanogaster. PLoS Genet 15: e1008318. https://doi.org/10.1371/journal.pgen.1008318 118. Martyn KD, Frederick LM, von Loehneysen K et al (2006) Functional analysis of Nox4 reveals unique characteristics compared to other NADPH oxidases. Cell Signal 18:69–82. https://doi.org/10. 1016/j.cellsig.2005.03.023 119. Lyle AN, Deshpande NN, Taniyama Y et al (2009) Poldip2, a novel regulator of Nox4 and cytoskeletal integrity in vascular smooth muscle cells. Circ Res 105:249–259. https://doi.org/10. 1161/CIRCRESAHA.109.193722 120. Pan X, Zhou G, Wu J et al (2012) Wolbachia induces reactive oxygen species (ROS)-dependent activation of the Toll pathway to control dengue virus in the mosquito Aedes aegypti. Proc Natl Acad Sci 109:E23–E31. https://doi.org/10.1073/pnas.1116932108 121. Jin R, Xiao Z, Nakai M, Huang GH (2023) Pest Manag Sci 79 (3):1123-1130. 10.1002/ps.7284 122. Hajjar C, Cherrier MV, Dias Mirandela G et al (2017) The NOX family of proteins is also present in bacteria. MBio 8. https://doi. org/10.1128/mBio.01487-17 123. Hewitt OH, Degnan SM (2022) Distribution and diversity of ROSgenerating enzymes across the animal kingdom with a focus on sponges (Porifera). BMC Biol 20(1) 212. 10.1186/s12915-02201414-z

NADPH Oxidases in Zebrafish

29

S. M. Sabbir Alam and Daniel M. Suter

Abstract

Zebrafish, a minnow family teleost fish, has emerged as a powerful in vivo vertebrate model system in biological research. Zebrafish have been used to investigate a number of biological functions mediated by NADPH oxidase (Nox), a transmembrane enzyme known for its role in the production of reactive oxygen species (ROS). Several Nox isoforms such as Nox1, Nox2, Nox4, Nox5, and Duox are expressed in zebrafish. Nox expression and activity change dynamically during zebrafish development and play a role in cell signaling, embryo development, physiology, and pathophysiology. A classical role of Nox is infection control mediated by macrophages and leukocytes, which produce ROS during the oxidative burst. Nox-derived hydrogen peroxide (H2O2) after injury acts as a critical signaling molecule and mediates migration of leukocytes during wound response. Nox also plays a role in several other aspects of zebrafish development and physiology including nervous system development, sensory axon regeneration, thyroid hormone synthesis, and circadian clock control. In addition, zebrafish are used as a model to study Nox function in human disease such as congenital hypothyroidism. Because of its many advantages, zebrafish will likely continue serving as an excellent model system for investigating Nox functions in various biological processes. Keywords

NADPH oxidase · Nox · Duox · ROS · Hydrogen peroxide · Zebrafish

S. M. S. Alam · D. M. Suter (✉) Department of Biological Sciences, Purdue Institute for Integrative Neuroscience, Purdue University, West Lafayette, IN, USA e-mail: [email protected]; [email protected]

1

Introduction

Zebrafish (Danio rerio) is a minnow family (Cyprinidae) tropical freshwater fish belonging to the order of Cypriniformes under the class Actinopterygii (ray-finned fishes) [1]. The word ‘zebrafish’ emanates from the presence of the horizontal blue stripes on the sides of their bodies, which look like the stripes of a zebra. In the wild, zebrafish are found in freshwaters of South Asia, especially in India, Nepal, and Bangladesh [2]. Zebrafish were first used as a research model in the early 1970s by Streisinger at the University of Oregon, who was looking for a model system to study development and function of the nervous system that is simpler than the mouse [3]. Because of its simplicity, high fecundity, development outside the mother, capability of in vivo imaging and genetic manipulations, zebrafish became one of the leading animal model systems in biomedical research over the past twenty years [4]. Furthermore, the availability of the complete genome made zebrafish an effective model filling the gap between invertebrate and higher vertebrate model systems [5]. Zebrafish have a significant physiological and genetic similarity with mammals as shown by the fact that 70% of human gene orthologs are present in zebrafish [5]. Most of the zebrafish genes have mammalian counterparts, and many of the human diseases can be modeled in zebrafish [5, 6]. NADPH oxidases (Nox) are a family of transmembrane enzymes that produce reactive oxygen species (ROS) [7]. Different Nox isoforms have been identified in a variety of species including in animals, fungi, and plants [8]. Nox1, Nox2, Nox4, Nox5, and Duox are expressed in zebrafish [9]. Nox enzymes except Nox4 are not constitutively active, and various stimuli promote translocation of associated proteins such as p47phox, p67phox, and p40phox to activate the Nox2 complex [7]. Each isoform exhibits distinct tissue distribution and activation mechanisms but shares common structural motifs. Nox is known as the only enzyme whose sole function is the production of ROS. Nox mediates

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_29

489

490

S. M. S. Alam and D. M. Suter

production of superoxide by transferring of one electron from cytosolic NADPH through FAD via two hemes to oxygen. Whereas initial studies focused on the function of Nox-mediated ROS production in phagocytic killing of microorganisms [10], follow-up work revealed that Nox enzymes regulate a variety of biological functions. In zebrafish, Duox and Nox2 have been investigated the most among the different Nox isoforms, especially with respect to infection control, injury responses, and regeneration. Furthermore, Nox plays critical roles in several aspects of zebrafish development and physiology including nervous system development [11, 12], cardiac development and functioning [13], maintaining biological clock [14], and thyroid hormone synthesis [15, 16]. This book chapter focuses on the role of different Nox isoforms in development, immunity, host defense, wound healing, and regeneration in zebrafish.

2

Zebrafish as a Model System: A Historical Perspective

Zebrafish has emerged as a research model in 1970s mainly through the work of Streisinger and colleagues at the University of Oregon [3, 17]. At the beginning, most zebrafish research was focused on the understanding of vertebrate embryonic development by taking advantage of in situ hybridization and of forward and reverse genetics [18, 19]. In 1988, techniques were developed to integrate foreign DNA fragments into the zebrafish genome, which could be transmitted to the next generation [20]. About a decade later, the first fish lines expressing fluorescent proteins in a tissue-specific manner were developed

[21, 22]. Zebrafish gets its impetus as a research model in the 1990s, when Nobel laureate Nüsslein-Volhard in Germany and Fishman in the USA used large scale genetic mutagenesis screens to study the regulation of development and organogenesis [23–25]. Zebrafish are now being used to explore many different biological processes utilizing its imaging, genetic manipulation, and high throughput screening benefits [4]. The Zebrafish International Resource Center (ZIRC; http://zebrafish.org) is a comprehensive resource of zebrafish transgenic and mutant fish lines [26]. The Zebrafish Information Network (ZFIN; https://zfin.org) is a database with important information about zebrafish genes, transgenic and mutant lines, expression profiles, and other resources [27]. As of December 2021, ZFIN database includes over 48,500 transgenic lines. Furthermore, the Sanger Institute hosts the whole zebrafish genome information available to the research community (http://www.sanger.ac.uk/Projects/ D_rerio/). The zebrafish model entered the field of redox biology starting in 2005 mostly through studies of respiratory burst, phagocytosis, ROS production in wound healing, pathogen control, and cellular signaling [28–30]. Numerous transgenic lines expressing EGFP, DsRED, mCherry, and other fluorescent proteins in neutrophils, macrophages, endothelial cells, and other cell types along with transgenic lines expressing redox probes became available to investigate oxidative stress in vivo [29, 31–34]. The use of zebrafish as a model to study Nox has strongly increased since 2010 as indicated by the number of publications (Fig. 29.1). In 2004–2010, there were only a few studies on zebrafish Nox; however, publications about Nox in zebrafish increased five-fold during 2010–2015, and then doubled again between 2016 and

Cumulative number of publications

Publications on Nox in Zebrafish 120 100 80 60 40 20 0 2004 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021

Year Fig. 29.1 Increase in publications on Nox in zebrafish as of 2021. Data derived from searching Google Scholar and PubMed databases with keywords ‘Nox AND zebrafish’, ‘NADPH oxidase AND zebrafish’, and ‘Duox AND zebrafish’

29

NADPH Oxidases in Zebrafish

491

2021. Early on, most studies on Nox in zebrafish have been focused on its role in inflammation, wound healing, infection control, and immunity [28, 29, 35, 36]. Development of a respiratory burst assay in zebrafish helped advancing the knowledge of Nox function in ROS production and host immune responses [28]. This was followed by research on toxicology of drugs controlling Nox activity, Nox role in development, cell signaling, circadian clock, and regeneration of the heart, kidney, and nervous system. The recent increase in zebrafish studies on Nox is largely due to the advantages of dynamic ROS imaging using genetically encoded biosensors, transgenic fish lines, and gene editing methods such as CRISPR-Cas9 to study Nox in development, physiology, and pathophysiology.

3

Zebrafish as a Model System to Investigate NADPH Oxidases and ROS

With the advent of genetic and in vivo imaging technologies, zebrafish were recognized as a wonderful vertebrate model organism to investigate a variety of biological processes [37]. All the advantages of zebrafish as a model system for biomedical research also apply to studies on Nox and ROS. In this section, we focus on four of these advantages and how they contributed to our understanding of Nox functions in zebrafish: (1) in vivo live fluorescent imaging, (2) genetic manipulation and generation of transgenic lines, (3) suitability for drug application and screening, and (4) similarity in protein structure and function between zebrafish and human species. Due to their small size, optical transparency, and development outside the mother, zebrafish larvae enable highresolution in vivo live imaging, which is otherwise very difficult for most other vertebrates [37]. The ability of celltype specific labeling in zebrafish larvae has been utilized in several Nox studies. For example, two different fish lines, Tg (mpo:GFP) and Tg(lysC:DsRED2), were employed to study leukocyte recruitment to the tail wound site through a tissue-

scale hydrogen peroxide gradient produced by Duox in epithelial cells [30]. Quantitative imaging of ROS levels in live cells and tissues with high resolution and specificity has been a challenge in the field of redox biology due to the fact that ROS are fast diffusing, small molecules with high reactivity [38]. Both fluorescent dyes and genetically encoded biosensors can be used to monitor general and specific ROS in zebrafish larvae (Tables 29.1 and 29.2). Since zebrafish larvae take up small, membrane-permeable molecules easily, labeling larvae with fluorescent ROS dyes is feasible. One of the most commonly used fluorescent ROS probes is 2′,7′-dihydrodichlorofluorescein diacetate (H2DCF-DA), which produces green fluorescent dichlorofluorescein upon oxidation by different ROS molecules (Table 29.1). For example, H2DCF-DA has been used to measure the respiratory burst in zebrafish kidney and embryos [28]. In order to detect H2O2 near tail wound sites, Rieger and Sagasti used another fluorescent probe, pentafluorobenzene sulphonyl fluorescein (Table 29.1), that is hydrogen peroxide-specific and does not depend on oxidation [32]. Although some fluorescent probes with specificity for certain ROS are available today, there is a need for additional probes that can distinguish between different ROS while providing high signal-to-noise. Penetration capacity, accumulation in particular cell compartments, slow reaction kinetics, auto-oxidation, and lack of ROS specificity have limited the use of fluorescent dyes for in vivo imaging in zebrafish. Genetically encoded ROS biosensors overcome these limitations due to targeted expression, reversible reactions, faster reaction kinetics, and photostability. Several genetically encoded sensors such as HyPer and roGFP have been used to monitor ROS in live zebrafish. These probes have the additional benefit that they are ratiometric and therefore do not suffer from issues with varying expression levels or volume changes affecting the quantification. HyPer consists of a circularly permuted yellow fluorescent protein (YFP), which is inserted into the regulatory domain of E. coli OxyR containing two cysteines [63]. Cysteines in OxyR form a disulfide bond upon reaction with H2O2, which induces a conformational change in YFP

Table 29.1 Commonly used fluorescent probes and genetically encoded biosensors to measure ROS in zebrafish Probe 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA) Pentafluorobenzene sulphonyl fluorescein Dihydroethidium (DHE) CellROX Deep Red Reagent HyPer roGFP2-Orp1 Probes are listed according to their ROS-specificity

Category Fluorescent dye Fluorescent dye Fluorescent dye Fluorescent dye Genetically encoded biosensor Genetically encoded biosensor

Detects H2O2, O2-
, ONOO-, HOCl, peroxides H2O2 O2-
H2O2, O2-
, ONOO-, HOCl, peroxides H2O2 H2O2

References [28, 39–46] [32, 47–49] [50] [51] [30, 52–54] [11, 12, 31, 55]

492

S. M. S. Alam and D. M. Suter

Table 29.2 Transgenic lines and genetic approaches used to study Nox and ROS in zebrafish Transgenic line Tg(ubi:HyPer) Tg(β-actin:Hyper) pku326 Tg(lyz:HyPer) Tg(Kdrl:roGFP2Orp1)uto41 Tg(myl7: roGFP2Orp1)uto40 nox2/cybb pu22/pu22 p22 phox-/- (sa11798) duoxck62a drf Tg(fabp10:GFPDNRac1)lri4 Morpholinos

Purpose Genetically encoded H2O2 biosensor ubiquitously expressed Genetically encoded H2O2 biosensor ubiquitously expressed

References [52] [56]

Genetically encoded H2O2 biosensor expressed in neutrophils Genetically encoded H2O2 biosensor expressed in endothelial cells

[53] [31]

Genetically encoded H2O2 biosensor expressed in myocardiocytes

[31]

lacks functional Nox2 lacks functional p22phox lacks functional Duox1 lacks functional myeloperoxidase Dominant negative Rac1 expressed in hepatocytes

[12] [57] [58] [53] [44]

Splice- and translation-blocking Morpholino oligonucleotides, which target different Nox isoforms

[30, 32, 36, 39, 40, 45, 47–49, 51, 54, 56, 59–62]

These fish lines have been used to image and manipulate ROS

and thereby a shift in the excitation spectrum from 420 nm to 500 nm, which can be detected by emission at 530 nm to measure H2O2 levels. HyPer has been used to visualize a tissue-scale H2O2-gradient that extended 100 μm from the wound margin following transection of the caudal fin in 3 days post fertilization (dpf) zebrafish larvae [30]. The authors demonstrated that this H2O2-gradient attracts leukocytes to the wound site. roGFP was the first genetically encoded redox biosensor, which contains two surface exposed cysteine residues introduced into the β-barrel structure of GFP [64]. These cysteines too form a disulfide bond upon oxidation, which alters fluorescent properties of the GFP. This probe has been further modified into the H2O2-specific biosensor roGFP2Orp1 by attaching the yeast peroxiredoxin Orp1 to roGFP2 [65]. roGFP2-Orp1 has been expressed in zebrafish to monitor H2O2-levels in the nervous system and cardiovascular system [11, 12, 31, 55]. Transgenic fish lines expressing these biosensors either ubiquitously or in specific cell types are listed in Table 29.2. The power of zebrafish genetics has led to the development of several fish lines that express mutant version of different Nox proteins (Table 29.2). Morpholino oligonucleotides have initially been the preferred approach to downregulate protein expression in zebrafish (Table 29.2); however, due to concerns of unspecific toxicity with Morpholinos, the field has moved towards other molecular approaches including CRISPR-Cas9 genome editing in recent years. Weaver et al. (2018) used this approach to create a fish line that lacks functional Nox2/Cybb, which exhibits severe defects in the formation of axonal tracts including retinotectal connections [12]. A p22 phox-/- line was instrumental to demonstrate that neutrophil-derived ROS limits invasive fungal growth [57]. A transgenic line expressing a dominant negative form of Duox1 (DN-Duox1) lacking the

carboxyl-terminal flavin domain has served as a tool to demonstrate the role of ATP in activation of Duox1 and leukocyte recruitment following injury [59]. Another benefit of zebrafish is their suitability for drug treatments and screening due to the easy application of chemical compounds in the media provided [66]. Several studies have applied different Nox inhibitors and antioxidants to larvae as summarized in Table 29.3. Treatment of zebrafish embryos with a Nox-specific inhibitor, VAS2870, at 4 hours post fertilization (hpf) caused a significant delay of epiboly due to reduced cell movements [67]. Nox inhibition of zebrafish embryos between 32 and 36 hours post fertilization (hpf) with celastrol resulted in a wider retinal ganglion cell layer and reduced innervation of the optic tectum [12]. Application of Diphenyleneiodonium (DPI) provided evidence that Nox-derived ROS plays a role in neutrophil and macrophage recruitment to the site of C. albicans infection [39]. The high similarities between zebrafish and human genes make zebrafish an attractive model to study human diseases [5, 6, 77]. Over 71% of all human genes and 82% of human genes associated with a disease have at least one zebrafish orthologue [5]. The Huttenlocher lab has recently used the p22 phox-/-(sa11798) mutant fish line, which is deficient in functional p22phox, as a model for human chronic granulomatous disease (CGD) [57]. They found that loss of p22phox does not limit the hyphal growth of Aspergillus nidulans similar to human CGD patients. Redox sensitive transcription factors are also conserved in zebrafish, such as hypoxiainducible factor (HIF) as well as proteins that regulate its oxidation (such as prolyl-hydroxylase domain (PHDs)containing enzymes) and its degradation (such as von Hippel-Lindau tumor suppressor protein, pVHL) [78]. Zebrafish vhl mutants that are deficient in functional pVHL exhibit systemic hypoxic responses including increased heart rate, cardiomegaly, and increased numbers

29

NADPH Oxidases in Zebrafish

493

Table 29.3 Chemical compounds used to control Nox activity and ROS in zebrafish Compound VAS2870 Celastrol Diphenyleneiodonium (DPI) Apocynin N-acetyl-L-cysteine (NAC) Xyloketal B Puerarin Micrometam C Fucoxanthin

Category Nox inhibitor Nox inhibitor Nox inhibitor Nox inhibitor antioxidant antioxidant antioxidant antioxidant, Nox inhibitor antioxidant

References [12, 14, 30, 52, 67, 68] [11, 12] [14, 39, 41, 44, 47, 54, 56, 59, 67, 69–73] [67, 69, 73, 74] [14, 44, 50, 75] [76] [46] [33] [75]

Different Nox inhibitors and antioxidants are listed with corresponding references

of hematopoietic stem cells and circulating erythroid precursors similar to human polycythemia patients [79]. In summary, as outlined in this section, zebrafish is an excellent model system to investigate functions of Nox proteins and derived ROS in normal biological processes as well as pathophysiology relevant to human diseases.

4

Expression of NADPH Oxidases in Zebrafish

The zebrafish genome encodes five Nox isoforms: nox1, nox2 (also known as cybb), nox4, nox5 and duox [8]. Figure 29.2a shows the relative location of different nox genes on zebrafish chromosomes (data was retrieved from NCBI database). nox2 is located on chromosome 11, nox1 is on chromosome 14, nox4 is on chromosome 15, whereas duox and nox5 are located on chromosome 25. nox3 was not found in the zebrafish genome so far [8]. nox4 although identified in zebrafish is poorly annotated. The presence of a second duox gene in another chromosomal location in zebrafish cannot be completely ruled out considering complex tetraploid genomes of Danio rerio and incomplete genome sequences [8]. Additional subunits such as p22phox, p40phox, p47phox, p67phox, NOXO1, and NOXA1 and several rac genes are also present in zebrafish genome [80]. Zebrafish possess a single duox maturation factor, or duoxA gene, which has been renamed duox2 in NCBI, although it is not a second duox gene. The Duox maturation factor is required for the exit of Duox from the endoplasmic reticulum. Only a few studies have investigated the expression of different Nox isoforms in zebrafish. Weaver et al. (2016) described expression of nox1, nox2, nox5, and duox isoforms during the early development of zebrafish (Fig. 29.2c) [9]. As determined by quantitative PCR, zebrafish nox1 expression is highest at 12 hours post fertilization (hpf). After this time point, expression decreases nearly ten fold and remains at a consistent level from 24 hpf to 48 hpf [9]. At 24–36 hpf, nox1 is expressed in the forebrain, eye, midbrain, hindbrain, and

anterior spinal cord region. At 48 hpf, nox1 expression is slightly elevated in the optic tectum and dorsal region of the spinal cord (Fig. 29.2c) [9]. These findings are consistent with a recent large mRNA expression study conducted by White et al. (2017) (Fig. 29.3), which revealed high nox1 expression during gastrula and early segmentation period, followed by a decline and rise again at 4 and 5 days dpf [81]. nox2 expression is more stable during the first two days of zebrafish development when compared to other Nox isoforms [9]. nox2 is expressed in the eye, forebrain, hindbrain and spinal cord region (Fig. 29.2c). A slight increase is observed in the optic tectum at 48 hpf [9]. The mRNA expression time course study by White et al. (2017) revealed high expression of nox2 from the gastrula to the segmentation (1–4 somites) stage, as well as at 4 and 5 dpf (Fig. 29.3) [81]. Whereas a nox4 gene sequence has been identified in the zebrafish genome, no expression data are available for zebrafish nox4 thus far. Zebrafish nox5 is expressed at significantly lower levels when compared to other Nox isoforms early in development, being highest at 12 hpf (Figs. 29.2c and 29.3) [9, 81, 83]. nox5 is broadly expressed in the nervous system at 24 and 36 hpf. At 48 hpf, a slight increase has been observed along the anterior posterior axis from the forebrain to the spinal cord (Fig. 29.2c) [9]. duox is widely expressed during zebrafish development including in intestinal epithelia, ultimobranchial bodies, epidermis [36, 84], nervous system, eye [9], kidney [47]. At 80 hpf, duox expression was also detected in thyroid tissue [84, 85]. duox expression is lower at 12 hpf when compared to nox2, and then expression increases by >10-fold at 36 hpf [9]. A relatively broad expression of duox has been found in the head and spinal cord at 24 and 36 hpf [9]. At 48 hpf, duox expression levels are increased on the dorsal side of tectal ventricles and the spinal cord [9]. White et al. (2017) reported higher duox expression levels from the gastrula to the hatching stage (Fig. 29.3) [81]. In conclusion, nox1, nox2, nox5, and duox are expressed relatively broadly during zebrafish development with certain isoforms exhibiting higher expression in specific regions of the larvae.

494

S. M. S. Alam and D. M. Suter

A

Nox5 Nox2 /Cybb

B

Phosphorylation Calcium GTP

Duox Nox1

Ch 11

Nox4

Ch 25

Ch 15 Ch 14

C

Nox1

Nox2

Nox5

Duox

24 HPF Fig. 29.2 Isoforms and expression pattern of Nox in zebrafish. (a) Relative location of nox1, nox2, nox4, nox5, and duox genes on zebrafish chromosomes. (b) Schematic of Nox activation complexes for Nox1, Nox2, Nox5, and Duox. Nox1 and Nox2 are activated by cytosolic subunits, whereas Nox5 and Duox are activated by calcium. Nox 4 (not shown) is constitutively active. (c) Broad expression of nox1, nox2, nox5, and duox during the first two days of zebrafish development. Figure shows in situ hybridization of whole mount zebrafish embryos

5

Biological Functions Regulated by NADPH Oxidases in Zebrafish

5.1

Nox in Zebrafish Development

Different Nox isoforms have been implicated in a variety of biological functions including development, wound healing, regeneration, antimicrobial defense, cancer, and circadian clock (Fig. 29.4). Recent evidence suggest that zebrafish embryos undergo major changes in H2O2-levels during

Nox1

Nox2

Nox5

Duox

48 HPF and cryosections with labeled riboprobes against nox1, nox2, nox5 and duox mRNA. Position of the sections is indicated in the top panels. Abbreviations: DC diencephalon, e/E eye, f forebrain, h hindbrain, m midbrain, MyC myelencephalon, MC mesencephalon, NC notochord, ON optic nerve, OV optic vesicle, SC/s spinal cord, SM somites, Tec tectum, TC telencephalon, TecV tectal ventricle, 3V third ventricle. Scale bar = 0.2 mm in whole mounts, and 100 μm in sections. Fig. (b–c) is adopted from [9] with permission by Wiley

early development [52]. Gauron et al. (2016) used the genetically encoded biosensor HyPer to monitor H2O2 during the first 3 dpf [52]. They showed that H2O2-levels significantly increase during gastrulation with a peak around 24 hpf, then decline again until 48 hpf with the exception of the heart and the nervous system. Elevated H2O2 was detected in the optic tectum, notochord, and the spinal cord region at 48 hpf. An inverse relationship between H2O2-levels and catalase activity, an enzyme that degrades H2O2, suggested that ROS degradation rather than production might control H2O2-levels during the first 2 days of zebrafish development.

29

NADPH Oxidases in Zebrafish

495

nox5 nox1 nox2/cybb duox Fig. 29.3 Expression of nox1, nox2/cybb, nox5, and duox mRNA during different stages of zebrafish embryonic development. During gastrula, duox is highly expressed when compared to other nox

isoforms. Expression levels are indicated as transcripts per million (TPM). Data source: Expression atlas [81, 82]

Furthermore, H2O2 depletion and Nox inhibition impaired projections of retinal ganglion cell (RGC) axons in the optic tectum [52]. Interestingly, ectopic activation of the sonic hedgehog pathway or H2O2 rescue could restore normal axonal projections, implicating sonic hedgehog as a potential downstream effector of Nox-derived H2O2 signaling [52]. A functional role for Nox-derived ROS during gastrulation was demonstrated by Mendieta-Serrano and colleagues [67]. Using H2DCF-DA-labeling and spinning disk confocal microscopy, they showed a dynamic pattern of ROS during epiboly with high ROS signals presenting as a ring-like structure at the migrating epiboly front. ROS was detected mostly in the interstitial space. Nox inhibitors reduced ROS,

caused delayed epiboly, reduced cell movements, F-actin cytoskeleton and E-Cadherin amounts at cell contacts, which could be rescued by the addition of H2O2. The authors concluded that Nox-derived ROS regulate cell movements during epiboly by controlling E-cadherin localization at the plasma membrane [67]. NADPH oxidases and their derived ROS have significant roles in regulating nervous system development, especially neurogenesis, neuronal differentiation, axonal growth and guidance as shown in different species [86–88]. In vitro studies conducted with rat hippocampal and Aplysia bag cell neurons revealed that Nox activity is required for growth cone motility, neurite outgrowth, and neuronal polarity [89–

Fig. 29.4 Major Nox functions in zebrafish. Nox isoforms have been implicated in a number of biological processes including infection control and immunity, wound healing, regeneration, cell signaling, development, circadian clock, tumorigenesis, and ROS balance

Infection control and immunity

Wound healing

Regeneration

Cell signaling

Development

Circadian clock

Redox balance Tumor

496

91]. In order to better understand the function of Nox in axonal growth and guidance in vivo, we took advantage of the zebrafish model system. In zebrafish, nox1, nox2, nox5, and duox exhibit a relatively broad expression pattern in the brain and spinal cord during the first two days of development [9]. However, ROS levels could also be controlled in specific parts of the body by either altering Nox activity through external stimuli, or the activity of redox defense systems such as thioredoxin, peroxiredoxin, glutathione as well as enzymes that degrade hydrogen peroxide such as catalase [52, 92]. In conclusion, both in vivo ROS imaging and expression analysis of different Nox isoforms have not been extremely informative for determining the precise steps in nervous system development that are regulated by Nox. Functional studies using Nox inhibitors and molecular approaches were needed to determine Nox function during neuronal development. Several studies conducted with rodents have provided evidence that Nox enzymes are critical for maintaining the neural stem cell pool and neurogenesis in both embryonic and adult brain [93–96]. A more detailed discussion of Nox isoforms regulating embryonic and adult neurogenesis can be found in a recent review from our laboratory [88]. Does Nox function in neural stem cell maintenance and neurogenesis in zebrafish? We have reported that the formation of the retina cell layers is affected in nox2-/- zebrafish larvae [12]. Specifically, the width of the retinal ganglion cell layer is enlarged in these Nox2-deficient fish when compared to wild type fish. Nestin (a marker for neural progenitor cells) expression was not significantly altered in these mutant fish. This might suggest that early steps of neural development are not affected in these nox2-/- zebrafish; however, we did not quantify the number of neural progenitor cells in the nox2-/- zebrafish. Thus, more investigations are needed to determine whether Nox2 functions in neurogenesis in zebrafish. A role for redox signaling in retinal progenitor cell proliferation vs. differentiation in zebrafish was shown by Albadri et al. (2019), although the source of H2O2 was not investigated in this study [97]. They found that the lipid peroxidation product 9-hydroxystearic acid (9-HAS) increases proliferation of embryonic and postembryonic retinal progenitor cells via Notch and Wnt activation. Furthermore, overexpression of catalase, which degrades H2O2, was able to trigger premature RGC differentiation, further supporting the idea that H2O2-levels during development may be controlled by degradation rather than by production. Nox2 controls the formation of retinotectal connections in the developing zebrafish brain [12, 52]. Both of these studies demonstrated that the area of the optic tectum innervated by RGC axons was diminished in embryos treated with Nox inhibitors VAS2870 and celastrol—an effect which could be rescued by H2O2 addition. These defects in tectal innervation were also confirmed in nox2-/- zebrafish [12]. We

S. M. S. Alam and D. M. Suter

detected additional abnormal axonal projections in the forebrain and spinal cord. Both in the embryonic retina and growth cones of cultured RGC neurons, H2O2 levels were not significantly different between nox2-/- and wild type zebrafish, which could be due to compensatory effects by other Nox isoforms. Using cultured RGCs from nox2-/and wild type zebrafish embryos, we also investigated whether specific axonal guidance cues act upstream of Nox2 in order to control retinotectal innervation [11]. Specifically, we showed that Nox2 activation is required for slit2mediated axonal retraction and guidance of RGC axons in vitro. These findings provided strong evidence for a cellautonomous role of Nox2 in the signal transduction pathway downstream of the repulsive guidance cue slit2 and its receptor robo2. Evidence for a shared pathway involving both robo2 and Nox2 in tectal innervation by RGC axons was also found in vivo by comparing the tectal innervation phenotypes of nox2- and robo2-deficient larvae [11]. Duox has a role in zebrafish development because it is involved in the synthesis of thyroid hormone. Iodide is taken up from the blood and extracellular fluid by the thyrocytes followed by oxidation via Duox-derived H2O2 [98]. Two nonsense mutant duox alleles (sa9892 and sa13017) resulted in congenital hypothyroidism phenotype in zebrafish including thyroid hormone synthesis failure, goiter (enlarged thyroid gland), growth retardation, pigmentation defects, and infertility [15]. These findings were confirmed using a duox-targeted CRISPR-Cas9-based genome editing approach [16, 58]. Thus, these zebrafish mutants could serve as valuable disease models of human congenital hypothyroidism. NADPH oxidases are expressed in several different cells of the cardiovascular system and implicated in cardiac development, blood pressure regulation, and cardiovascular function [99]. Razaghi and colleagues showed that loss of hace 1 (HECT domain and Ankyrin repeat Containing E3 ubiquitin-protein ligase 1) by genetic intervention induces overactivity of Rac1 (a small GTPase participating in the activation of Nox1 and Nox2) and causes abnormalities of the atrium, ventricle, and atrioventricular junction. These malformations can be rescued by pharmacological and genetic inhibition of nox1 and nox2, which suggests a possible role of Nox in normal cardiac development in zebrafish [13]. Taken together, all Nox isoforms have been implicated in the development of different organ systems in zebrafish including eyes, brain, thyroid gland, and heart.

5.2

Nox in Zebrafish Wound Healing and Regeneration

In previous sections, we have highlighted the many advantages of zebrafish as a model system to investigate the molecular and cellular mechanisms of normal and

29

NADPH Oxidases in Zebrafish

497

tail fin cut Duox

[H2O2]

Neutrophil migration

Lyn kinase

Fig. 29.5 Tail fin wounding recruits neutrophils via H2O2-gradient. The wound generates a Duox-mediated H2O2-gradient that attracts neutrophils to the wound site [30]. Lyn kinase acts as the redox target in neutrophil migration to the wound site [34]

pathophysiology. An additional benefit of using zebrafish is its exceptional capacity to regenerate various tissues and organs including fins, retina, spinal cord, and heart throughout adulthood. Due to its simple structure and relatively transparent nature, the tail fin provides a convenient tissue to study vascular wounding and regeneration [100, 101]. Complete regeneration of the tail fin happens within two weeks of injury. Furthermore, regeneration capacity seems to be unlimited, as it was shown that tail fins full regenerate even after 29 consecutive amputations [102]. Niethammer and colleagues used HyPer to investigate the role of hydrogen peroxide in the initial wound response following a tail fin cut in 3 dpf zebrafish larvae [30]. They found that a tissue-scale H2O2 gradient developed over 100 μm from the cut site within the first 20 min of injury. This gradient was required for neutrophil recruitment (Fig. 29.5). Pharmacological Nox inhibition and Morpholino-mediated duox knockdown provided evidence that Duox activity is necessary for both wound margin H2O2-production and leukocyte recruitment [30]. How can this long-range signaling gradient be established considering the high concentration of antioxidant systems such as peroxiredoxin? Using HyPer imaging along with a computational modeling approach, Niethammer’s group later has shown that the H2O2-gradient can extend deep into the tissue but overcome antioxidant barriers only within around 30 μm of the wound margin [103]. What are the downstream effector proteins of H2O2mediated neutrophil recruitment to the wound site? Redoxsensitive cysteines are common targets of H2O2-signaling for example by inactivation of tyrosine phosphatases or activation of tyrosine kinases [104]. The Huttenlocher laboratory identified the Src family kinase Lyn as an important redox target during neutrophil recruitment [34]. The same group also demonstrated that H2O2-induced activation of another

Src family kinase member, Fynb, in epithelial cells is critical for fin regeneration [105]. A different Src family kinase, Yrk, is critical for macrophage recruitment and migration of neutrophils away from the edge of a wound during the process of fin regeneration [60]. In conclusion, Nox-derived ROS and different members of the Src family kinase are involved in the initial wound response and following regeneration after a tail fin injury. Taking advantage of myeloperoxidase-deficient zebrafish, it was also demonstrated that myeloperoxidase in neutrophils downregulates the H2O2-levels caused by wounding [53]. Furthermore, zebrafish have an amazing capability to regenerate their tail fins during adulthood [106]. Interestingly, sustained production of ROS 10–12 hours post injury in adult zebrafish is required for blastema formation via apoptosis and Jnk activation during regeneration [107]. Thus, ROS may have a different role in initial wound healing and regeneration based on the different time scale of these ROS signals. A very recent study showed that homozygous duox-/- mutants exhibit significantly reduced adult tail fin regeneration that is correlated with reduced ROS levels [108]. Adult zebrafish have also been used as a model system for heart regeneration [106]. Following heart injury, pre-existing cardiomyocytes dedifferentiate to generate new cardiac muscle. H2O2-levels increase near the injury site but on a slower time scale compared to tail fin regeneration [109]. Nox-derived H2O2 is required and sufficient to mediate cardiac regeneration through destabilizing the redoxsensitive phosphatase Dusp6, which results in Erk1/2 activation. A role for Duox-derived H2O2 in promoting kidney repair and regeneration in zebrafish has also been reported [47]. Nox-derived ROS also play a critical role in axonal regeneration in zebrafish. Rieger and Sagasti demonstrated that H2O2 promotes regrowth of peripheral sensory axons after injury [32]. First, they showed that tail amputation increased axonal growth in the tail fin of 3 dpf larvae. Next, they provided evidence that Duox-derived H2O2 produced by tail amputation promotes axonal regeneration. Lastly, keratinocyte ablation and chimeric transplantation studies suggested that H2O2 produced by keratinocyte promotes axonal regeneration [32]. Duox is also critical for axonal regeneration in the central nervous system. CRISPR-Cas9generated duox null mutant larvae exhibited reduced axonal regeneration in Mauthner cells, which are large hindbrain neurons responsible for the fast escape reflex [110]. These duox null mutant larvae had reduced expression levels of several genes that are critical for mitochondrial function such as mitofusin-2. Furthermore, mitochondrial size was increased and transport dynamics was reduced in these duox mutants, suggesting that reduced axonal regeneration in Duox-deficient larvae could be due to altered mitochondrial

498

S. M. S. Alam and D. M. Suter

dynamics [110]. In conclusion, there is accumulating evidence that Nox-derived H2O2—especially generated by Duox—has a critical role in regulating tissue regeneration in both larval and adult zebrafish.

5.3

Nox in Antimicrobial Defense in Zebrafish

The first well-characterized function of Nox enzymes is the killing of bacteria by Nox-derived superoxide produced by the phagocytic leukocyte during the respiratory burst [111]. Hohn and Lehrer’s work provided the first definitive evidence of Nox involvement in generating superoxide using the CGD model [112, 113]. Due to the ability of cell typespecific labeling and genetic manipulations, several studies have taken advantage of the zebrafish model system to investigate the role of neutrophils and macrophages in antimicrobial defense and inflammation [114–118]. Neutrophils and macrophages are the key cells of the innate immune systems acting as first responders to tissue injury and infection [114, 116]. In addition to phagocytosing microbes and cellular debris, they accomplish a multitude of other functions to control infection and inflammation including secretion of cytokines and formation of neutrophil extracellular traps. The first investigation of the phagocytic respiratory burst conducted in zebrafish embryos was reported by Hermann and colleagues in 2004 [28]. These authors used H2DCF-DA to visualize phorbol ester-induced ROS production to study the respiratory burst in macrophages and neutrophils in zebrafish kidney and embryos (Table 29.1). In zebrafish embryo, neutrophils develop in the caudal hematopoietic tissue until 4 dpf, when the kidney starts to become the main site of hematopoiesis [119]. Zebrafish kidney leukocytes produce ROS in a rhythmic pattern with a maximum of cellular ROS before dawn, whereas phagocytosis of E. coli peak late in the day [29]. These findings provided evidence that zebrafish immune cell activity is under circadian control, which we will discuss in more detail in a later section. Using Morpholino-mediated knockdown of both p47phox and Nox2, Brothers and colleagues showed that Nox2 activation by leukocytes is required for the killing of Candida albicans in vivo [40]. The same group provided evidence that both Nox2 and Duox activity are required for leukocyte recruitment to sites of Candida albicans infection, although the cellular source of ROS (leukocyte vs epithelial cell) was not clear from this study [39]. The role of Duox in the zebrafish intestine in limiting the infection by Salmonella enterica has also been demonstrated [36]. In contrast, Deng and colleagues used a localized infection model with Pseudomonas aeruginosa bacteria injected into the inner ear or the yolk sac to show that Duox-generated H2O2 from the tissue is not required for neutrophil recruitment to sites of bacterial

infections but is required for tail wound healing [54]. The difference among these findings about the role of H2O2 for neutrophil recruitment to sites of infections could be due to distinct immune responses triggered by specific microbes. A slightly different antimicrobial role of zebrafish neutrophils has been reported for the defense against Mycobacterium marinum, a natural fish pathogen. In contrast to macrophages, neutrophils do not directly interact with mycobacteria at the initial site of infection. However, they respond to signals from infected macrophages, which recruit neutrophils to the granuloma (a site of macrophage aggregation), where neutrophils phagocytose dying macrophages and kill internalized bacteria through a Nox2-dependent mechanisms [61]. Nox2-mediated superoxide production by neutrophils can also kill E. Coli bacteria at a distance without phagocytosis [120]. Using a notochord injection model in larval zebrafish, it was found that neutrophils can be recruited to the infection site but they are unable to reach and engulf the bacteria due to a collagen barrier. Pharmacological Nox inhibition and p47phox Morpholino approach revealed that Nox2 in neutrophils is required to kill E. coli without phagocytosis [120]. Infections by the fungus Aspergillus nidulans are quite common in human patients suffering from CGD, which contributes significantly to its morbidity and mortality. Schoen and colleagues developed a zebrafish model of CGD and demonstrated a critical role of Nox2 in controlling the invasive hyphal growth using a p22phox-/- mutant fish line [57]. Re-expression of p22phox in neutrophils rescued the p22phox-deficient larvae from fungal growth and excessive inflammation. This and other fish lines that lack functional components of the phagocytic Nox2 complex will serve as excellent tools for further investigations of the underlying molecular and cellular mechanisms of host defense against microbial infections and related disorders.

5.4

Nox in Zebrafish Tumorigenesis

The relationship between cellular ROS levels and tumorigenesis is complex. Cancer cells generally have elevated ROS levels when compared to normal cells; therefore, antioxidant strategies have been employed in order to kill cancer cells [121]. However, since the cellular redox defense systems tend to be upregulated in cancer cells as well, therapeutic antioxidant treatments have frequently proven to be ineffective to treat cancer. Therefore, pro-oxidant treatments of cancer cells have been considered as well. Due to the significant conservation between the zebrafish and human genome including cell cycle genes and oncogenes, zebrafish has emerged as a popular model system for cancer research over the last two decades [122]. Zebrafish can readily develop tumors after exposure to mutagens or through

29

NADPH Oxidases in Zebrafish

transgenesis, which are genetically and histologically comparable to human cancers [123]. Cancer growth can be visualized in vivo at single cell resolution using transparent adult and larval zebrafish [123]. Different Nox isoforms have been associated with distinct types of human cancers with respect to tumor initiation, progression, and suppression. For example, increased expression of Nox1 and Duox2 has been observed in colorectal cancer as well as elevated expression levels of Nox4 and Nox5 in malignant melanomas [124]. Furthermore, exogenous expression of Nox1 in mouse fibroblast cells caused increased levels of H2O2 as well as cell transformation and tumor formation [125]. Oncogenic Ras activates Nox expression, which leads to ROS-mediated cell transformation [126]. Zebrafish expressing oncogenic human Ras exhibit elevated ROS levels due to increased Nox4 expression, whereas Nox inhibition mitigates the oncogenic effects of Ras expression [50]. Nox4-dependent increase of ROS mediates abnormal cell proliferation and simultaneously stimulates activation of DNA damage response (DDR), which is known for its role in cellular senescence [50]. Feng and colleagues used a transgenic fish line in which melanocytes were transformed with human oncogenic Ras- or Src to demonstrate that Duox-derived H2O2 recruits leukocytes towards the cancer cells similarly as in the wound response [48]. Further evidence for Nox-mediated ROS in tumorigenesis comes from the work of Daugaard and colleagues, who found that tumor suppressor HACE1 (HECT domain and Ankyrin repeat Containing E3ubiquitin-protein ligase 1) functions as a negative regulator of Rac1-dependent Nox activation. HACE1 specifically ubiquitinates Rac1 for proteasomal degradation, thereby controlling Nox activation and ROS production in zebrafish [69]. Using zebrafish in vivo model along with human tumor cell line and mice model, Daugaard and colleagues found that loss of HACE1 leads to elevated ROS levels due to overactivation of Nox1, which increases both ROS-mediated DNA damage and cyclin D1 expression leading to tumor formation [69]. Zebrafish embryos have been successfully used for highthroughput screening of drugs to treat human cancers, and some of these drugs have entered into clinical trials [123]. Several ROS-modulating chemicals have been tested in zebrafish for their anti-cancer properties. Using transgenic and xenograft zebrafish models, anti-angiogenic and antitumor effects of several anti-oxidants such as nobiletin [127], polymethoxylated flavonoids [128], quercetin4′-O-β-D-glucopyranoside (QODG) [129], quercetin [130], and eupatilin [131] have been reported. More investigations are needed to get a better understanding of Nox function in tumor formation and of therapeutic Nox manipulations to treat cancers in zebrafish.

499

5.5

Nox in Zebrafish Circadian Clock

Circadian rhythms are regulated by multiple factors most notably by light detected by the eye and are critical for maintaining homeostasis of the organism including the immune system. In humans, several cardiovascular and immunological parameters including heart rate, white blood cell count, and phagocytic index undergo a circadian rhythm with peak activity at night [132, 133]. Circadian oscillations of cellular redox states have been detected in plants [134], animals [135], and fungi [136]. Circadian rhythms controlled by expression of the clock gene have also been identified in zebrafish [137, 138]. Kaplan and colleagues observed a rhythmic pattern of phagocytosis and ROS level changes in zebrafish leukocytes throughout a twenty-four hour period with a peak of ROS generation before dawn [29]. Hirayama and colleagues showed that light controls H2O2-levels via oscillation of catalase expression, which is antiphasic with zebrafish clock gene expression regulating circadian rhythm [139]. Expression of zebrafish clock gene cry1a and per2 is regulated by H2O2, which is generated by the light stimulus [139]. Nox inhibition with diphenyleneiodonium chloride (DPI) provided evidence that Nox might be the sensor for light-mediated expression of the zebrafish clock genes cry1a and per2 [41]. This work was followed up by the Vallone group, who showed that these zebrafish clock genes are transcriptionally activated by D-Box enhancer elements and that Nox-derived ROS act upstream in this gene activation process [14]. Exposure of blue light to zebrafish PAC-2 fibroblast cells increased Nox-mediated ROS production and subsequent clock gene expression providing evidence for a light>Nox > ROS>MAPK>D-Box>clock gene signaling pathway in the control of circadian rhythm [14]. Interestingly, in mammalian cells ROS do not induce D-box driven clock gene expression although light still induces ROS increase, which indicates an evolutionary change in the role of ROS for circadian rhythm [14]. The molecular details of how light affects Nox activity remain to be determined.

6

Conclusion

Because of the availability of transgenic lines and in vivo imaging, zebrafish has emerged as an excellent model system to investigate Nox function in various biological processes. Whereas we have learned much about the role of Nox in wound healing, regeneration, infection control, and inflammation, there are still significant gaps in our understanding of the molecular pathways immediately upstream and downstream of Nox. Furthermore, the spatial and temporal profiles of the cellular ROS changes produced by Nox and other

500

systems controlling ROS levels are not well understood. These challenges are largely due to the limitations of specific ROS imaging probes and approaches of Nox inhibition. Whereas molecular approaches can provide isoform specificity, knocking out one Nox isoform can result in compensation by other isoforms. Finally, the similarity between human and zebrafish genome as well as the suitability for drug screening, make zebrafish an excellent model system to further investigate molecular and cellular mechanisms that underlie Nox functions in human diseases. Acknowledgements Work in the Suter Lab is supported by NIH grant R01NS117701 and a grant from the Indiana Spinal Cord and Brain Injury Research Fund.

References 1. Marshall RA, Osborn DP (2016) Zebrafish: a vertebrate tool for studying basal body biogenesis, structure, and function. Cilia 5(1): 1–9 2. Sundin J, Morgan R, Finnøen MH et al (2019) On the observation of Wild Zebrafish (Danio rerio) in India. Zebrafish 16(6):546–553 3. Streisinger G, Walker C, Dower N et al (1981) Production of clones of homozygous diploid zebra fish (Brachydanio rerio). Nature 291(5813):293–296. https://doi.org/10.1038/291293a0 4. Ge W (2018) Zebrafish. In: Skinner MK (ed) Encyclopedia of reproduction, 2nd edn. Academic, Oxford, pp 704–710. https:// doi.org/10.1016/B978-0-12-809633-8.20618-3 5. Howe K, Clark MD, Torroja CF et al (2013) The zebrafish reference genome sequence and its relationship to the human genome. Nature 496(7446):498–503. https://doi.org/10.1038/nature12111 6. Santoriello C, Zon LI (2012) Hooked! Modeling human disease in zebrafish. J Clin Invest 122(7):2337–2343. https://doi.org/10.1172/ jci60434 7. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87(1):245–313. https://doi.org/10.1152/physrev.00044.2005 8. Kawahara T, Quinn MT, Lambeth JD (2007) Molecular evolution of the reactive oxygen-generating NADPH oxidase (Nox/Duox) family of enzymes. BMC Evol Biol 7(1):1–21 9. Weaver CJ, Leung YF, Suter DM (2016) Expression dynamics of NADPH oxidases during early zebrafish development. J Comp Neurol 524(10):2130–2141. https://doi.org/10.1002/cne.23938 10. Rada B, Leto TL (2008) Oxidative innate immune defenses by Nox/Duox family NADPH oxidases. In: Trends in innate immunity, vol 15. Karger Publishers, pp 164–187 11. Terzi A, Roeder H, Weaver CJ et al (2021) Neuronal NADPH oxidase 2 regulates growth cone guidance downstream of slit2/ robo2. Dev Neurobiol 81(1):3–21. https://doi.org/10.1002/dneu. 22791 12. Weaver CJ, Terzi A, Roeder H et al (2018) nox2/cybb deficiency affects zebrafish retinotectal connectivity. J Neurosci 38(26): 5854–5871. https://doi.org/10.1523/jneurosci.1483-16.2018 13. Razaghi B, Steele SL, Prykhozhij SV et al (2018) hace1 Influences zebrafish cardiac development via ROS-dependent mechanisms. Dev Dyn 247(2):289–303 14. Pagano C, Siauciunaite R, Idda ML et al (2018) Evolution shapes the responsiveness of the D-box enhancer element to light and reactive oxygen species in vertebrates. Sci Rep 8(1):13180. https://doi.org/10.1038/s41598-018-31570-8

S. M. S. Alam and D. M. Suter 15. Chopra K, Ishibashi S, Amaya E (2019) Zebrafish duox mutations provide a model for human congenital hypothyroidism. Biology Open 8(2). https://doi.org/10.1242/bio.037655 16. Sun F, Fang Y, Zhang M-M et al (2021) Genetic manipulation on zebrafish duox recapitulate the clinical manifestations of congenital hypothyroidism. Endocrinology 162(8):1–14. https://doi.org/10. 1210/endocr/bqab10 17. Walker C, Streisinger G (1983) Induction of mutations by γ-rays in pregonial germ cells of zebrafish embryos. Genetics 103(1): 125–136 18. Lawson ND, Wolfe SA (2011) Forward and reverse genetic approaches for the analysis of vertebrate development in the zebrafish. Dev Cell 21(1):48–64 19. Kimmel CB (1989) Genetics and early development of zebrafish. Trends Genet 5(8):283–288. https://doi.org/10.1016/0168-9525 (89)90103-0 20. Stuart GW, McMurray JV, Westerfield M (1988) Replication, integration and stable germ-line transmission of foreign sequences injected into early zebrafish embryos. Development 103(2): 403–412 21. Higashijima S, Okamoto H, Ueno N et al (1997) High-frequency generation of transgenic zebrafish which reliably express GFP in whole muscles or the whole body by using promoters of zebrafish origin. Dev Biol 192(2):289–299. https://doi.org/10.1006/dbio. 1997.8779 22. Long Q, Meng A, Wang H et al (1997) GATA-1 expression pattern can be recapitulated in living transgenic zebrafish using GFP reporter gene. Development 124(20):4105–4111 23. Driever W, Solnica-Krezel L, Schier AF et al (1996) A genetic screen for mutations affecting embryogenesis in zebrafish. Development 123:37–46 24. Haffter P, Granato M, Brand M et al (1996) The identification of genes with unique and essential functions in the development of the zebrafish, Danio rerio. Development 123:1–36 25. Stainier DY, Fouquet B, Chen JN et al (1996) Mutations affecting the formation and function of the cardiovascular system in the zebrafish embryo. Development 123:285–292 26. Varga ZM (2011) Aquaculture and husbandry at the zebrafish international resource center. Methods Cell Biol 104:453–478. https://doi.org/10.1016/b978-0-12-374814-0.00024-0 27. Sprague J, Clements D, Conlin T et al (2003) The Zebrafish Information Network (ZFIN): the zebrafish model organism database. Nucleic Acids Res 31(1):241–243 28. Hermann AC, Millard PJ, Blake SL et al (2004) Development of a respiratory burst assay using zebrafish kidneys and embryos. J Immunol Methods 292(1–2):119–129 29. Kaplan JE, Chrenek RD, Morash JG et al (2008) Rhythmic patterns in phagocytosis and the production of reactive oxygen species by zebrafish leukocytes. Comp Biochem Physiol A Mol Integr Physiol 151(4):726–730 30. Niethammer P, Grabher C, Look AT et al (2009) A tissue-scale gradient of hydrogen peroxide mediates rapid wound detection in zebrafish. Nature 459(7249):996–999 31. Panieri E, Millia C, Santoro MM (2017) Real-time quantification of subcellular H2O2 and glutathione redox potential in living cardiovascular tissues. Free Radic Biol Med 109:189–200 32. Rieger S, Sagasti A (2011) Hydrogen peroxide promotes injuryinduced peripheral sensory axon regeneration in the zebrafish skin. PLoS Biol 9(5):e1000621. https://doi.org/10.1371/journal.pbio. 1000621 33. Tang H, Ge H, Chen ZB et al (2015) Micrometam C protects against oxidative stress in inflammation models in zebrafish and RAW264.7 macrophages. Mar Drugs 13(9):5593–5605. https:// doi.org/10.3390/md13095593

29

NADPH Oxidases in Zebrafish

34. Yoo SK, Starnes TW, Deng Q et al (2011) Lyn is a redox sensor that mediates leukocyte wound attraction in vivo. Nature 480(7375):109–112 35. Martin P, Feng Y (2009) Wound healing in zebrafish. Nature 459(7249):921–923 36. Flores MV, Crawford KC, Pullin LM et al (2010) Dual oxidase in the intestinal epithelium of zebrafish larvae has anti-bacterial properties. Biochem Biophys Res Commun 400(1):164–168. https://doi.org/10.1016/j.bbrc.2010.08.037 37. Choe CP, Choi S-Y, Kee Y et al (2021) Transgenic fluorescent zebrafish lines that have revolutionized biomedical research. Lab Animal Res 37(1):1–29 38. Maghzal GJ, Krause K-H, Stocker R et al (2012) Detection of reactive oxygen species derived from the family of NOX NADPH oxidases. Free Radic Biol Med 53(10):1903–1918 39. Brothers KM, Gratacap RL, Barker SE et al (2013) NADPH oxidase-driven phagocyte recruitment controls Candida albicans filamentous growth and prevents mortality. PLoS Pathog 9(10): e1003634 40. Brothers KM, Newman ZR, Wheeler RT (2011) Live imaging of disseminated candidiasis in zebrafish reveals role of phagocyte oxidase in limiting filamentous growth. Eukaryot Cell 10(7): 932–944 41. Osaki T, Uchida Y, Hirayama J et al (2011) Diphenyleneiodonium chloride, an inhibitor of reduced nicotinamide adenine dinucleotide phosphate oxidase, suppresses light-dependent induction of clock and DNA repair genes in zebrafish. Biol Pharm Bull 34(8): 1343–1347 42. Tell RM, Kimura K, Palić D (2012) Rac2 expression and its role in neutrophil functions of zebrafish (Danio rerio). Fish Shellfish Immunol 33(5):1086–1094 43. Goody MF, Peterman E, Sullivan C et al. (2013) Quantification of the respiratory burst response as an indicator of innate immune health in zebrafish. J Vis Exp JoVE 79 44. Nussbaum JM, Liu LJ, Hasan SA et al (2013) Homeostatic generation of reactive oxygen species protects the zebrafish liver from steatosis. Hepatology 58(4):1326–1338 45. Zhang Y, Shimizu H, Siu KL et al (2014) NADPH oxidase 4 induces cardiac arrhythmic phenotype in zebrafish. J Biol Chem 289(33):23200–23208 46. Lu XL, Liu JX, Wu Q et al (2014) Protective effects of puerarin against Aß40-induced vascular dysfunction in zebrafish and human endothelial cells. Eur J Pharmacol 732:76–85. https://doi.org/10. 1016/j.ejphar.2014.03.030 47. Chen J, Yu T, He X et al (2019) Dual roles of hydrogen peroxide in promoting zebrafish renal repair and regeneration. Biochem Biophys Res Commun 516(3):680–685. https://doi.org/10.1016/j. bbrc.2019.06.052 48. Feng Y, Santoriello C, Mione M et al (2010) Live imaging of innate immune cell sensing of transformed cells in zebrafish larvae: parallels between tumor initiation and wound inflammation. PLoS Biol 8(12):e1000562. https://doi.org/10.1371/journal.pbio. 1000562 49. de Oliveira S, Boudinot P, Calado  et al (2015) Duox1-derived H2O2 modulates Cxcl8 expression and neutrophil recruitment via JNK/c-JUN/AP-1 signaling and chromatin modifications. J Immunol 194(4):1523–1533. https://doi.org/10.4049/jimmunol. 1402386 50. Ogrunc M, Di Micco R, Liontos M et al (2014) Oncogene-induced reactive oxygen species fuel hyperproliferation and DNA damage response activation. Cell Death Differ 21(6):998–1012. https://doi. org/10.1038/cdd.2014.16 51. Bernut A, Dupont C, Ogryzko NV et al (2019) CFTR protects against mycobacterium abscessus infection by fine-tuning host oxidative defenses. Cell Rep 26(7):1828–1840. e1824. https://doi. org/10.1016/j.celrep.2019.01.071

501 52. Gauron C, Meda F, Dupont E et al (2016) Hydrogen peroxide (H2O2) controls axon pathfinding during zebrafish development. Dev Biol 414(2):133–141 53. Pase L, Layton JE, Wittmann C et al (2012) Neutrophil-delivered myeloperoxidase dampens the hydrogen peroxide burst after tissue wounding in zebrafish. Curr Biol 22(19):1818–1824 54. Deng Q, Harvie EA, Huttenlocher A (2012) Distinct signalling mechanisms mediate neutrophil attraction to bacterial infection and tissue injury. Cell Microbiol 14(4):517–528. https://doi.org/ 10.1111/j.1462-5822.2011.01738.x 55. Terzi A, Alam SMS, Suter DM (2021) ROS live cell imaging during neuronal development. J Visual Exp JoVE 168. https:// doi.org/10.3791/62165 56. Yan B, Han P, Pan L et al (2014) IL-1β and reactive oxygen species differentially regulate neutrophil directional migration and Basal random motility in a zebrafish injury-induced inflammation model. J Immunol 192(12):5998–6008. https://doi.org/10.4049/jimmunol. 1301645 57. Schoen TJ, Rosowski EE, Knox BP et al (2019) Neutrophil phagocyte oxidase activity controls invasive fungal growth and inflammation in zebrafish. J Cell Sci 133(5). https://doi.org/10.1242/jcs. 236539 58. Park JS, Choi TI, Kim OH et al (2019) Targeted knockout of duox causes defects in zebrafish growth, thyroid development, and social interaction. J Genet Genomics 46(2):101–104. https://doi.org/10. 1016/j.jgg.2019.01.004 59. de Oliveira S, López-Muñoz A, Candel S et al (2014) ATP modulates acute inflammation in vivo through dual oxidase 1-derived H2O2 production and NF-κB activation. J Immunol 192(12):5710–5719. https://doi.org/10.4049/jimmunol.1302902 60. Tauzin S, Starnes TW, Becker FB et al (2014) Redox and Src family kinase signaling control leukocyte wound attraction and neutrophil reverse migration. J Cell Biol 207(5):589–598. https:// doi.org/10.1083/jcb.201408090 61. Yang CT, Cambier CJ, Davis JM et al (2012) Neutrophils exert protection in the early tuberculous granuloma by oxidative killing of mycobacteria phagocytosed from infected macrophages. Cell Host Microbe 12(3):301–312. https://doi.org/10.1016/j.chom. 2012.07.009 62. Venkateswaran A, Sekhar KR, Levic DS et al (2014) The NADH oxidase ENOX1, a critical mediator of endothelial cell radiosensitization, is crucial for vascular development. Cancer Res 74(1): 38–43. https://doi.org/10.1158/0008-5472.can-13-1981 63. Belousov VV, Fradkov AF, Lukyanov KA et al (2006) Genetically encoded fluorescent indicator for intracellular hydrogen peroxide. Nat Methods 3(4):281–286. https://doi.org/10.1038/nmeth866 64. Dooley CT, Dore TM, Hanson GT et al (2004) Imaging dynamic redox changes in mammalian cells with green fluorescent protein indicators. J Biol Chem 279(21):22284–22293. https://doi.org/10. 1074/jbc.M312847200 65. Gutscher M, Sobotta MC, Wabnitz GH et al (2009) Proximitybased protein thiol oxidation by H2O2-scavenging peroxidases. J Biol Chem 284(46):31532–31540. https://doi.org/10.1074/jbc. M109.059246 66. Rennekamp AJ, Peterson RT (2015) 15 years of zebrafish chemical screening. Curr Opin Chem Biol 24:58–70 67. Mendieta-Serrano MA, Mendez-Cruz FJ, Antúnez-Mojica M et al (2019) NADPH-Oxidase-derived reactive oxygen species are required for cytoskeletal organization, proper localization of E-cadherin and cell motility during zebrafish epiboly. Free Radic Biol Med 130:82–98. https://doi.org/10.1016/j.freeradbiomed. 2018.10.416 68. Giusti N, Gillotay P, Trubiroha A et al (2020) Inhibition of the thyroid hormonogenic H(2)O(2) production by Duox/DuoxA in zebrafish reveals VAS2870 as a new goitrogenic compound. Mol

502 Cell Endocrinol 500:110635. https://doi.org/10.1016/j.mce.2019. 110635 69. Daugaard M, Nitsch R, Razaghi B et al (2013) Hace1 controls ROS generation of vertebrate Rac1-dependent NADPH oxidase complexes. Nat Commun 4(1):1–13 70. Candel S, de Oliveira S, López-Muñoz A et al (2014) Tnfa signaling through tnfr2 protects skin against oxidative stress-induced inflammation. PLoS Biol 12(5):e1001855. https://doi.org/10. 1371/journal.pbio.1001855 71. Ratanayotha A, Kawai T, Okamura Y (2019) Real-time functional analysis of Hv1 channel in neutrophils: a new approach from zebrafish model. Am J Phys Regul Integr Comp Phys 316(6): R819–R831 72. Robinson B, Gu Q, Ali SF et al (2019) Ketamine-induced attenuation of reactive oxygen species in zebrafish is prevented by acetyl l-carnitine in vivo. Neurosci Lett 706:36–42. https://doi.org/10. 1016/j.neulet.2019.05.009 73. Dikova V, Vorhauser J, Geng A et al (2020) Metabolic interaction of hydrogen peroxide and hypoxia in zebrafish fibroblasts. Free Radic Biol Med 152:469–481. https://doi.org/10.1016/j. freeradbiomed.2019.11.015 74. Choi J, Im GJ, Chang J et al (2013) Protective effects of apocynin on cisplatin-induced ototoxicity in an auditory cell line and in zebrafish. J Appl Toxicol 33(2):125–133 75. Seo MJ, Seo YJ, Pan CH et al (2016) Fucoxanthin suppresses lipid accumulation and ROS production during differentiation in 3T3-L1 adipocytes. Phytother Res 30(11):1802–1808. https://doi. org/10.1002/ptr.5683 76. Li ZX, Chen JW, Yuan F et al (2013) Xyloketal B exhibits its antioxidant activity through induction of HO-1 in vascular endothelial cells and zebrafish. Mar Drugs 11(2):504–522. https://doi. org/10.3390/md11020504 77. Bradford YM, Toro S, Ramachandran S et al (2017) Zebrafish models of human disease: gaining insight into human disease at ZFIN. ILAR J 58(1):4–16. https://doi.org/10.1093/ilar/ilw040 78. Elks PM, Renshaw SA, Meijer AH et al (2015) Exploring the HIFs, buts and maybes of hypoxia signalling in disease: lessons from zebrafish models. Dis Model Mech 8(11):1349–1360 79. van Rooijen E, Voest EE, Logister I et al (2009) Zebrafish mutants in the von Hippel-Lindau tumor suppressor display a hypoxic response and recapitulate key aspects of Chuvash polycythemia. Blood 113(25):6449–6460 80. Kawahara T, Lambeth JD (2007) Molecular evolution of Phoxrelated regulatory subunits for NADPH oxidase enzymes. BMC Evol Biol 7:178. https://doi.org/10.1186/1471-2148-7-178 81. White RJ, Collins JE, Sealy IM et al (2017) A high-resolution mRNA expression time course of embryonic development in zebrafish. elife 6:e30860 82. Papatheodorou I, Fonseca NA, Keays M et al (2017) Expression Atlas: gene and protein expression across multiple studies and organisms. Nucleic Acids Res 46(D1):D246–D251. https://doi. org/10.1093/nar/gkx1158 83. Wang J, Xiao J, Meng X et al (2021) NOX5 is expressed aberrantly but not a critical pathogenetic gene in Hirschsprung disease. BMC Pediatr 21(1):153. https://doi.org/10.1186/s12887-021-02611-5 84. Oehlers SH, Flores MV, Hall CJ et al (2011) The inflammatory bowel disease (IBD) susceptibility genes NOD1 and NOD2 have conserved anti-bacterial roles in zebrafish. Dis Model Mech 4(6): 832–841. https://doi.org/10.1242/dmm.006122 85. Opitz R, Maquet E, Zoenen M et al (2011) TSH receptor function is required for normal thyroid differentiation in zebrafish. Mol Endocrinol 25(9):1579–1599. https://doi.org/10.1210/me. 2011-0046 86. Olguín-Albuerne M, Morán J (2018) Redox signaling mechanisms in nervous system development. Antioxid Redox Signal 28(18): 1603–1625

S. M. S. Alam and D. M. Suter 87. Wilson C, Muñoz-Palma E, González-Billault C (2018) From birth to death: a role for reactive oxygen species in neuronal development. In: Semin Cell Dev Biol. Elsevier, pp 43–49 88. Terzi A, Suter DM (2020) The role of NADPH oxidases in neuronal development. Free Radic Biol Med 154:33–47. https://doi.org/ 10.1016/j.freeradbiomed.2020.04.027 89. Wilson C, Núñez MT, González-Billault C (2015) Contribution of NADPH oxidase to the establishment of hippocampal neuronal polarity in culture. J Cell Sci 128(16):2989–2995. https://doi.org/ 10.1242/jcs.168567 90. Munnamalai V, Weaver CJ, Weisheit CE et al (2014) Bidirectional interactions between NOX2-type NADPH oxidase and the F-actin cytoskeleton in neuronal growth cones. J Neurochem 130(4): 526–540. https://doi.org/10.1111/jnc.12734 91. Munnamalai V, Suter DM (2009) Reactive oxygen species regulate F-actin dynamics in neuronal growth cones and neurite outgrowth. J Neurochem 108(3):644–661. https://doi.org/10.1111/j. 1471-4159.2008.05787.x 92. Rampon C, Volovitch M, Joliot A et al. (2018) Hydrogen peroxide and redox regulation of developments. Antioxidants (Basel, Switzerland) 7(11). https://doi.org/10.3390/antiox7110159 93. Yoneyama M, Kawada K, Gotoh Y et al (2010) Endogenous reactive oxygen species are essential for proliferation of neural stem/progenitor cells. Neurochem Int 56(6):740–746. https://doi. org/10.1016/j.neuint.2009.11.018 94. Le Belle JE, Orozco NM, Paucar AA et al (2011) Proliferative neural stem cells have high endogenous ROS levels that regulate self-renewal and neurogenesis in a PI3K/Akt-dependant manner. Cell Stem Cell 8(1):59–71 95. Dickinson BC, Peltier J, Stone D et al (2011) Nox2 redox signaling maintains essential cell populations in the brain. Nat Chem Biol 7(2):106–112 96. Nayernia Z, Colaianna M, Robledinos-Antón N et al (2017) Decreased neural precursor cell pool in NADPH oxidase 2-deficiency: from mouse brain to neural differentiation of patient derived iPSC. Redox Biol 13:82–93. https://doi.org/10.1016/j. redox.2017.04.026 97. Albadri S, Naso F, Thauvin M et al (2019) Redox signaling via lipid peroxidation regulates retinal progenitor cell differentiation. Dev Cell 50(1):73–89e76. https://doi.org/10.1016/j.devcel.2019. 05.011 98. Song Y, Driessens N, Costa M et al (2007) Roles of hydrogen peroxide in thyroid physiology and disease. J Clin Endocrinol Metabol 92(10):3764–3773 99. Santillo M, Colantuoni A, Mondola P et al (2015) NOX signaling in molecular cardiovascular mechanisms involved in the blood pressure homeostasis. Front Physiol 6:194 100. Iovine MK (2007) Conserved mechanisms regulate outgrowth in zebrafish fins. Nat Chem Biol 3(10):613–618. https://doi.org/10. 1038/nchembio.2007.36 101. Sehring IM, Weidinger G (2020) Recent advancements in understanding fin regeneration in zebrafish. Wiley Interdiscip Rev Dev Biol 9(1):e367. https://doi.org/10.1002/wdev.367 102. Azevedo AS, Grotek B, Jacinto A et al (2011) The regenerative capacity of the zebrafish caudal fin is not affected by repeated amputations. PLoS One 6(7):e22820 103. Jelcic M, Enyedi B, Xavier JB et al (2017) Image-based measurement of H2O2 reaction-diffusion in wounded zebrafish larvae. Biophys J 112(9):2011–2018 104. Poole LB, Nelson KJ (2008) Discovering mechanisms of signaling-mediated cysteine oxidation. Curr Opin Chem Biol 12(1):18–24. https://doi.org/10.1016/j.cbpa.2008.01.021 105. Yoo SK, Freisinger CM, LeBert DC et al (2012) Early redox, Src family kinase, and calcium signaling integrate wound responses and tissue regeneration in zebrafish. J Cell Biol 199(2):225–234

29

NADPH Oxidases in Zebrafish

106. Sehring IM, Jahn C, Weidinger G (2016) Zebrafish fin and heart: what’s special about regeneration? Curr Opin Genet Dev 40:48–56. https://doi.org/10.1016/j.gde.2016.05.011 107. Gauron C, Rampon C, Bouzaffour M et al (2013) Sustained production of ROS triggers compensatory proliferation and is required for regeneration to proceed. Sci Rep 3:2084. https://doi.org/10. 1038/srep02084 108. Chopra K, Folkmanaitė M, Stockdale L et al. (2021) The roles of NADPH oxidases during adult zebrafish fin regeneration. bioRxiv. https://doi.org/10.1101/2021.07.13.452203 109. Han P, Zhou XH, Chang N et al (2014) Hydrogen peroxide primes heart regeneration with a derepression mechanism. Cell Res 24(9): 1091–1107. https://doi.org/10.1038/cr.2014.108 110. Yang LQ, Chen M, Ren DL et al (2020) Dual oxidase mutant retards mauthner-cell axon regeneration at an early stage via modulating mitochondrial dynamics in zebrafish. Neurosci Bull 36(12):1500–1512. https://doi.org/10.1007/s12264-020-00600-9 111. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. J Clin Invest 52(3):741–744. https://doi.org/ 10.1172/jci107236 112. Hohn DC, Lehrer RI (1975) NADPH oxidase deficiency in X-linked chronic granulomatous disease. J Clin Invest 55(4): 707–713 113. Rossi F, Zatti M (1964) Biochemical aspects of phagocytosis in poly-morphonuclear leucocytes. NADH and NADPH oxidation by the granules of resting and phagocytizing cells. Experientia 20(1): 21–23 114. Harvie EA, Huttenlocher A (2015) Neutrophils in host defense: new insights from zebrafish. J Leukoc Biol 98(4):523–537 115. Rosowski EE (2020) Determining macrophage versus neutrophil contributions to innate immunity using larval zebrafish. Dis Model Mech 13(1). https://doi.org/10.1242/dmm.041889 116. Yoshida N, Frickel E-M, Mostowy S (2017) Macrophage–microbe interactions: lessons from the zebrafish model. Front Immunol 8: 1703 117. Torraca V, Masud S, Spaink HP et al (2014) Macrophage-pathogen interactions in infectious diseases: new therapeutic insights from the zebrafish host model. Dis Model Mech 7(7):785–797 118. Hogan D, Wheeler RT (2014) The complex roles of NADPH oxidases in fungal infection. Cell Microbiol 16(8):1156–1167 119. Bennett CM, Kanki JP, Rhodes J et al (2001) Myelopoiesis in the zebrafish, Danio rerio. Blood 98(3):643–651. https://doi.org/10. 1182/blood.v98.3.643 120. Phan QT, Sipka T, Gonzalez C et al (2018) Neutrophils use superoxide to control bacterial infection at a distance. PLoS Pathog 14(7):e1007157. https://doi.org/10.1371/journal.ppat.1007157 121. Trachootham D, Alexandre J, Huang P (2009) Targeting cancer cells by ROS-mediated mechanisms: a radical therapeutic approach? Nat Rev Drug Discov 8(7):579–591. https://doi.org/10. 1038/nrd2803 122. Fazio M, Ablain J, Chuan Y et al (2020) Zebrafish patient avatars in cancer biology and precision cancer therapy. Nat Rev Cancer 20(5):263–273. https://doi.org/10.1038/s41568-020-0252-3 123. White R, Rose K, Zon L (2013) Zebrafish cancer: the state of the art and the path forward. Nat Rev Cancer 13(9):624–636

503 124. Roy K, Wu Y, Meitzler JL et al (2015) NADPH oxidases and cancer. Clin Sci 128(12):863–875 125. Arnold RS, Shi J, Murad E et al (2001) Hydrogen peroxide mediates the cell growth and transformation caused by the mitogenic oxidase Nox1. Proc Natl Acad Sci USA 98(10):5550–5555. https://doi.org/10.1073/pnas.101505898 126. Kamata T (2009) Roles of Nox1 and other Nox isoforms in cancer development. Cancer Sci 100(8):1382–1388 127. Lam KH, Alex D, Lam IK et al (2011) Nobiletin, a polymethoxylated flavonoid from citrus, shows anti-angiogenic activity in a zebrafish in vivo model and HUVEC in vitro model. J Cell Biochem 112(11):3313–3321 128. Lam IK, Alex D, Wang YH et al (2012) In vitro and in vivo structure and activity relationship analysis of polymethoxylated flavonoids: identifying sinensetin as a novel antiangiogenesis agent. Mol Nutr Food Res 56(6):945–956. https://doi.org/10. 1002/mnfr.201100680 129. Lin C, Wu M, Dong J (2012) Quercetin-4′-O-β-D-glucopyranoside (QODG) inhibits angiogenesis by suppressing VEGFR2-mediated signaling in zebrafish and endothelial cells. PLoS One 7(2): e31708. https://doi.org/10.1371/journal.pone.0031708 130. Zhao D, Qin C, Fan X et al (2014) Inhibitory effects of quercetin on angiogenesis in larval zebrafish and human umbilical vein endothelial cells. Eur J Pharmacol 723:360–367. https://doi.org/10. 1016/j.ejphar.2013.10.069 131. Lee JY, Bae H, Yang C et al (2020) Eupatilin promotes cell death by calcium influx through ER-mitochondria axis with SERPINB11 inhibition in epithelial ovarian cancer. Cancers 12(6). https://doi. org/10.3390/cancers12061459 132. Melchart D, Martin P, Hallek M et al (1992) Circadian variation of the phagocytic activity of polymorphonuclear leukocytes and of various other parameters in 13 healthy male adults. Chronobiol Int 9(1):35–45 133. Scheiermann C, Kunisaki Y, Frenette PS (2013) Circadian control of the immune system. Nat Rev Immunol 13(3):190–198 134. Lai AG, Doherty CJ, Mueller-Roeber B et al (2012) Circadian Clock-Associated 1 regulates ROS homeostasis and oxidative stress responses. Proc Natl Acad Sci USA 109(42):17129–17134. https://doi.org/10.1073/pnas.1209148109 135. Fanjul-Moles ML (2013) ROS signaling pathways and biological rhythms: perspectives in crustaceans. Front Biosci 18:665–675. https://doi.org/10.2741/4129 136. Yoshida Y, Iigusa H, Wang N et al (2011) Cross-talk between the cellular redox state and the circadian system in Neurospora. PLoS One 6(12):e28227. https://doi.org/10.1371/journal.pone.0028227 137. Delaunay F, Thisse C, Marchand O et al (2000) An inherited functional circadian clock in zebrafish embryos. Science 289(5477):297–300 138. Whitmore D, Foulkes NS, Strähle U et al (1998) Zebrafish Clock rhythmic expression reveals independent peripheral circadian oscillators. Nat Neurosci 1(8):701–707 139. Hirayama J, Cho S, Sassone-Corsi P (2007) Circadian control by the reduction/oxidation pathway: catalase represses lightdependent clock gene expression in the zebrafish. Proc Natl Acad Sci 104(40):15747–15752

Part VI Structure

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex

30

Jing-Xiang Wu, Ji Sun, and Lei Chen

Abstract

Dual oxidases (DUOX), dedicated enzymes for regulated hydrogen peroxide production, play essential roles in many biological processes. Recent advance in the determination of DUOX1 structures provides unprecedented atomic insights into how DUOX family members work. Here, we review our current structural understanding of the assembly, electron transfer pathway, and regulatory mechanism of DUOX. Keywords

DUOX · Calcium · Hydrogen peroxide · Reactive oxygen species · EF hand · FAD · NADPH

1

Introduction

Dual oxidase 1 (DUOX1) is a member of the NADPH oxidase (NOX) family and mediates cross-membrane electron transfer from cytosolic NADPH to extracellular oxygen to produce hydrogen peroxide (H2O2) [1]. The NOX family consists of seven members, NOX1–5 and DUOX1–2, representing a cluster of enzymes that produce reactive J.-X. Wu · L. Chen (✉) State Key Laboratory of Membrane Biology, College of Future Technology, Institute of Molecular Medicine, Peking University, Beijing Key Laboratory of Cardiometabolic Molecular Medicine, Beijing, China Peking-Tsinghua Center for Life Sciences, Peking University, Beijing, China Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, China National Biomedical Imaging Center, Peking University, Beijing, China e-mail: [email protected]; [email protected] J. Sun Department of Structural Biology, St. Jude Children’s Research Hospital, Memphis, TN, USA e-mail: [email protected]

oxygen species (ROS) [2]. Initially described as a thyroid oxidase for its identification in the thyroid gland [3–5], DUOX1 has been detected in various cell types and tissues, including epithelial cells of the epidermis, urothelium and respiratory tract, and non-epithelial cells such as T cells and macrophages [6–10]. Though the physiological function of DUOX1 is not fully understood yet, H2O2 production by DUOX1 has been shown to play essential roles in host defense, hormone biosynthesis, mitogenesis, apoptosis, etc. [11]. Additionally, abnormal DUOX1 expression is implicated in multiple diseases, such as allergy, asthma, lung fibrosis, and cancers [11–13]. High-level DUOX1 expression is proposed to be a prognostic marker in patients with hepatocellular carcinoma [14]. DUOX1 forms catalytically active complexes with an accessory subunit, DUOX activator 1 (DUOXA1) [15]. The human DUOX1 contains 1551 amino acids and is composed of the peroxidase homology domain (PHD), pleckstrin homology-like domain (PHLD), EF-hand calcium-binding module (EF module), transmembrane domain (TMD), and dehydrogenase (DH) domain (Fig. 30.1a) [16, 17]. The DH domain contains FAD-binding and NADPH-binding domains (FBD and NBD, respectively). The TMD and DH domains make up the catalytic module of all NOX proteins [2]. DUOXA1 is a transmembrane glycoprotein, expression of which rescues the catalytic function of DUOX1 in reconstituted systems [15]. DUOXA1 regulates DUOX1 by promoting endoplasmic reticulum (ER)-to-Golgi transition, maturation, translocation to the plasma membrane, and protein stability [18, 19]. The enzymatic activity of DUOX1-DUOXA1 complexes is tightly regulated. DUOX1-DUOXA1-mediated H2O2 production is Ca2+-dependent [20–22]. The cytoplasmic EF module of DUOX1 could sense intracellular Ca2+ signals; mutations of calcium-binding residues in the EF module abolish the enzymatic activities of DUOX1-DUOXA1 complexes. The function of DUOX1-DUOXA1 complexes is also modulated by post-translational modifications. For

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_30

507

508

J.-X. Wu et al.

a M0

PHD 22

TMD

EF

PHLD

591 619 643

756

DH

960 1012

1270

1551

N-ter

PHD PHD C

A

M1

M0

M2

preM1

M3

E

M4

M6

In

DH

EF

preM1M6

M0

D

B

PHLD

M5

Out

PHLD

DH

EF

b

c CBS1

PHLD

CBS2

αB

R395 βD βB

S176

D397

βC

βA

βG

βF βE

T332

βH T170

αA

e

d

PHD-TMD interface

DH domain FBD

NBD PHD

β13

β5 α5

α2

C118-C1165 disulfide bond

loop C

α3 loop A

β11 TMD

Fig. 30.1 Structure of the human DUOX1 subunit. (a) Top: scheme of DUOX1 with domain boundaries shown. The PHD, M0, PHLD, EF module, TMD, and DH domains of DUOX1 are colored in violet,

M1

M3 M2

M0

yellow, lightblue, orange, marine, and cyan, respectively. Same color codes will be used in this chapter unless otherwise noted. Bottom left: cartoon scheme of DUOX1. Bottom right: cryo-EM structure of human

30

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex

example, N-glycosylation of DUOX1 and DUOXA1 is essential for protein maturation, and phosphorylation of DUOX1 could boost H2O2 production [21]. Furthermore, the dimerization of DUOX1 PHD has been reported, suggesting a potential regulatory role of protein oligomerization [23]. Recent structural studies have illustrated the highresolution cryo-EM structures of both human and mouse DUOX1-DUOXA1 complexes and shed light on key questions regarding the overall architecture, complex assembly, electron transfer, and Ca2+-regulations [16, 17]. This chapter will briefly summarize the structural findings and the functional implications of the DUOX1-DUOXA1 complex.

2

Architecture of DUOX1 and DUOXA1

The cryo-EM structure of DUOX1-DUOXA1 complexes reveals a 2:2 heterotetrameric assembly with 140 Å × 105 Å × 160 Å in dimension and an overall twofold rotational symmetry (Fig. 30.1b). Each tetrameric complex has an overall two-fold rotational symmetry, with the symmetry axis running perpendicular to the cell membrane. To dissect this complex machinery, we will first introduce structural elements of each subunit using the human DUOX1DUOXA1 complex (Fig. 30.1b) [17]. DUOX1, the catalytic subunit, contains the PHD, PHLD, EF module, TMD, and DH domain from N-terminus to C-terminus (Fig. 30.1a). The PHD shares structural homology to peroxidases [16, 17]. However, it lacks the key histidines needed for heme coordination, and no heme densities are observed in both human and mouse DUOX1 structures, suggesting that PHD is not catalytically active as a peroxidase [16, 17, 24]. The PHD contains two putative cation binding sites (CBS1 and CBS2) with strong cryo-EM densities [17]. The cation in CBS1 is coordinated by the side chains of D397 and T332, and the main chain carbonyl groups of V399, T332, and R395; the second cation is coordinated by the side chains of D109, D174, S176, and T170, and the carbonyl groups of T170 and W172 (Fig. 30.1c). The identities of these cations remain to be characterized. In CBS2, peroxidases such as LPO (PDB 5B72) and DdPxoA (PDB 6ERC) accommodate a calcium ion in the corresponding position, but the ion density persists in the presence of EGTA in human DUOX1 [17, 25, 26]. On the

509

other hand, residues constituting CBS1 and CBS2 are evolutionally conserved, indicating their functional importance; mutations in both the CBS1 (D397A + T332A) and CBS2 (D109A + D174A) disrupt the DUOX1-DUOXA1 complex assembly [17]. Lastly, the PHD is glycosylated, and sugar moieties were observed on N94, N342, and N534 in both human and mouse structures (Fig. 30.1a and b) [16, 17]. Following PHD are the cytosolic PHLD and EF modules; a transmembrane helix, M0, connects PHD and PHLD. PHLD is named for its structural similarity to the β sheetsrich pleckstrin homology domain (Fig. 30.1d). The EF module contains two Ca2+-binding motifs, EF1 and EF2. PHLD and the EF module are responsible for sensing Ca2+ signals. The TMD contains an amphipathic helix (preM1) and six transmembrane helices (M1–M6) (Fig. 30.1a). PreM1 floats on the inner leaflet of the plasma membrane, providing an anchor for M1 and the interacting DH domain. This has been observed in the csNOX5 as well [27] and is probably a shared feature for all NOX proteins. M1–M6 adopts the folding of a ferric oxidoreductase domain and binds two heme molecules to catalyze cross-membrane electron transfer (Fig. 30.1a). The cytosolic DH domain is composed of FBD and NBD (Fig. 30.1e). The FBD is a cylindrical beta-domain, consisting of two orthogonal sheets (β1-β7-β4-β5 vs. β2-β6β3) separated by one alpha helix (α1) (Fig. 30.1f). The NBD adopts a “Rossmann-like fold“characterized by an alternating motif of β-α-β secondary structures (Fig. 30.1e). A conserved ‘GXGXG’ motif is located in the tight loop connecting the first β-strand (β8) and α-helix (α2), contributing to the binding of the NADPH diphosphate group [16]. Different from a canonical “Rossmann fold” [28], the NBD of DUOX1 has an insertion (between β10 and α5), which is involved in interaction with the EF-module (Fig. 30.1e) and probably plays a critical role in DUOX1 activation (Fig. 30.1f). Interdomain interactions contribute to the assembly of DUOX1 and connect intracellular, transmembrane, and extracellular modules. The PHD sits on the top of TMD and interacts with loop A (between M1 and M2) and loop C (between M3 and M4) (Fig. 30.1f). In addition, a disulfide bond between C118 of PHD and C1165 on loop C further helps to anchor the PHD (Fig. 30.1f). On the cytosolic side, the DH domain packs on the intracellular side of TMD (Fig. 30.1a). DUOXA1 is a membrane protein with five transmembrane helices (M1–M5) (Fig. 30.2). The extracellular domain of

ä

Fig. 30.1 (Continued) DUOX1 in the presence of high Ca2+. (b) The potential cation binding sites in PHDs. (c) The structure of PHLD with secondary structure labeled. The secondary structure nomenclature in this chapter is consistent with (Wu et al., 2021). (d) The structure of DH

domain. FBD and NBD are separated by a double-dashed line. The “Rossmann fold” is circled in gray dashed lines. The interface for EF module interaction is indicated by a yellow line. (e) The PHD-TMD interface. The disulfide bond between PHD and TMD is labeled

510

J.-X. Wu et al.

DUOXA1 mainly contains the N-terminal peptide (NTP), M2–M3, and M4–M5 loops. NTP involves in interaction with DUOX1, and the M2–M3 loop adopts a compact claudin-like domain (CLD) structure with two helices held in the four beta-strands (Fig. 30.2) [29]. The M4–M5 loop forms a hairpin structure (Fig. 30.2). Furthermore, the extracellular domain is glycosylated, and the cryo-EM structure study was able to resolve 11 sugar moieties on N109, mediating extensive interdomain interactions between DUOX1 and DUOXA1 (Fig. 30.3a). The cytosolic C-terminal part of DUOXA1 (~14 residues) is disordered based on the cryo-EM densities. Of note, M2–M5 and the extracellular part of DUOXA1 share structural similarities to the claudin superfamily [29].

3

Interdomain Interaction and DUOX1-DUOXA1 Complex Architecture

Looking from the cell membrane, the complex can be divided into three layers: the extracellular layer, the transmembrane layer, and the intracellular layer [17]. The extracellular layer involves interactions between DUOX1 PHD domains and between PHD and DUOXA1 extracellular domains. Overall, the two large N-terminal PHDs of DUOX1 pack against each other diagonally and are buttressed by the extracellular domain of DUOXA1 from beneath. Mutation of residues at the PHD dimer interface

(Fig. 30.3b) could disrupt the complex assembly. DUOXA1 interacts with PHDs from both DUOX1. On one side, the CLD of DUOXA1 holds one PHD domain on the top (Fig. 30.3a), burying a surface area of ~1000 Å2. Interestingly, Q30 of DUOX1 is located at this interface, whose mutation in the corresponding position (Q36) of DUOX2, could lead to congenital hypothyroidism. It is conceivable that mutation of Gln36 could disrupt the interaction between DUOX2 and DUOXA2 [30]. On the other side, the NTP and glycan on N109 of DUOXA1 insert into the “neck” region between PHD and TMD of the second DUOX1. In addition, linkers connecting β1-β2, β3-β 4, and H1-H2 of DUOXA1 also interact with the PHD (Fig. 30.3a). The transmembrane layer contains 24 transmembrane helices (Fig. 30.3c). Inter-subunit interaction in this layer is less intensive, comparing to the extracellular layer. On the extracellular leaflet side, the M5–M6 linker of DUOX1 interacts with M3 of DUOXA1 with contributions from four other residues: G1169 and V1190 of DUOX1 and L134 and N171 of DUOXA1. An ordered lipid is sandwiched between S1 of DUOX1 and M1 of DUOXA1 in the mouse complex structure [16]. On the intracellular leaflet side, M6 of DUOX1 interacts with M3 and M4 of DUOXA1 mainly through hydrophobic interactions and preM1 of DUOX1 packs against the M1 and M1–2 linker of DUOXA1 (Fig. 30.3d). At the center of the transmembrane layer, there is a large cavity without discernable protein densities (Fig. 30.3c). The interior surface of this cavity is highly

CLD

NTP NTP

N109

Out

M3

M1

M2

M3

M4

M4

M5 M1

M2

M5

In

Fig. 30.2 Structure of the human DUOXA1 subunit. Left: cartoon scheme of the DUOXA1 subunit. NTP: N-terminal peptide. CLD: claudin-like domain. Right: the cryo-EM structure of human DUOXA1

30

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex

511

extracellular layer

a

105 Å

160 Å

transmembrane layer

140 Å

90º

intracellular layer

180º

b PHD

PHD N121

F8

N84

P9 P9

H6

M3

F10

NTP

PHD

N109

P16

Y11

F18

M2

M1

TMD

M5

M4

DUOXA1 c

d

e

DUOXA1’ M5

M5

R507

M1

M3 M0

M2

R50 E41

DUOX1

M1

M3

F313

DUOX1-DUOXA1

M1

M2

M4

M4

M6

M6

M3

DUOX1’

M3

M1

M4 M2

M1

hydrophobic, and there are several lipid molecules bound on this surface, suggesting this cavity is probably filled by phospholipids on the cell membrane. The intracellular layer is comprised of the catalytic DH domain and regulatory PHLD and EF modules of DUOX1, while the cytosolic part of DUOXA1 is mostly disordered. PHLD and EF modules are located at the periphery, DH domains in the middle. Interdomain interactions in the intracellular layer are minimal [17].

I45 G44

M5

DUOXA1

Fig. 30.3 Architecture of DUOX1-DUOXA1 complexes. (a) Detailed interaction between DUOX1 and DUOXA1. Right: 90-degree rotation of the interface indicated by an arrow in the middle panel. (b) The interactions between DUOX1 and DUOXA1 at the extracellular side. (c) The PHD-PHD interface residues whose mutations could disrupt the

I41

R1042 I1038

M2

M2

V40

M5

M4 M0

I1045

M1

P43

E1039

complex assembly. (d) The top view of the cross-section of the transmembrane layer of the cryoEM map (filtered at 6 Å). The large cavity in the transmembrane layer is indicated by a dashed oval. (e) The DUOX1DUOXA1 interface at the intracellular leaflet of the membrane

4

Substrate and Cofactor Binding Sites for NADPH, FAD, Heme, and Oxygen

DUOX1 is the central catalytic subunit of the complex, which harbors the binding sites for the critical cofactors and substrates involved in the electron transfer (Fig. 30.4a). All of the available structures of the DUOX1-DUOXA1 complex show strong densities for two hemes bound within

512

J.-X. Wu et al.

a

b

c oxygen-reducing center

oxygen-reducing center

M5

M3

M2

M4

out Outer heme Out H1148

TMD H1238

R1087

Inner Inn heme

in

H1130

H1225

H1144

FAD NADPH

FBD

NBD

M4 M2 Outer heme

M3

M5 Inner heme

DH M5

M3

bits

3 2 1

D H HH I KN A

ALV

Q

R

S

D S

HK H

R

E

E

F AI

KRLVI Y ALAVI ST L G

T

D

N A

Q V

S

M

M A

W

M

T

C

M

I

Q

L W M

V F

S

S

L

S

A

L A H A

G

M

G

F

T

V

N

A

M

F

L

F

FL

T C T I VC VGVV VL Q I I

W

1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537

0

M4

Inner heme M3

A

S

L I

G

A

T

G

F

M

VL GS A

M

S F L

L

......

RRSFRGFWLTHHLLYLLLYVLLIIHGSF

4

CN

3

V I

A

A

I L

F

Y

C N

T

2

F

L V A TI VV I V LF V AG I GD

consensus D K N I V T F H K L V A Y M I A V F S

e

*

V L H S V G H V V N N V Y L F S I S

FS Y

AE I

N

A

I

AV

N

I

Q

T

D

N

V

I

C

S

0

Y

A L H T I A H L V C N F Y R F S V S

S

V

N

G

TY RN YLWS L A A

I

G A

I L D

T

K

Q

V

F I A

Q

R

NI

I

F

P

V

L

WVF

F

L

A

W

M

I

M

N S

L

V

MA I M GMCF

FLY

L

C

M

V

L

I I V AGTVL L

S

I

FSV C

M

W

F

G

Y

L

T

A

G

S

AF

AL VTY A

CVN Y T I E S

P

F

M

S

T

LGSFEVFWYTHHLGFIVFYILLIIHGSG

f

M2

FH H

L F WT L

G E Y H GFYVI FFY I LL I G Y H

N

H

H

L

L

I

W

V

A

V

L

P

1

S

YRL

1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747

4

*

D A A V V D F H R L I A S T A I V L T

1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578

hDUOX1

bits

d

R1131

M3

M1

lipid Y1041

D1128

N1040

V1127

E1039

R1214

loop D

R1036

loop B

preM1

FAD

FBD

NADPH -10 kT

Fig. 30.4 Substrate and cofactor binding sites on the DUOX catalytic subunit. (a) Side view of hDUOX1 catalytic domain (the TMD and DH domains, PDB:7D3E), the cofactors and substrates are shown in sticks. The putative oxygen-reducing center is indicated by the arrow. (b) The outer heme binding site and the adjacent oxygen-reducing center in the four-helical transmembrane bundle (M2–M5) of TMD. The coordinating residues are shown as sticks. (c) The inner heme binding site in the four-helical transmembrane bundle of TMD. (d) The highly conserved coordinating histidines for hemes and oxygen. The hDUOX1 sequence was aligned with 106 NOX/DUOX homologs. The top

10 kT

NBD

sequence represents the hDUOX1 sequence. The bottom consensus sequence represents the most common amino acid at each position. The height of the residues in the logo plot stands for their frequency in the sequence alignment. The logo plot of residue frequencies was generated by the web tool (weblogo.berkeley.edu). (e) The FAD binding site. FAD is shown as spheres. The surface representation of DH is colored according to electrostatic potential. (f) The NADPH binding site. NADPH is shown as spheres. The interacting residues are shown as sticks. The adjacent lipid is colored in purple. The surface representation of DH is colored according to electrostatic potential

30

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex

the transmembrane region of DUOX1 [16, 17]. The two hemes are sandwiched in the four-helical transmembrane bundle (M2–M5), which is nearly orthogonal to the lipid bilayer (Fig. 30.4a). This elegant design for electron transfer across lipid bilayer has been found in different oxidases [31]. Two pairs of histidines of M3 and M5 helices coordinate two hemes: H1144 and H1238 for the outer heme, and H1130 and H1225 for the inner heme, respectively (Fig. 30.4b–c). These four histidines have been previously found to be absolutely conserved among the NOX family proteins (Fig. 30.4d) [32]. Therefore, other NOX members likely use a similar set of histidines for heme coordination. In fact, all members of the NOX family show the characteristics of hexa-coordinated heme with the Soret γ-band at the 414 nm and reduction by dithionite shifts the peaks to the 427 nm [27, 33, 34]. In particular, the importance of these histidines is emphasized by the disease-causing mutations (NOX2 in CGD, and DUOX2 in the hypothyroidism) [32, 35, 36] and by the previous mutagenesis experiments [37]. Adjacent to the outer heme, a pocket with a spherical electron density inside was identified in both human and mouse DUOX1 [16, 17]. This small cavity is surrounded by the hydrophilic residues from M2 and M3 helices: the R1087, H1148, and H1144 in hDUOX1 (Fig. 30.4b). In csNOX5, a water molecule was identified in this small cavity, and mutations of the corresponding residues impaired the re-oxidation of hemes in the TMD of csNOX5 by oxygen [27], suggesting that this cavity is the binding site of dioxygen substrate. The bound dioxygen might interact with the adjacent positively charged residues and the propionate group of heme rather than directly with the iron of the outer heme. These positively charged residues are strictly conserved in the NOX family (Fig. 30.4d) [32]. Therefore, it is proposed that superoxide is ultimately generated by electron transfer from the reduced outer heme to oxygen but without direct contact between oxygen and the iron of outer heme. Adjacent to the inner heme, the cytosolic DH domain docks onto the cytosolic surface of TMD in all of the available structures except the dimer-of-dimer configuration of mouse DUOX1-DUOXA1 complex with a highly flexible DH domain [16, 17]. The high-resolution structures of hDUOX1 show clear densities for both the FAD cofactor and the substrate NADPH, which are sandwiched at the interface between DH and TMD [17]. The FBD subdomain of DUOX1 DH shapes a pocket to bind the riboflavin group of FAD (Fig. 30.4e), similar to the csNOX5 and other ferredoxin NADP+ reductases (FNR) [27, 38, 39]. Moreover, extensive interactions are also observed between FAD and TMD in the complete structure of the DUOX1 subunit (Fig. 30.4e). The highly acidic phosphate group of FAD interacts electrostatically with the

513

positively charged residues of TMD (R1214 and R1131), and the ribose of FAD makes hydrogen bonding with the D1128 of TMD (Fig. 30.4e). The extensive interactions between FAD and TMD lead to the proximity between the adenosine group of FAD and the inner heme for electron transfer (Fig. 30.4e). The extensive interactions among FAD, DH, and TMD tightly coordinate FAD in the DUOX enzyme, which is consistent with the high affinity of FAD on NOX proteins [16, 40–42]. When NADPH substrate was supplemented into the cryoEM sample, its electron density was clearly observed at the interface of the NBD and TMD (Fig. 30.4f). NADPH binds in the negatively charged crevice of NBD (Fig. 30.4f). In addition, extensive interactions are also observed between NADPH and the preM1 of TMD in the complete structure of the DUOX1 subunit (Fig. 30.4f). The adenosine ring of NADPH not only forms hydrogen bonding with the hydrophilic residues on the preM1 (E1039 and N1040) but also makes cation-π interaction with the R1036 on preM1. Remarkably, the phosphate group of NADPH ribose forms multiple electrostatic interactions with basic residues from both NBD (R1495 and R1424) and the preM1 of TMD (R1036), explaining why NADPH has a higher affinity than NADH [43]. In addition, a lipid molecule was observed adjacent to the NADPH in these structures (Fig. 30.4f). Lipids have been suggested to play an important role in the regulation of NOX activity [41, 44, 45], but the identity and function of this lipid await further investigation. Taken together, both DH and TMD of DUOX1 contribute to the binding of FAD and NADPH.

5

Electron Transfer Pathway

Central to the function of DUOX is its capability to transfer electrons across the membrane. The structures of DUOX1 allow us to visualize the complete electron transfer pathway. On the cytosolic DH domain, the coordinated NADPH acts as the electron donor; electrons are transferred to the neighboring FAD cofactor, relied to the inner heme, and then to the outer heme in the TMD; the electrons reach O2 substrate to generate superoxide anions initially and then hydrogen peroxide finally. The edge-to-edge distances from the NADPH to the outer heme are shown in Fig. 30.5a. Along the pathway, the distance between NADPH and FAD is larger than 7.5 Å in all of the DUOX structures available (Fig. 30.5a). This distance is larger than the completely contactable 3.2 Å distance in the active state of the canonical spinach FNR (Fig. 30.5b) (sFNR, PDB ID: 1QFZ) [38]. Therefore, the conformation of DH is not optimal for highly efficient electron transfer. A mobile NBD has been found to change the conformation of DH to shorten the distance between NADPH and FAD [46]. The edge-to-edge

514

J.-X. Wu et al. a

Tunnels

e

b

Loop C

Loop A

oxygen-reducing center FAD

8.2 Å

NADPH

NTP

B

Out

C

Loop E FBD

Outer heme 3.7 Å 6.7 Å

F1097

A

NBD

18.4 Å

D

TMD

3.2 Å

hDUOX1 (7D3F)

Inner heme 3.9 Å

In

M1M6 M1 M6 M2 M3

8.2 Å

FAD

3.2 Å

FAD

NADPH

f

M5 M4

M0

Loop E

M1 M2

NADPH

Q1245 FBD

R1087 NBD

R1062 A

DH

R1248

14.3 Å

sFNR (1QFZ)

c

Loop C

oxygen-reducing center

Loop A

g

Loop C

NTP

NTP Loop C

F1167

Loop A

Loop A

Out Loop E

F1155

B

I1244

Out

I1083

Loop E Outer heme

Loop E

M2

Loop C

h F1155

W1184

In

In

I1083

W1185

C

I1244 I1243

d

P1191

Exposed oxygen-reducing center

F1241

Loop E i

Out

Out

M0

M3

M4 Loop C

F1186

F1155 V1152

L602 D

V1149

L1199

I1244

M3

In

In

hDUOX1 (7D3F)

csNOX5 (5O0X)

Fig. 30.5 Electron transfer pathway in the DUOX1 complex. (a) The electron transfer pathway of DUOX1 (PDB: 7D3F) with key components shown in sticks and the edge-to-edge distances labelled.

-10 kT

M4

Loop E 10 kT

(b) The edge-to-edge distances between NADPH and FAD of hDUOX1 (top) and the active sFNR (bottom). The distances between Cα atoms of the Arg1337 (Lys110 in sFNR) and Cys1520 (Cys266 in sFNR) were

30

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex

distance between the FAD and inner heme is around 4 Å (Fig. 30.5a), which might allow direct electron transfer between them. In the TMD, the edge-to-edge distance between two hemes is about 7 Å in all of these solved structures, which is too large for efficient electron transfer, suggesting there might be a structural element that mediates the electron transfer. There are several hydrophobic residues between two hemes. Notably, the phenyl ring of F1097 is within Van der Waals contact distance from both hemes (Fig. 30.5a). Therefore, the two hemes together with F1097 form a tightly packed network to mediate the crossmembrane electron transfer. Finally, the electrons are transferred to the substrate oxygen at the oxygen-reducing center to produce the superoxide as the primary product in NOX family proteins [47, 48], while DUOX can further convert the intermediate superoxide anion to hydrogen peroxide as the final product. The structures of csNOX5 and DUOX1 provide some clues on the mechanism underlying their different product specificity. The structure of csNOX5 show the oxygen-reducing center is exposed to solution largely enough for the entrance of the substrate oxygen and quick release of the product superoxide, while the structures of DUOX1 show that the extracellularfacing oxygen-reducing center is completely caged below the several extracellular loops of DUOX1 (Fig. 30.5c-d), indicating the structure of extracellular loops may affect the release of superoxide. NOX family members all possess three extracellular loops: loop A, loop C, and loop E (Fig. 30.5c), meanwhile DUOX enzymes have an additional extracellular PHD domain. Sequence alignment of human NOX family members shows that the loop A of DUOX is 6 amino acids longer than that of NOX1–5 [49]. Mutations of loop A have been found to switch the product specificity from hydrogen peroxides to superoxide anions [49]. Because the oxygenreducing center of the DUOX complex is buried, the oxygen substrate and superoxide anion have to enter and exit through tunnels in DUOX. Using the oxygen-reducing center as the starting point and the probe radius at around 1 Å, four routes (tunnels A–D) have been found in the high calcium state of hDUOX1 (PDB ID: 7D3F) (Fig. 30.5e). Strikingly, tunnel A is the only path that is embraced by hydrophilic residues on M1, M2, M6, and loop E (Fig. 30.5f), while tunnels B–D are mainly hydrophobic (Fig. 30.5g–i). The negatively charged superoxide anions would be favorably attracted by the highly positively charged constriction in tunnel A, and then two

515

unstable superoxide anions probably dismutate to the uncharged hydrogen peroxide for release. It is noteworthy that tunnel A is adjacent to tunnel B, which is mainly enclosed by the flexible loops (loop A, C, and E), and flexible loop A is stabilized by the DUOXA1 NTP through their tight association (Fig. 30.5g). This agrees with that DUOX1 loop A and the associated DUOXA1 NTP were suggested to prevent the leakage of superoxide anions [49–51]. Therefore, manipulations that may alter the constrictions of tunnel A–D would affect the leakage of superoxide anion intermediate.

6

Activation and Modulation

DUOX oxidases are highly expressed in the thyroid tissue and play a vital role in the formation of the thyroid hormone, and the functional DUOX in the thyroid has been identified to be dependent on calcium [4, 52–54]. In addition, DUOX enzymes have also been found to function in other tissue with calcium-dependent activity [55, 56]. In vitro, the purified DUOX1 protein complex showed calcium-activated activity [17, 22]. Two canonical EF-hands that are capable of calcium-binding have been predicted in the sequence of DUOX oxidases. The structures of hDUOX1 at both low and high calcium states provide structural insights into how calcium modulation might be achieved [17]. The alignment of structures at both low and high calcium states shows no obvious differences in the extracellular and transmembrane layers (Fig. 30.6a). However, large conformational changes are observed at the regulatory PHLD and EF modules in the cytosolic layer upon the removal of calcium (Fig. 30.6a–d). In the high calcium state, the EF module is in an extended conformation (Fig. 30.6b) and two typical EF-hands form a large crevice to enclose α4 and the post α4 loop of the DH domain (Fig. 30.6c). In detail, the hydrophobic crevice is formed mainly by residues from both EF1 and EF2, and the hydrophobic region of α4 and the post α4 loop of the DH domain bind inside the crevice (Fig. 30.6b–c). This binding mode observed in the high calcium state is similar to that of calcineurin in the calcium-bound state [57]. The key residues on EF to chelate calcium are highly conserved in the DUOX family [17]. Some mutations on these sites of DUOX have been reported to impair the calcium activation and result in congenital hypothyroidism [21, 58, 59]. Upon the removal of

ä

Fig. 30.5 (Continued) used to indicate the closure of DH. (c) The oxygen-reducing center of DUOX and csNOX5. The extracellular loops (Loops A, C, E) are highlighted with purple (DUOX) and black (csNOX5). The associated NTP of DUOXA is shown in green. (d) The difference of the oxygen-reducing center between DUOX (buried) and csNOX5 (exposed). The surface representation of TMD is colored

according to electrostatic potential. (e) Four predicted tunnels at the oxygen-reducing center for the exchange of substrate and product with exterior solutions. The starting point at the oxygen-reducing center is indicated by a red sphere. (e–i) The close-up views of the tunnels. The surface representation of TMD is colored according to electrostatic potential. Residues surrounding the tunnels are indicated as sticks

516

J.-X. Wu et al.

DUOXA1

a

d

c PHD

DH post α4

DH post α4

α4

αC

αK αE

αJ αI

Out

α4

αH

αC

αD

αF

αG

αF

αK

αH αG

αE

αI

TMD

αD EF1

EF1

EF2

αJ

In

EF2

PHLD αD

αD

αC

αC EF1

EF1

DH

αG

Calcium

αK

αE

αH αI

EF

αE αG

αJ αF

b

EF2

αF

αH

EF DH αI

EF2

αK αJ

high-calcium state

low-calcium state

f

e

TMD R1216

K653

PHLD 17.2°

R1215

DH E1348 cytosolic layer

cytosolic layer

I1349

R674

αA

High-calcium state

low-calcium state

αA

Fig. 30.6 Calcium activation of DUOX1 complex. (a) Comparison of DUOX-DUOXA1 structure between the high-calcium state (colored, PDB: 7D3F) and the low-calcium state (gray, PDB: 7D3E). The protein complex is shown as the cartoon. The stable regions are shaded with surface representation. The regulatory PHLD and the EF module with large conformational changes are boxed. (b) Zoomed-in view of the EF-hand module highlights its conformational change upon calcium removal. The movement of EF2 between the high-calcium and low-calcium states is measured by the indicated marker (Cα atom of A894 on αJ helix). The interacting residues between EF and DH are shown as sticks, and the hydrogen bonds are indicated with dashed lines. (c) The extended EF module in the high calcium state tightly interacts

with α4 and the post α4 loop of the DH domain. The interacting residues from both EF1 and EF2 are shown as sticks. (d) The contracted EF module in the low calcium state. (e) Close-up view of PHLD highlights its conformational change upon calcium removal. The angle between αA helices in the high-calcium and low-calcium states is measured. The interacting residues between PHLD and DH or TMD are shown as sticks, and the hydrogen bonds are indicated with dashed lines. The Cα atom of the interacting residues from PHLD is shown as spheres, and arrows indicate the disrupted interaction upon calcium removal. (f) The mobile cytosolic layer in the high-calcium (colored, left) and low-calcium (gray, right) states, relative to the stable extracellular and transmembrane layer

30

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex

517

calcium, the EF module is contracted (Fig. 30.6b) and EF2 moves away from the DH domain. Therefore, the EF module is no longer tightly packed with α4 and the post α4 loop of the DH domain (Fig. 30.6b, d). We also found the adjacent PHLD rotates outwardly and is completely dissociated from both the DH and TMD, which disrupts the interaction nexus of PHLD with both the DH and TMD (Fig. 30.6e). Therefore, upon the removal of calcium, not only the inter-domain interactions of the cytosolic layer but also the docking of DH onto TMD is destabilized, resulting in a more mobile cytosolic layer in the low-calcium state (Fig. 30.6f), evidenced by the markedly altered distribution of multibody refinement [17]. The mobility of the DH domain in the cytosolic layer is negatively correlated with the electron transfer efficiency from DH to TMD. In addition, because NAPDH interacts with both DH and TMD, the motion of DH relative to TMD would reduce the binding affinity of NADPH as well. This agrees with reduced Kcat and increased Km(NADPH) in the low-calcium state [17]. DUOX1 complex was observed as both heterodimer and heterotetramer in the purified mouse complex [16] (Sun 2020), suggesting the assembly mechanism could provide a layer for modulation. However, the human DUOX1DUOXA1 structural study only shows a heterotetrameric conformation. Whether this difference is due to different protein preparation procedures or different species (mouse vs. human) remains elusive. Interestingly, functional regulation of NADPH oxidase by protein oligomerization has also been reported in the plant RbohD protein [60]. Therefore, it will be worthwhile to investigate such regulation in the context of live cells.

peptidisc compared to DUOX1 in the membrane [17], suggesting lipid bilayer may have some influence on DUOX1 activity. However, this result might be obscured by the accuracy of protein quantification in the crude membrane, since it is possible that only a fraction of functional DUOX1 protein was solubilized out of the membrane for quantification. 2. DUOX1 enzyme is dedicated to hydrogen peroxide production. According to our structure, we propose a model that the trapping of immediate product superoxide anions in the vicinity of oxygen reducing center would promote the dismutation reaction of superoxide anions to generate hydrogen peroxide [16, 17]. In contrast to this, the related NOX4 enzyme also produces hydrogen peroxide as the major product but using a distinct mechanism: the superoxide anion was stably bound by residues of NOX4 and reduced by another electron to generate hydrogen peroxide [61]. Therefore, the detailed mechanism of hydrogen peroxide generation awaits further investigation. 3. We observed that the structural changes of EF domains upon calcium binding are associated with the immobilization of the DH domain, and docking of the DH domain below TMD allows electron transfer, thus the activation of DUOX. Because the activity of DUOX is regulated by calcium, it is reasonable to propose the structural changes of EF domains are upstream of the immobilization of the DH domain. But how EF domains control the conformational stability of the DH domain remains elusive. In addition, whether the cytosolic PHLD, which also shows positional changes upon calcium binding, is involved in this calcium-dependent activation process is unknown.

7

While the canonical single-particle cryo-EM methods provide protein structures at discrete states, the biological functionality of macromolecules, such as DUOX, requires dynamic motions. Therefore, the answers to the aforementioned outstanding questions demand complementary methods such as enzyme kinetics, spectrometric and computational methods, accompanied by extensive mutagenesis studies. Moreover, DUOX proteins are implicated in several diseases and are important to human health [1, 13, 62], but there is no specific pharmacological inhibitor or activator targeting DUOX protein available. The mechanistic understanding of DUOX protein based on structural information would certainly aid drug discovery targeting related diseases, such as thyroid diseases, allergic diseases, and certain cancers [1, 13, 62].

Perspective

The current progress in structural determination of the DUOX1-DUOXA1 complex at multiple assembly states [16] and different functional states [17] markedly advance our understanding of the DUOX enzyme. However, several open questions about the molecular mechanism of DUOX remain elusive. 1. What is the structure of fully active DUOX? We observe a larger distance (~4 Å) between the FBD and NBD of DUOX1 (PDB ID: 7D3F) than the related sFNR protein (PDB ID 1QFZ), suggesting the electron transfer efficiency in the current structure of DUOX1 is not as optimal as sFNR. To reach higher electron transfer efficiency, a conformational change of DH, which will bring FBD and NBD closer, is required. Alternatively, a bridging protein residue or cofactor needs to be placed between FAD and NADPH to mediate the efficient electron transfer. One related observation is the lower activity of DUOX1 in

Acknowledgements The work is supported by grants from Ministry of Science and Technology of China (National Key R&D Program of China, 2022YFA1303000 to L.C.), National Natural Science Foundation of China (91957201, 32225027, and 31821091 L.C., 31900859 to J.-X. W.), China Postdoctoral Science Foundation (2016 M600856,

518 2017 T100014, 2019 M650324, and 2019 T120014 to J.-X.W.). J.-X.W. is supported by the Boya Postdoctoral Fellowship of Peking University and the postdoctoral foundation of the Peking-Tsinghua Center for Life Sciences, Peking University (CLS). J.S. is supported by NIH (R00 HL143037 and R01 GM141357) and American Lebanese Syrian Associated Charities (ALSAC).

References 1. De Deken X, Corvilain B, Dumont JE et al (2014) Roles of DUOXmediated hydrogen peroxide in metabolism, host defense, and signaling. Antioxid Redox Signal 20:2776–2793 2. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev 87:245–313 3. Dupuy C, Ohayon R, Valent A et al (1999) Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. Cloning of the porcine and human cdnas. J Biol Chem 274:37265–37269 4. Leseney AM, Deme D, Legue O et al (1999) Biochemical characterization of a Ca2+/NAD(P)H-dependent H2O2 generator in human thyroid tissue. Biochimie 81:373–380 5. De Deken X, Wang D, Many MC et al (2000) Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275:23227–23233 6. Donko A, Ruisanchez E, Orient A et al (2010) Urothelial cells produce hydrogen peroxide through the activation of Duox1. Free Radic Biol Med 49:2040–2048 7. Hirakawa S, Saito R, Ohara H et al (2011) Dual oxidase 1 induced by Th2 cytokines promotes STAT6 phosphorylation via oxidative inactivation of protein tyrosine phosphatase 1B in human epidermal keratinocytes. J Immunol 186:4762–4770 8. Rada B, Park JJ, Sil P et al (2014) NLRP3 inflammasome activation and interleukin-1beta release in macrophages require calcium but are independent of calcium-activated NADPH oxidases. Inflamm Res 63:821–830 9. Habibovic A, Hristova M, Heppner DE et al (2016) DUOX1 mediates persistent epithelial EGFR activation, mucous cell metaplasia, and airway remodeling during allergic asthma. JCI Insight 1: e88811 10. Meziani L, Gerbe De Thore M, Hamon P et al (2020) Dual oxidase 1 limits the IFNgamma-associated antitumor effect of macrophages. J Immunother Cancer 8 11. Ashtiwi NM, Sarr D, Rada B (2021) DUOX1 in mammalian disease pathophysiology. J Mol Med (Berl) 99:743–754 12. Ris-Stalpers C (2006) Physiology and pathophysiology of the DUOXes. Antioxid Redox Signal 8:1563–1572 13. Van Der Vliet A, Danyal K, Heppner DE (2018) Dual oxidase: a novel therapeutic target in allergic disease. Br J Pharmacol 175: 1401–1418 14. Chen S, Ling Q, Yu K et al (2016) Dual oxidase 1: a predictive tool for the prognosis of hepatocellular carcinoma patients. Oncol Rep 35:3198–3208 15. Grasberger H, Refetoff S (2006) Identification of the maturation factor for dual oxidase. Evolution of an eukaryotic operon equivalent. J Biol Chem 281:18269–18272 16. Sun J (2020) Structures of mouse DUOX1-DUOXA1 provide mechanistic insights into enzyme activation and regulation. Nat Struct Mol Biol 27:1086–1093 17. Wu JX, Liu R, Song K et al (2021) Structures of human dual oxidase 1 complex in low-calcium and high-calcium states. Nat Commun 12:155 18. Luxen S, Noack D, Frausto M et al (2009) Heterodimerization controls localization of Duox-DuoxA NADPH oxidases in airway cells. J Cell Sci 122:1238–1247

J.-X. Wu et al. 19. Korzeniowska A, Donko AP, Morand S et al (2019) Functional characterization of DUOX enzymes in reconstituted cell models. Methods Mol Biol 1982:173–190 20. Deme D, Virion A, Hammou NA et al (1985) NADPH-dependent generation of H2O2 in a thyroid particulate fraction requires Ca2+. FEBS Lett 186:107–110 21. Rigutto S, Hoste C, Grasberger H et al (2009) Activation of dual oxidases Duox1 and Duox2: differential regulation mediated by camp-dependent protein kinase and protein kinase C-dependent phosphorylation. J Biol Chem 284:6725–6734 22. Ameziane-El-Hassani R, Morand S, Boucher JL et al (2005) Dual oxidase-2 has an intrinsic Ca2+-dependent H2O2-generating activity. J Biol Chem 280:30046–30054 23. Meitzler JL, Hinde S, Banfi B et al (2013) Conserved cysteine residues provide a protein-protein interaction surface in dual oxidase (DUOX) proteins. J Biol Chem 288:7147–7157 24. Meitzler JL, Ortiz De Montellano PR (2009) Caenorhabditis elegans and human dual oxidase 1 (DUOX1) “peroxidase” domains: insights into heme binding and catalytic activity. J Biol Chem 284: 18634–18643 25. Singh PK, Sirohi HV, Iqbal N et al (2017) Structure of bovine lactoperoxidase with a partially linked heme moiety at 1.98A resolution. Biochim Biophys Acta Proteins Proteom 1865:329–335 26. Nicolussi A, Dunn JD, Mlynek G et al (2018) Secreted heme peroxidase from Dictyostelium discoideum: insights into catalysis, structure, and biological role. J Biol Chem 293:1330–1345 27. Magnani F, Nenci S, Millana Fananas E et al (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci USA 114:6764–6769 28. Hanukoglu I (2015) Proteopedia: Rossmann fold: a beta-alpha-beta fold at dinucleotide binding sites. Biochem Mol Biol Educ 43: 206–209 29. Vecchio AJ, Stroud RM (2019) Claudin-9 structures reveal mechanism for toxin-induced gut barrier breakdown. Proc Natl Acad Sci USA 116:17817–17824 30. Varela V, Rivolta CM, Esperante SA et al (2006) Three mutations (p.Q36H, p.G418fsX482, and g.IVS19-2A>C) in the dual oxidase 2 gene responsible for congenital goiter and iodide organification defect. Clin Chem 52:182–191 31. Oosterheert W, Reis J, Gros P et al (2020) An elegant four-helical fold in NOX and STEAP enzymes facilitates electron transport across biomembranes-similar vehicle, different destination. Acc Chem Res 53:1969–1980 32. Kawahara T, Quinn MT, Lambeth JD (2007) Molecular evolution of the reactive oxygen-generating NADPH oxidase (Nox/Duox) family of enzymes. BMC Evol Biol 7:109 33. Iizuka T, Kanegasaki S, Makino R et al (1985) Studies on neutrophil b-type cytochrome in situ by low temperature absorption spectroscopy. J Biol Chem 260:12049–12053 34. Miki T, Fujii H, Kakinuma K (1992) EPR signals of cytochrome b558 purified from porcine neutrophils. J Biol Chem 267:19673– 19675 35. Ishibashi F, Nunoi H, Endo F et al (2000) Statistical and mutational analysis of chronic granulomatous disease in Japan with special reference to gp91-phox and p22-phox deficiency. Hum Genet 106: 473–481 36. Tsuda M, Kaneda M, Sakiyama T et al (1998) A novel mutation at a probable heme-binding ligand in neutrophil cytochrome b558 in atypical X-linked chronic granulomatous disease. Hum Genet 103: 377–381 37. Biberstine-Kinkade KJ, Deleo FR, Epstein RI et al (2001) Hemeligating histidines in flavocytochrome b(558): identification of specific histidines in gp91(phox). J Biol Chem 276:31105–31112 38. Deng Z, Aliverti A, Zanetti G et al (1999) A productive NADP+ binding mode of ferredoxin-NADP+ reductase revealed by protein engineering and crystallographic studies. Nat Struct Biol 6:847–853

30

Structural Insights into the Mechanism of DUOX1-DUOXA1 Complex

39. Aliverti A, Pandini V, Pennati A et al (2008) Structural and functional diversity of ferredoxin-NADP(+) reductases. Arch Biochem Biophys 474:283–291 40. Clark RA, Leidal KG, Pearson DW et al (1987) NADPH oxidase of human neutrophils. Subcellular localization and characterization of an arachidonate-activatable superoxide-generating system. J Biol Chem 262:4065–4074 41. Nozaki M, Takeshige K, Sumimoto H et al (1990) Reconstitution of the partially purified membrane component of the superoxidegenerating NADPH oxidase of pig neutrophils with phospholipid. Eur J Biochem 187:335–340 42. Hajjar C, Cherrier MV, Dias Mirandela G et al (2017) The NOX family of proteins is also present in bacteria. mBio 8 43. Hohn DC, Lehrer RI (1975) NADPH oxidase deficiency in X-linked chronic granulomatous disease. J Clin Invest 55:707–713 44. Knoller S, Shpungin S, Pick E (1991) The membrane-associated component of the amphiphile-activated, cytosol-dependent superoxide-forming NADPH oxidase of macrophages is identical to cytochrome b559. J Biol Chem 266:2795–2804 45. Shpungin S, Dotan I, Abo A et al (1989) Activation of the superoxide forming NADPH oxidase in a cell-free system by sodium dodecyl sulfate. Absolute lipid dependence of the solubilized enzyme. J Biol Chem 264:9195–9203 46. Hammerstad M, Hersleth HP (2021) Overview of structurally homologous flavoprotein oxidoreductases containing the low Mr thioredoxin reductase-like fold – a functionally diverse group. Arch Biochem Biophys 702:108826 47. Leto TL, Morand S, Hurt D et al (2009) Targeting and regulation of reactive oxygen species generation by Nox family NADPH oxidases. Antioxid Redox Signal 11:2607–2619 48. Sumimoto H (2008) Structure, regulation and evolution of Nox-family NADPH oxidases that produce reactive oxygen species. FEBS J 275:3249–3277 49. Ueyama T, Sakuma M, Ninoyu Y et al (2015) The extracellular A-loop of dual oxidases affects the specificity of reactive oxygen species release. J Biol Chem 290:6495–6506 50. Morand S, Ueyama T, Tsujibe S et al (2009) Duox maturation factors form cell surface complexes with Duox affecting the

519

specificity of reactive oxygen species generation. FASEB J 23: 1205–1218 51. Hoste C, Dumont JE, Miot F et al (2012) The type of DUOXdependent ROS production is dictated by defined sequences in DUOXA. Exp Cell Res 318:2353–2364 52. Ohye H, Sugawara M (2010) Dual oxidase, hydrogen peroxide and thyroid diseases. Exp Biol Med (Maywood) 235:424–433 53. Dupuy C, Deme D, Kaniewski J et al (1988) Ca2+ regulation of thyroid NADPH-dependent H2O2 generation. FEBS Lett 233:74– 78 54. Carvalho DP, Dupuy C, Gorin Y et al (1996) The Ca2+- and reduced nicotinamide adenine dinucleotide phosphate-dependent hydrogen peroxide generating system is induced by thyrotropin in porcine thyroid cells. Endocrinology 137:1007–1012 55. Razzell W, Evans IR, Martin P et al (2013) Calcium flashes orchestrate the wound inflammatory response through DUOX activation and hydrogen peroxide release. Curr Biol 23:424–429 56. Forteza R, Salathe M, Miot F et al (2005) Regulated hydrogen peroxide production by Duox in human airway epithelial cells. Am J Respir Cell Mol Biol 32:462–469 57. Ye Q, Feng Y, Yin Y et al (2013) Structural basis of calcineurin activation by calmodulin. Cell Signal 25:2661–2667 58. Peters C, Nicholas AK, Schoenmakers E et al (2019) DUOX2/ DUOXA2 mutations frequently cause congenital hypothyroidism that evades detection on newborn screening in the United Kingdom. Thyroid 29:790–801 59. Maruo Y, Nagasaki K, Matsui K et al (2016) Natural course of congenital hypothyroidism by dual oxidase 2 mutations from the neonatal period through puberty. Eur J Endocrinol 174:453–463 60. Hao H, Fan L, Chen T et al (2014) Clathrin and membrane microdomains cooperatively regulate RbohD dynamics and activity in arabidopsis. Plant Cell 26:1729–1745 61. Nisimoto Y, Diebold BA, Cosentino-Gomes D et al (2014) Nox4: a hydrogen peroxide-generating oxygen sensor. Biochemistry 53: 5111–5120 62. Faria CC, Fortunato RS (2020) The role of dual oxidases in physiology and cancer. Genet Mol Biol 43:e20190096

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP) Enzymes

31

Insights into a Transmembrane Oxidoreductase Family Related to NADPH Oxidases Wout Oosterheert, Sara Marchese, and Andrea Mattevi Abstract

Keywords

NADPH Oxidases (NOXs) generate reactive oxygen species (ROS) by transferring electrons from intracellular NADPH to molecular oxygen at the opposite side of the membrane. The ability of NOXs to transport electrons across the membrane is not unique but is shared by various other redox enzymes, such as by oxidoreductases that participate in the mitochondrial electron transport chain. In fact, because of the fold adopted by their transmembrane domain, NOXs are members of a large superfamily of specialized transmembrane oxidoreductases known as the ferric reductase (FRD) superfamily, which besides NOXs, comprises several bacterial, fungal, plant and metazoan oxidoreductases. The six-transmembrane epithelial antigen of the prostate (STEAP) family represents another iconic member of the FRD superfamily. STEAP proteins are exclusively found in metazoans and share distal sequence homology with NOXs. In this chapter, we provide an overview of the current knowledge of the structure and function of STEAP enzymes, with a particular emphasis on the molecular mechanisms underlying the enzymatic activity of STEAPs. Finally, we provide a detailed structural comparison of STEAPs and NOXs, and highlight the similarities and differences between these important classes of transmembrane oxidoreductases.

Six-transmembrane epithelial antigen of the prostate (STEAP) · Metalloreductase · Metalloenzyme · Transmembrane protein · Electron transport · Cryo-electron microscopy · Cancer antigen · Therapeutic target

Parts of this chapter have been adapted from: Oosterheert W (2021) Molecular mechanisms of tumor-cell markers: structural insights into the STEAP and tetraspanin membrane protein families – PhD Thesis, Utrecht University, The Netherlands. DOI: 10.33540/505. W. Oosterheert (✉) Department of Structural Biochemistry, Max Planck Institute of Molecular Physiology, Dortmund, Germany e-mail: [email protected] S. Marchese · A. Mattevi Department of Biology and Biotechnology ‘L. Spallanzani’, University of Pavia, Pavia, Italy

1

Identification of STEAPs as Transmembrane Oxidoreductases

In 1999, in search of new genes upregulated in metastatic prostate cancer, the Afar lab identified a gene encoding for a 339 amino-acid protein highly expressed in advanced prostate cancer [1]. Because in-silico analysis of the protein’s amino-acid sequence revealed six potential transmembrane helices, they christened this newly identified protein as ‘sixtransmembrane epithelial antigen of the prostate’, abbreviated as STEAP. In the years following the identification of STEAP (or STEAP1), three more genes encoding for proteins belonging to the same family were cloned: STEAP2, also known as six transmembrane protein of prostate 1 (STAMP1) [2, 3]; STEAP3, also referred to as tumor suppressor-activated pathway 6 (TSAP6) and pHyde [4]; and STEAP4, also known as six transmembrane protein of prostate 2 (STAMP2) and tumor necrosis factor alphainduced adipose-related protein (TIARP) [5, 6]. Although the four STEAPs share a similar transmembrane domain, STEAP2–4 contain an additional N-terminal intracellular dehydrogenase domain that is absent in STEAP1 (Fig. 31.1a). Following their identification, STEAP proteins were initially assumed to function as ion channels or transporters based on their predicted secondary structure of six transmembrane helices. However, a bioinformatics analysis revealed that the common six-helical transmembrane domains of the four STEAP proteins harbor a single heme-b binding site [7], which thus suggested a potential role of STEAPs in electron transport. Interestingly, this study

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_31

521

522

W. Oosterheert et al.

Fig. 31.1 Domain organization and molecular function of the STEAP protein family. (a) Cartoon representation of the topological arrangement of STEAP proteins in the membrane. The cofactor and substrate binding sites are depicted in the cartoon. (b) Schematic cartoon

representation of the transferrin cycle, in which STEAP3 was identified as the major ferric reductase in erythrocyte cells. Abbreviations: Tf transferrin, TfR transferrin receptor, DMT1 divalent-metal transporter 1

represented the first report of sequence homology between STEAPs and NOXs. A functional role of STEAP proteins as oxidoreductases was experimentally confirmed in 2005, when the Fleming lab identified STEAP3 as the dominant ferric reductase in the transferrin cycle in erythroid cells [8]. In this process (Fig. 31.1b), Fe3+-loaded transferrin (Tf) is endocytosed with the transferrin receptor (TfR), after which Fe3+ is released at endosomal pH. STEAP3 then reduces ferric iron (Fe3+) to ferrous iron (Fe2+), and the reduced iron is imported into the cell by metal transporter DMT1. Apo-transferrin is subsequently secreted and can be loaded with Fe3+ again. The Fe3+ to Fe2+ reduction step by STEAP3 is essential in the cycle because only Fe2+ can get

internalized. A follow-up study by the Fleming lab established that also STEAP2 and STEAP4, but not STEAP1, exhibit ferric (Fe3+) and cupric (Cu2+) reductase activity when over-expressed in human cells [9]. When combined with the observation that all STEAPs partially co-localize with the transferrin receptor, these results define three of the four STEAP proteins as enzymes that reduce extracellular iron(III) and copper(II) through electrons donated by a substrate bound to their intracellular oxidoreductase domain. Conversely, STEAP1 lacks a dehydrogenase domain (Fig. 31.1a). Hence, STEAP1 cannot function as an independent metalloreductase.

31

2

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP). . .

STEAPs in Metabolic Homeostasis

Consistent with a functional role in metal-ion reduction and transport, STEAPs, and in particular STEAP4, have been shown to be involved in cellular metabolism [10, 11]. The expression of STEAP4 is regulated by a diverse set of metabolic and inflammatory stimuli. For example, STEAP4 expression is induced by cytokines such as tumor necrosis factor α, and interleukins 6 and 17, whereas it is inhibited by leptin [5, 12, 13]. Additionally, STEAP4 expression is regulated by nutrients such as fatty acids [14]; hormones, including growth hormone and testosterone [15]; and several transcription factors [14, 16]. Functionally, the upregulation of STEAP4 through stimuli in a diabetic mouse model results in reduced inflammatory effects and better regulated glucose metabolism [16]. Conversely, the knock-down of STEAP4 leads to abnormal responses to nutrients in adipocytes, leading to tissue dysfunction and insulin resistance [17, 18]; and STEAP4-/- knockout mice develop spontaneous metabolic disease under a normal diet [14]. In accordance with these experimental observations in cell lines and animal models, single-nucleotide polymorphisms in the STEAP4 gene are associated with metabolic syndrome in a female Uygur population and in a Han Chinese population [19–21]. An epidemiological study revealed that common variations in the STEAP4 gene are unlikely to significantly contribute to a higher risk of metabolic syndrome in French Caucasians, although these variations may be associated with higher triglyceride and fasting glucose levels, and higher fat intake [22]. When taken together, these data suggest that STEAP4 is a key player in maintaining metabolic homeostasis.

3

STEAPs Are Overexpressed in Cancer and Represent Potential Therapeutic Targets

As mentioned in section “Identification of STEAPs as Transmembrane Oxidoreductases”, STEAP1 was discovered in search of genes highly expressed in prostate cancer tissue [1]. Follow-up work has since then established that STEAP1 is expressed in various human cancers, that besides prostate carcinoma include bladder, colorectal, lung, ovarian, and breast carcinoma; malignant melanoma; and Ewing sarcoma [10, 23–29]. Among the other STEAP homologs, STEAP2 and STEAP4 are also upregulated in certain cancers: STEAP2 is overexpressed in prostate cancer and may have a role in prostate cancer progression [3, 30]. STEAP4 is overexpressed in prostate cancer as well [6], and a study revealed that the therapeutic silencing of STEAP4 strongly inhibited tumor growth in two preclinical prostate cancer models in mice [31], indicating that STEAP4 is critical for

523

prostate cancer growth in vivo. In addition, recent findings indicate that STEAP4 promotes human colorectal cancer growth and that its expression correlates with poor prognosis [32]. Because of the high expression of STEAP1, STEAP2 and STEAP4, on the plasma membrane of various tumor cells, STEAP proteins represent potential targets for cancer diagnostics and for anti-cancer therapies [10, 23, 33, 34]. What is the molecular function of STEAPs in cancer cells? For STEAP4, the link between the enzyme’s metalloreductase activity and its role in cancer growth is confirmed by several reports. Specifically, STEAP4 increases oxidative stress in prostate cancer through the reduction of Fe3+ [31], because the generated ferrous Fe2+ can form reactive-oxygen species through the Fenton reaction [35]. In colorectal cancer, the inflammation-induced expression of STEAP4 results in mitochondrial iron dysregulation, which is critical for colon tumorigenesis [32]. Although STEAP1 represents the most prevalent cancer antigen among the STEAP proteins, the role of STEAP1 in cancer remains less well understood, mainly because its cellular function is unknown. Accordingly, it also remains obscure if the overexpression of STEAP1 contributes to cancer pathogenesis. Namely, STEAP1 is a biomarker for poor prognosis in prostate, ovarian and colorectal cancer patients [27, 36, 37], whereas high STEAP1 expression is associated with improved outcome of Ewing sarcoma and breast cancer patients [38, 39]. These contrasting reports therefore suggest that the beneficial or detrimental effects of STEAP1 depends on the microenvironment and metabolism of the specific tumor. Despite this ambiguous role of STEAP1 in cancer, it is widely studied as a target in anti-cancer clinical trials [40–42], highlighting its therapeutic potential [10, 23, 33].

4

STEAPs Enable Transmembrane Electron Transport Through an Array of Cofactors

In the last two decades, biochemical and structural studies have revealed how STEAPs transport intracellular electrons across the membrane to extracellular metal ions through a complex molecular interplay between their intracellular dehydrogenase domain (DHD) and six-helical transmembrane domain (TMD). The intracellular DHD is often referred to as oxidoreductase domain in the STEAP-field, but we refer to it as dehydrogenase domain here to stay consistent with the terminology used in the NOX-field. The DHD of STEAP2–4 is unique in eukaryotes, but shares sequence homology with F420H2:NADP+ oxidoreductases (FNOs) from methanogenic and sulfate-reducing archaea. FNOs are soluble, homodimeric enzymes whose primary function is to catalyze the formation of NADPH from NADP+ by hydride (H-) transfer from a reduced F420 cofactor. F420 is a flavin derivative (8-hydroxy-5-deazaflavin) that is exclusively found in

524

W. Oosterheert et al.

prokaryotes. In 2001, the crystal structure of Archaeoglobus fulgidus FNO in complex with NADP+ and F420 revealed the structural basis of the catalytic activity of FNO enzymes [43]; NADP+ binds in a protein pocket and the tricyclic deazaflavin ring of F420 stacks with the nicotinamide ring of NADP+ (Fig. 31.2a). The sequence homology between FNOs and the intracellular DHD of STEAP2–4 indicated that STEAPs could use NADPH as electron donating substrate for the reduction of extracellular metal ions. The crystal structures of the isolated oxidoreductase domain of human STEAP3 [44] and rat STEAP4 [45] were published in, respectively, 2008 and 2013, and indeed structurally verified that the DHDs of STEAP proteins adopt an architecture similar to FNO. The structures revealed how STEAPs harbor a GXGXXG/A fingerprint motif to bind NADPH through a typical Rossmann fold of six parallel β-sheets flanked by five α-helices (Fig. 31.2b, c). Contrary to FNO, the isolated DHDs of human STEAP3 and rat STEAP4 could not be crystallized with a flavin molecule stacked with NADPH. In fact, enzymatic assays showed that eukaryotic flavins interact with the NADPH-bound DHD of STEAP4 with low, non-physiological affinities (Kd > 100 μM) [45]. A follow up study showed that full-length STEAP3 in cell membranes exhibits a much higher affinity (Kd ~ 1 μM) for flavinadenine dinucleotide (FAD), providing biochemical evidence for a FAD-binding site that spans both the DHD and intracellular side of the TMD of STEAPs [46]. The same study experimentally confirmed earlier bioinformatical work [7, 47] by showing that STEAPs bind just a single heme cofactor in their TMD. Remarkably, STEAPs are the only known eukaryotic transmembrane oxidoreductases that harbor a single heme binding site; all other transmembrane oxidoreductases, including NOXs, bind two axially aligned

heme ligands in their TMD [48] (further discussed in section “STEAPs Share a Common Transmembrane Architecture with NOX Enzymes”). When taken together, these studies support a model in which STEAP2–4 transport electrons from intracellular NADPH to extracellular metal ions through FAD and a single heme cofactor. In 2018, two single-particle cryo-electron microscopy (cryo-EM) structures of human STEAP4 provided the first insights into the architecture of a STEAP enzyme containing both its DHD and TMD [49]. The structures were solved in the presence of NADPH, FAD and heme (termed cofactorbound state), as well as in the presence of NADP+, FAD, heme and substrate Fe3+-nitrilotriacetic acid (Fe3+-NTA) (termed cofactor/substrate-bound state). To obtain the structures, STEAP4 was expressed in HEK293 cells and subsequently purified from cell membranes in the detergent digitonin. Importantly, STEAP4 retained its bound heme cofactor, indicating that the STEAP4 TMD remained correctly folded during the purification procedure required for structure determination. Enzymatic assays showed that purified STEAP4 displays ferric reductase activity in the presence of FAD and NADPH, thereby defining STEAP4 as an independent oxidoreductase that does not require stoichiometric amounts of accessory proteins. The cryo-EM structures of human STEAP4 revealed that the enzyme adopts a trimeric arrangement (Fig. 31.3a). The TMD of each subunit comprises six transmembrane α-helices (h1– h6), whereas an additional short helix between transmembrane helices h3 and h4 forms the major extracellular region of the protein. The fold of the intracellular N-terminal DHD in the cryo-EM structures of STEAP4 is virtually identical to those of the isolated DHDs of STEAP3 and STEAP4 solved by X-ray crystallography (Fig. 31.2b, c) [44, 45]. Within the

Fig. 31.2 Fold of the intracellular dehydrogenase domain. (a–c) Crystal structures of A. fulgidus FNO in complex with NADP+ and F420 (a, brown, pdb 1jay), human STEAP3 residues 28–209 in complex with

NADP+ (b, red, pdb 2vq3) and rat STEAP4 residues 19–195 in complex with NADP+ (c, orange, pdb 2yjz). NADP+ and F420 are shown in stick representation with carbon atoms colored yellow

31

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP). . .

525

Fig. 31.3 Cryo-EM structures of human STEAP4 and Fab120.545bound human STEAP1. Left panel: sharpened cryo-EM density maps of STEAP4 (a, EMD-0199) and Fab120.545-bound STEAP1 (b, EMD-10735) shown parallel to the membrane as sideview. Middlepanel: Atomic model of STEAP4 (a, pdb 6hcy) and Fab120.545bound STEAP1 (b, pdb 6y9b) shown parallel to the membrane. Right panel: atomic models shown orthogonal to the membrane from the

cytoplasmic side. All membrane helices are annotated. In panel a, one STEAP4 subunit is colored orange, the other two are colored grey. . In panel b, one STEAP1 subunit is colored green, the other two are colored grey. One Fab120.545 fragment is colored magenta, the other two fragments are colored beige. Cofactors are shown in sphere representation with carbon atoms colored yellow. Domains and membrane helices are annotated

trimer, the DHD of a single STEAP4 subunit resides beneath the TMD of an adjacent subunit (Fig. 31.3a), yielding a domain-swapped arrangement of STEAP4. The cryo-EM structures of STEAP4 were solved in the presence of redox cofactors and substrate. Hence, the structures allowed for the analysis of the electron transport and metal ion reduction mechanisms of STEAP enzymes in the context of a structural framework (Fig. 31.4a). In the following sections, the major

steps of these mechanisms will be discussed. During the writing of this chapter, a preprint describing the structure of homotrimeric, human STEAP2 confirmed that STEAP2 also adopts a domain-swapped architecture and exhibits comparable molecular mechanism as STEAP4 [50]. Thus, the mechanisms described here based on structures of STEAP4 are relevant for all STEAP family members that contain an intracellular DHD.

Fig. 31.4 Molecular mechanisms of metal-ion reduction by STEAPs. (a) Schematic representation of the arrangement of cofactors in the DHD and TMD of adjacent STEAP4-subunits. Residue F359 is shown as stick. Distances between cofactors are annotated (b) Structure of human STEAP4 shown as surface. (c) FAD-binding site that bridges the intracellular DHD with the TMD. Residues W140 and R190 are shown as sticks. Membrane helix h5 is omitted from the

figure because it resides in front of FAD. (d) Heme-binding site at the extracellular membrane leaflet of the TMD. The heme coordination histidines H304 and H397 are shown as sticks. (e) Metal-ion substrate binding site above the heme. The red density represents the positive difference density attributed to Fe3+-NTA (retrieved from EMD-0199). The amino-acid residues that form the basic ring surrounding the substrate-binding site are shown as stick

526 W. Oosterheert et al.

31

4.1

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP). . .

Hydride Transfer Between NADPH and FAD

The first step of electron transport by STEAPs is the net transfer of a hydride (H-) from NADPH to FAD, which requires the stacking of the nicotinamide ring of NADPH and the tricyclic flavin ring of FAD. In the cryo-EM structures of STEAP4, NADP+ is bound to the DHD in an identical conformation as in the crystal structures of the isolated DHD of STEAP3 and STEAP4 (Fig. 31.2b, c). The FAD cofactor is anchored at the intracellular-leaflet side of the membrane with its flavin ring in a cavity formed by membrane helices h2–h5, whereas the adenine moiety of FAD interacts with a conserved tryptophan residue of the DHD (W140 in STEAP4), yielding an FAD-binding site that bridges two STEAP4 subunits (Fig. 31.4c). The nicotinamide ring of NADP+ and the flavin ring of FAD do not adopt a stacked arrangement in the structure. In fact, the distance between the closest NADP+-FAD pair is ~18 Å in adjacent STEAP4 subunits (Fig. 31.4a), making the observed arrangement incompatible with hydride transfer. As a result, a rearrangement needs to occur to yield a ring stacking conformation. Within the TMD, both faces of the FAD flavin ring are sandwiched between membrane helices, making the flavin ring inaccessible in this binding mode. In contrast, the nicotinamide ring of NADP+ in the DHD is accessible, so that stacking with a flavin ring, as observed in the structure of the homologous Archaeoglobus fulgidus FNO (Fig. 31.2a) [43], would not lead to major clashes within STEAP4. The presence of a tunnel in between the NADP+ and FAD-binding sites in the STEAP4 trimer between two subunits, combined with the moderate Kd of ~1 μM of STEAPs for FAD [46, 49], suggest that the FAD cofactor may be mobile. When taken together, this yields a putative model in which the flavin ring of FAD diffuses from its TMD-bound conformation to a short-lived stacked conformation with NADPH in the adjacent STEAP-subunit to accept a H- ion; and then shuttles back to the TMD (Fig. 31.4a).

4.2

Electron Transport Across the Membrane from FAD to Heme

Following the acceptance of a hydride ion, the reduced flavin FADH2 is capable of transferring electrons to heme. At the extracellular leaflet side of the TMD, STEAP4 binds its b-type heme cofactor through two strictly conserved histidine residues (H304 and H397 in human STEAP4) that coordinate the central heme iron (Fig. 31.4d). The propionate arms of the heme orient towards the extracellular space, whereas the vinyl groups point towards the center of the membrane. The heme is surrounded by membrane helices h2–h5 and is axially aligned above the flavin ring of the FAD cofactor. The

527

dimethyl benzyl moiety of the flavin ring points towards the vinyl-4 group (Fisher nomenclature) of the heme, with a minimum distance of 9.6 Å (Fig. 31.4a), making the observed arrangement consistent with electron transfer between two cofactors in a protein scaffold [51, 52]. Interestingly, a phenylalanine residue (F359) is located right in between the FAD and heme and resides within 4 Å of both cofactors (Fig. 31.4a). These data lead to the following model for transmembrane-electron transport by STEAP4: the reduced FADH2 injects an electron into the TMD; the electron then travels through F359 and is accepted by vinyl-4 of the heme [49]. Other STEAP homologs harbor leucine, phenylalanine or tyrosine at the equivalent position of F359, which suggests that an aromatic residue is not essential for electron transfer. In conclusion, the STEAP4 structures highlight how STEAPs employ their TMD-bound FAD to perform transmembrane electron transport with just a single heme cofactor.

4.3

Metal Ion Reduction

After receiving an electron from FAD, the central heme iron adopts the reduced state (Fe2+), making the heme cofactor capable of reducing extracellular Fe3+ and Cu2+ ions. Although the two available STEAP4 structures were solved in both the absence and presence of substrate Fe3+-NTA, both structures are virtually identical within their determined resolution limits, indicating that Fe3+-NTA does not induce any substantial conformational changes in STEAP4. However, a difference map between the reconstructions with and without substrate revealed a highly significant difference density for Fe3+-NTA, above the propionate moieties of the heme cofactor. Substrate iron binds in a ~16 Å wide ring of mainly basic amino acids, such as arginine and lysine residues (Fig. 31.4e, Fe3+-NTA density colored red). The binding of non-complexed Fe3+ and Cu2+ ions in this highly positively charged ring would lead to charge repulsion, suggesting that these ions bind to STEAP4 while they remain complexed to a negatively charged chelator like NTA, or citrate in a physiological setting. This observation is further supported by redox-chemistry data on STEAP1 [53]. The basic amino acid ring may therefore represent a second coordination shell and facilitate metal-ion reduction by (1) positioning the metal-chelator complex towards the heme, and (2) polarizing the metal-chelator complex to make it more prone to electron uptake [49]. Because the distance between the heme and the Fe3+-NTA density in the structure is >6 Å, the heme likely does not form a covalent intermediate with the chelated metal-ion substrate during the catalysis, defining an outer-sphere reduction mechanism. All in all, because the basic ring above the heme-binding site is a conserved feature of all proteins within the STEAP family, it can be concluded

528

W. Oosterheert et al.

that STEAP enzymes exhibit a common mechanism of metalion reduction [49, 50, 54].

5

The Elusive Cellular Function of STEAP1

Although the molecular mechanisms underlying the enzymatic activity of STEAP2–4 have been established, STEAP1 lacks an intracellular DHD and hence cannot function as an independent metalloreductase. In fact, based on its predicted secondary structure, STEAP1 has often been described as a channel or transporter protein in the literature [1, 23, 55, 56]. However, a pioneering study revealed that detergent-purified STEAP1 binds a single heme [53], similar to the other STEAPs. In addition, its heme-redox potential of -114 to -118 mV [53] is compatible with the reduction of Fe3+ and Cu2+ ions. Further evidence for a role for STEAP1 in transmembrane-electron transport and metal-ion reduction emerged from the cryo-EM structure of human STEAP1 (Fig. 31.3b) [54]. The structure showed that STEAP1 adopts a trimeric arrangement that is highly similar to that of STEAP4 (Fig. 31.3a, b), with a comparable heme-binding site. The reductase-like architecture of STEAP1 thus suggested a functional role in transmembrane-electron transport and metal-ion reduction, albeit STEAP1 would require a source of intracellular electrons. Interestingly, STEAP1 is often co-upregulated with STEAP2 in cancer [10] and both proteins can be co-purified in detergent [53], suggesting that they may form a physiologically relevant oligomer. Additionally, the domain-swapped architecture of STEAP4 (Fig. 31.3a), with the intracellular DHD of a STEAP4 subunit positioned beneath the TMD of an adjacent subunit, provides further clues for functional heterotrimers of STEAP1 in complex with STEAP2–4. Namely, in a hetero-oligomeric complex, the TMD of STEAP1 could receive electrons from NADPH bound to the domain-swapped DHD of

STEAP2–4. Indeed, it was revealed that a fusion protein, comprised of the DHD of STEAP4 and the TMD of STEAP1, exhibits ferric reductase activity when expressed in HEK293 cells [54]. Thus, the TMD of STEAP1 is capable of transmembrane-electron transport and Fe3+-reduction in a cellular setting. What could be the physiological function of heterotrimers of STEAP1 with STEAP2–4 family members? In homotrimers of STEAP2, STEAP3 or STEAP4, all three subunits are capable of reducing metal ions, whereas a STEAP1 homotrimer is enzymatically inactive due to the absence of an NADPH-binding site (Fig. 31.5). Accordingly, STEAP-heterotrimers that contain STEAP1 would harbor fewer active subunits per trimer and thus would display lower iron reduction rates than STEAP2–4 homotrimers (Fig. 31.5). The incorporation of STEAP1 in heterotrimers with other STEAP family members could therefore represent a cellular mechanism to prevent iron overload that leads to severe cell damage. However, this model remains highly speculative and should be confirmed or refuted in future studies, thereby focusing on the functional association of STEAP1 with STEAP family members and other unidentified partner proteins in cancer tissue. Of interest, the structure of STEAP1 was solved in complex with the Fab fragment of a monoclonal antibody (mAb120.545) (Fig. 31.3b), which facilitated the cryo-EM structure determination process [54]. Humanized variants of mAb120.545 have been utilized in clinical trials for the imaging and therapeutic treatment of prostate cancer [40, 57, 58], underlining the potential of STEAP1 as a therapeutic target. Consistent with mAb120.545 binding to cancer cells, the three Fab fragments bind to the extracellular region of the STEAP1 trimer (Fig. 31.3b). Interestingly, the Fab fragment also inhibits the cellular ferric reductase activity of the fusion protein comprising the DHD of STEAP4 and the TMD of STEAP1 [54]. Thus, antibodies that inhibit the

Fig. 31.5 Putative model of STEAP hetero-trimerization. Schematic cartoon representation of different combinations of STEAP homo- and heterotrimers. Arrows depict active electron transport and metal-ion reduction. A STEAP1 homo-trimer exhibits no metalloreductase activity

31

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP). . .

enzymatic activity of STEAPs may be explored in therapeutic strategies to target STEAP proteins in cancer.

6

STEAPs Share a Common Transmembrane Architecture with NOX Enzymes

How do the molecular mechanisms of STEAP proteins relate to those of the distantly related NOXs? STEAP and NOX enzymes are emblematic members of the FRD superfamily. Despite the important role of NOX enzymes in human physiology, which sees them starring in host defense [59], cell differentiation, senescence, and apoptosis [60] and despite their broad involvement in inflammatory and oncogenic diseases, their mechanism of action at the molecular level remained undiscovered for a long time, in part because of the lack of high-resolution models of the full-length enzymes. Only in 2017, the crystal structures of the TMD and of the DHD domain of the NADPH oxidase isoform 5 (NOX5) provided, for the first time, a detailed model that unveiled the architecture and the reductase mechanism of NOX enzymes at the molecular level [61]. This also represented, to the best of our knowledge, the first atomic model of any member of the FRD superfamily. Because it was difficult to obtain a well-behaving full-length NOX protein for structural and mechanistic investigation, it was reasoned that the most attractive system for structural studies was a simpler NOX isoform, and hence the single-subunit calcium-regulated NOX5 was selected. Additionally, Cylindrospermum stagnale NOX5 (csNOX5), bearing a very notable 40% sequence identity to human NOX5, was found to be the most promising ortholog for structural studies. A ‘divide and conquer’ approach was embraced by solving the crystal structures of the two individual domains forming the TMD and DHD catalytic core typical to all NOXs, to get over the issues faced with the full-length protein. NOXs and STEAPs enzymes are functionally and evolutionary related transmembrane oxidoreductases, with a shared mechanism noted as electron hopping [62] that sees the transport of intracellular electrons across the cellular membranes towards two different substrates, a difference that delineates their own distinct and unique roles in human physiology. On one hand the structures of STEAP4 and csNOX5 highlight common features conserved among the two families of enzymes, on the other hand the two structures reveal critical differences [48]. Indeed, although STEAPs and NOXs employ identical, non-covalently bound cofactors to facilitate the electron transfer across membranes, the two structures show crucial differences in cofactor arrangements and binding stoichiometry. Contrary to STEAP4 that recruits NADPH through an N-terminal FNO-like DHD, NOX5 coordinates the substrate

529

NADPH through a C-terminal ferredoxin-NADP+ reductaselike (FNR-like) DHD. Additionally, as already mentioned in section “STEAPs Enable Transmembrane Electron Transport Through an Array of Cofactors”, the cryo-EM structure revealed STEAP4 to adopt a trimeric, domain-swapped architecture [49], whereas the crystal structure of the TMD and DHD of NOX5 did not exhibit evidence for physiologically relevant oligomerization states. Interestingly, although STEAP4 and NOX5 share only 16% sequence identity, in line with their different role and mechanism, a structural superposition of the two transmembrane domains reveals that they share the same topology, with helices h2–h5 adopting a remarkably similar orientation and conformation (Fig. 31.6a). Conversely, helices h1 and h6 do not present a comparable orientation. Helices h2–h5 are fundamental for transmembrane electron transport because they form the fourhelical core that binds the di-heme motif and the FAD/heme motif in NOX5 and STEAP4, respectively. An overlay of the conserved four-helical core (Fig. 31.6b) shows in detail that the heme of STEAP4 resides at the same membrane depth as and presents the same orientation of the outer heme of NOX5, with their planes orthogonal to the lipid bilayer. Accordingly, the central heme irons are coordinated by a pair of histidine residues located at equivalent positions in the outer membrane leaflet side of helices h3 and h5. These histidine residues are conserved in all the members of the STEAP and NOX families, as well as in other proteins belonging to the FRD superfamily [47], pointing out a common mechanism for outer-hemes binding. Remarkably, an overlay of the intracellular leaflet side of the four-helical core shows that the inner heme of NOX5 and other NOXs binds at a similar height as the FAD of STEAP4 (Fig. 31.6c). Differently from the outer leaflet side and instead of the histidine residues that coordinate the iron-heme of NOX5, STEAP4 harbors arginine and glutamine residues at the equivalent position, which coordinate the phosphates and flavin ring of FAD, respectively. These studies confirm that STEAPs evolutionary diverged from the other FRD-family enzymes by losing the histidine residues that coordinate the heme cofactor in the intracellular membrane leaflet side [46]. Instead, the second heme-binding site diverged into a flavin binding site, even though the same four-helical core architecture to coordinate the cofactor is retained. This is a rare case where critical amino acid substitutions and local conformational changes allow a cofactor swap between two structurally and functionally conserved scaffolds. Moreover, STEAP4 and NOX5, with their shared cofactors-binding four-helical bundle that enables the transmembrane electron transport and reduction of their substrate at the extracellular membrane side, are protagonists of an outer sphere reduction mechanism unique among known heme-depending proteins that traditionally function through

530

W. Oosterheert et al.

Fig. 31.6 Structural resemblances between NOX and STEAP enzymes. (a) Superimposition of the C-alpha traces of the transmembrane domains of C. stagnale NOX5 (forest green, pdb 5O0T) and human STEAP4 (orange, residues 192–454, pdb 6HCY) viewed parallel to the membrane as sideview. The helices and cofactor-binding sites are annotated (outer heme of csNOX5 in ruby, outer heme of hSTEAP4 in yellow, inner heme of csNOX5 in ruby, FAD of hSTEAP4 in yellow). (b) Overlay of the outer heme-binding pockets of csNOX5 and

hSTEAP4 viewed orthogonal to the membrane from the extracellular side. The amino acid residues coordinating the iron atom of the heme cofactors are indicated. (c) Overlay of the binding sites of the inner heme in csNOX5 and of the inner FAD of hSTEAP4 viewed orthogonal to the membrane from the cytoplasmic side. The amino acids residues coordinating the iron-heme of csNOX5 and the phosphates and the flavin ring of FAD of hSTEAP4 are indicated

a covalent intermediate between heme-iron and ligand/substrate. Indeed, as mentioned in section “STEAPs Enable Transmembrane Electron Transport Through an Array of Cofactors”, STEAP4 provides a favorable chemical environment at the extracellular leaflet side for the reduction of its substrate (Fe3+), through a ring-shaped arrangement of positively charged amino acids residues that can facilitate metal reduction; on the one hand by controlling the substrate position with respect to heme, and on the other hand by making Fe3+ more susceptible to the electron uptake [49] (Figs. 31.4e, 31.7a). Similarly, the outer heme of NOXs transfers the electrons to its substrate, namely molecular oxygen. In this regard, an accurate examination of the csTM NOX5 structure revealed a compelling feature: a

small cavity that is located above the outer heme, occupied by a highly ordered water molecule [61] (Fig. 31.7b). This water molecule is lined by the proprionate 7 of the outer heme (Fisher nomenclature) and by the amino acid residues R256, H317 and the iron-coordinating H313. This hydrogenbonding environment suggests that the cavity-bound water molecule occupies the position of the substrate. Moreover, the presence of a positively charged amino acid residue as R256 can electrostatically promote the catalytic reduction of the dioxygen molecule with the subsequent production of superoxide, as typically observed in other oxygen-reducing enzymes [63]. Finally, mutagenesis and kinetic experiments have strenghtened the central role of the two residues R256 and H217: the reoxidation of chemically reduced

31

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP). . .

Fig. 31.7 Substrate-binding sites in the outer leaflet membrane side of STEAP4 and NOX5 enzymes. (a) A ring-shaped arrangement of positively charged amino acids residues facilitate the reduction of the hSTEAP4 substrate Fe3+. (b) A highly ordered water molecule is present

csTM NOX5 was indeed found to be improved in mutants R256S and H317R [61]. All these observations together let to conclude that, similarly to STEAP4, NOXs’ substrate interacts non-covalently with the heme prosthetic group and the surrounding hydrophilic amino-acid residues. Therefore, product formation is characterized by a, unique to NOXs and STEAPs, outersphere reduction mechanism, which does not require the covalent binding of the substrate to the heme. For a more broad perspective on the structure, mechanism and evolution of transmembrane oxidoreductases in general, we refer the reader to our previous review covering this topic [48]. Overall, STEAPs and NOXs are atypical oxidoreductases, characterized by distinctive features: indeed, they catalyze the reduction of different substrates, have a low sequence identity, exhibit a different cofactor arrangement and stoichiometry, and recruit NADPH through unrelated DH domains. However, they share a highly similar cofactor-binding fourhelical bundle that enables electron transport, and they exemplify a rare case where critical amino acids substitutions allow a cofactor exchange between two very well conserved structural and functional scaffolds. Additionally, their singular outersphere substrate reduction mechanism makes them unique heme-depending proteins that function without forming a covalent intermediate between the heme and the substrate.

531

in a cavity of csNOX5 lined by the outer heme and by the amino acid residues R256, H317, and H313 coordinating the iron-heme. Hydrogen bonds are indicated with grey dashed lines. The water molecule is indicated

7

Outlook

More than two decades after the identification of the first STEAP protein in prostate cancer cells, the four enzymes of the STEAP protein family have emerged as important players in cellular metal homeostasis. It furthermore became clear that STEAPs represent attractive therapeutic targets because of their high expression in various cancers. Biochemical and structural studies have revealed a molecular model for understanding the metalloreductase activity of STEAP2–4, and have provided pivotal insights into a potential enzymatic function of STEAP1. Importantly, the mechanism of hydride transfer from NADPH to FAD and and the subsequent electron transfer to the outer heme in STEAPs could be relevant for elucidating the electron transport mechanisms of NOX1–3; and how electron flow in these enzymes is regulated by cytosolic components and heterodimer formation with p22phox. Within the STEAP field, the challenge now is to uncover how STEAP proteins exert their enzymatic activity in the context of partner proteins in a crowded cellular environment. In contrast to NOXs, for which several regulatory mechanisms have been described, it remains to be established how the metalloreductase activity of STEAPs is regulated, and how it is prevented that STEAPs leak electrons in the membrane milieu in the absence of metal-

532

ion substrates. We envision that a detailed discription of the molecular mechanisms and regulation of STEAPs in relevant physiological and cancer tissues will ultimately yield new strategies for using STEAP proteins in cancer diagnostics and therepeutics. Acknowledgements W.O. gratefully acknowledges his PhD mentor Piet Gros and current postdoctoral advisor Stefan Raunser for invaluable guidance and support. We furthermore thank Ramon M. van den Bos for proofreading of the chapter. The research on STEAPs performed by W.O. in the laboratory of Piet Gros at Utrecht University was accomplished with financial support from by the Dutch Research Council (NWO), fund Nieuwe Chemische Innovaties (NCI) Technology Area (project no. 731.015.201). Research on NOX proteins in the laboratory of A.M. was supported by the Associazione Italiana per la Ricerca sul Cancro (AIRC; IG19808). W.O. is supported by a postdoctoral fellowship from the Alexander von Humboldt Foundation.

Competing Interests The authors declare no competing interests.

References 1. Hubert RS, Vivanco I, Chen E et al (1999) STEAP: a prostatespecific cell-surface antigen highly expressed in human prostate tumors. Proc Natl Acad Sci USA 96:14523–14528. https://doi.org/ 10.1073/pnas.96.25.14523 2. Porkka KP, Helenius MA, Visakorpi T (2002) Cloning and characterization of a novel six-transmembrane protein STEAP2, expressed in normal and malignant prostate. Lab Investig 82:1573–1582. https://doi.org/10.1097/01.LAB.0000038554.26102.C6 3. Korkmaz KS, Elbi C, Korkmaz CG et al (2002) Molecular cloning and characterization of STAMP1, a highly prostate-specific six transmembrane protein that is overexpressed in prostate cancer. J Biol Chem 277:36689–36696. https://doi.org/10.1074/jbc. M202414200 4. Zhang X, Steiner MS, Rinaldy A, Lu Y (2001) Apoptosis induction in prostate cancer cells by a novel gene product, pHyde, involves caspase-3. Oncogene 20:5982–5990. https://doi.org/10.1038/sj.onc. 1204831 5. Moldes M, Lasnier F, Gauthereau X et al (2001) Tumor Necrosis Factor-α-induced Adipose-related Protein (TIARP), a cell-surface protein that is highly induced by tumor necrosis factor-α and adipose conversion. J Biol Chem 276:33938–33946. https://doi.org/10. 1074/jbc.M105726200 6. Korkmaz CG, Korkmaz KS, Kurys P et al (2005) Molecular cloning and characterization of STAMP2, an androgen-regulated six transmembrane protein that is overexpressed in prostate cancer. Oncogene 24:4934–4945. https://doi.org/10.1038/sj.onc.1208677 7. Sanchez-pulido L, Rojas AM, Valencia A et al (2004) ACRATA: a novel electron transfer domain associated to apoptosis and cancer. BMC Cancer 4. https://doi.org/10.1186/1471-2407-4-98 8. Ohgami RS, Campagna DR, Greer EL et al (2005) Identification of a ferrireductase required for efficient transferrin-dependent iron uptake in erythroid cells. Nat Genet 37:1264–1269. https://doi.org/ 10.1038/ng1658 9. Ohgami RS, Campagna DR, McDonald A, Fleming MD (2006) The Steap proteins are metalloreductases. Blood 108:1388–1394. https:// doi.org/10.1182/blood-2006-02-003681

W. Oosterheert et al. 10. Grunewald TGP, Bach H, Cossarizza A, Matsumoto I (2012) The STEAP protein family: versatile oxidoreductases and targets for cancer immunotherapy with overlapping and distinct cellular functions. Biol Cell 104:641–657. https://doi.org/10.1111/boc. 201200027 11. Scarl RT, Lawrence CM, Gordon HM, Nunemaker CS (2017) STEAP4: its emerging role in metabolism and homeostasis of cellular iron and copper. J Endocrinol 234:R123–R134. https://doi.org/ 10.1530/JOE-16-0594 12. Chen X, Zhu C, Ji C et al (2010) STEAP4, a gene associated with insulin sensitivity, is regulated by several adipokines in human adipocytes. Int J Mol Med 25:361–367. https://doi.org/10.3892/ ijmm 13. Sparna T, Rétey J, Schmich K et al (2010) Genome-wide comparison between IL-17 and combined TNF-alpha/IL-17 induced genes in primary murine hepatocytes. BMC Genomics 11. https://doi.org/ 10.1186/1471-2164-11-226 14. Wellen KE, Fucho R, Gregor MF et al (2007) Coordinated regulation of nutrient and inflammatory responses by STAMP2 is essential for metabolic homeostasis. Cell 129:537–548. https://doi.org/10. 1016/j.cell.2007.02.049 15. Maneschi E, Morelli A, Filippi S et al (2012) Testosterone treatment improves metabolic syndrome-induced adipose tissue derangements. J Endocrinol 215:347–362. https://doi.org/10.1530/ JOE-12-0333 16. Chuang CT, Guh JY, Lu CY et al (2015) Steap4 attenuates high glucose and S100B-induced effects in mesangial cells. J Cell Mol Med 19:1234–1244. https://doi.org/10.1111/jcmm.12472 17. Moreno-Navarrete JM, Ortega F, Serrano M et al (2011) Decreased STAMP2 expression in association with visceral adipose tissue dysfunction. J Clin Endocrinol Metab 96:1816–1825. https://doi. org/10.1210/jc.2011-0310 18. Konstantopoulos N, Foletta VC, Segal DH et al (2011) A gene expression signature for insulin resistance. Physiol Genomics 43: 110–120. https://doi.org/10.1152/physiolgenomics.00115.2010 19. Nanfang L, Yanying G, Hongmei W et al (2010) Variations of Six Transmembrane Epithelial Antigen of Prostate 4 (STEAP4) gene are associated with metabolic syndrome in a female uygur general population. Arch Med Res 41:449–456. https://doi.org/10.1016/j. arcmed.2010.08.006 20. Qi Y, Yu Y, Wu Y et al (2015) Genetic variants in six-transmembrane epithelial antigen of prostate 4 increase risk of developing metabolic syndrome in a Han Chinese Population. Genet Test Mol Biomarkers 19:666–672. https://doi.org/10.1089/gtmb. 2015.0104 21. Chen X, Huang Z, Zhou B et al (2014) STEAP4 and insulin resistance. Endocrine 47:372–379. https://doi.org/10.1007/s12020-0140230-1 22. Miot A, Maimaitiming S, Emery N et al (2010) Genetic variability at the six transmembrane protein of prostate 2 locus and the metabolic syndrome: the Data from an Epidemiological Study on the Insulin Resistance Syndrome (DESIR) study. J Clin Endocrinol Metab 95: 2942–2947. https://doi.org/10.1210/jc.2010-0026 23. Gomes IM, Maia CJ, Santos CR (2012) STEAP proteins: from structure to applications in cancer therapy. Mol Cancer Res 10: 573–587. https://doi.org/10.1158/1541-7786.MCR-11-0281 24. Yang D, Holt GE, Velders MP et al (2001) Murine six-transmembrane epithelial antigen of the prostate, prostate stem cell antigen, and prostate-specific membrane antigen: prostatespecific cell-surface antigens highly expressed in prostate cancer of transgenic adenocarcinoma mouse prostate mice. Cancer Res 61: 5857–5860 25. Lee CH, Chen SL, Sung WW et al (2016) The prognostic role of STEAP1 expression determined via immunohistochemistry staining in predicting prognosis of primary colorectal cancer: a survival

31

Structure, Function and Mechanism of Six-Transmembrane Epithelial Antigen of the Prostate (STEAP). . .

analysis. Int J Mol Sci 17:1–9. https://doi.org/10.3390/ ijms17040592 26. Zhuang X, Herbert JMJ, Lodhia P et al (2015) Identification of novel vascular targets in lung cancer. Br J Cancer 112:485–494. https:// doi.org/10.1038/bjc.2014.626 27. Jiao Z, Huang L, Sun J et al (2020) Six-transmembrane epithelial antigen of the prostate 1 expression promotes ovarian cancer metastasis by aiding progression of epithelial-to-mesenchymal transition. Histochem Cell Biol 154:215–230. https://doi.org/10.1007/s00418020-01877-7 28. Kobayashi H, Nagato T, Sato K et al (2007) Recognition of prostate and melanoma tumor cells by six-transmembrane epithelial antigen of prostate-specific helper T lymphocytes in a human leukocyte antigen class II-restricted manner. Cancer Res 67:5498–5504. https://doi.org/10.1158/0008-5472.CAN-07-0304 29. Grunewald TGP, Diebold I, Esposito I et al (2012) STEAP1 is associated with the invasive and oxidative stress phenotype of ewing tumors. Mol Cancer Res 10:52–65. https://doi.org/10.1158/ 1541-7786.MCR-11-0524 30. Whiteland H, Spencer-Harty S, Morgan C et al (2014) A role for STEAP2 in prostate cancer progression. Clin Exp Metastasis 31: 909–920. https://doi.org/10.1007/s10585-014-9679-9 31. Jin Y, Wang L, Qu S et al (2015) STAMP2 increases oxidative stress and is critical for prostate cancer. EMBO Mol Med 7:315–331. https://doi.org/10.15252/emmm.201404181 32. Xue X, Bredell BX, Anderson ER et al (2017) Quantitative proteomics identifies STEAP4 as a critical regulator of mitochondrial dysfunction linking inflammation and colon cancer. Proc Natl Acad Sci USA 114:E9608–E9617. https://doi.org/10.1073/pnas. 1712946114 33. Moreaux J, Kassambara A, Hose D, Klein B (2012) STEAP1 is overexpressed in cancers: a promising therapeutic target. Biochem Biophys Res Commun 429:148–155. https://doi.org/10.1016/j.bbrc. 2012.10.123 34. Barroca-Ferreira J, Pais JP, Santos MM et al (2017) Targeting STEAP1 protein in human cancer: current trends and future challenges. Curr Cancer Drug Targets 18:222–230. https://doi.org/ 10.2174/1568009617666170427103732 35. Torti SV, Torti FM (2013) Iron and cancer: more ore to be mined. Nat Rev Cancer 13:342–355. https://doi.org/10.1038/nrc3495 36. Nakamura H, Takada K, Arihara Y et al (2019) Six-transmembrane epithelial antigen of the prostate 1 protects against increased oxidative stress via a nuclear erythroid 2-related factor pathway in colorectal cancer. Cancer Gene Ther 26:313–322. https://doi.org/10. 1038/s41417-018-0056-8 37. Ihlaseh-Catalano SM, Drigo SA, de Jesus CMN et al (2013) STEAP1 protein overexpression is an independent marker for biochemical recurrence in prostate carcinoma. Histopathology 63:678– 685. https://doi.org/10.1111/his.12226 38. Grunewald TGP, Ranft A, Esposito I et al (2012) High steap1 expression is associated with improved outcome of ewing’s sarcoma patients. Ann Oncol 23:2185–2190. https://doi.org/10.1093/annonc/ mdr605 39. Xie J, Yang Y, Sun J et al (2019) STEAP1 inhibits breast cancer metastasis and is associated with epithelial–mesenchymal transition procession. Clin Breast Cancer 19:e195–e207. https://doi.org/10. 1016/j.clbc.2018.08.010 40. Danila DC, Szmulewitz RZ, Vaishampayan U et al (2019) Phase I Study of DSTP3086S, an antibody-drug conjugate targeting six-transmembrane epithelial antigen of prostate 1, in metastatic castration-resistant prostate cancer. J Clin Oncol 37:3518–3527. https://doi.org/10.1200/JCO.19.00646 41. Nolan-Stevaux O (2020) Abstract DDT02-03: AMG 509: a novel, humanized, half-Life extended, bispecific STEAP1 × CD3 T cell recruiting XmAb® 2+1 antibody. Cancer Res 80:DDT02-03

533

LP-DDT02-03. https://doi.org/10.1158/1538-7445.AM2020DDT02-03 42. Lin TY, Park JA, Long A et al (2021) Novel potent anti-STEAP1 bispecific antibody to redirect T cells for cancer immunotherapy. J Immunother Cancer 9:1–15. https://doi.org/10.1136/jitc2021-003114 43. Warkentin E, Mamat B, Sordel-Klippert M et al (2001) Structures of F420H2: NADP+ oxidoreductase with and without its substrates bound. EMBO J 20:6561–6569 44. Sendamarai AK, Ohgami RS, Fleming MD, Lawrence CM (2008) Structure of the membrane proximal oxidoreductase domain of human Steap3, the dominant ferrireductase of the erythroid transferrin cycle. Proc Natl Acad Sci USA 105:7410–7415. https://doi.org/ 10.1073/pnas.0801318105 45. Gauss GH, Kleven MD, Sendamarai AK et al (2013) The crystal structure of six-transmembrane epithelial antigen of the prostate 4 (Steap4), a ferri/cuprireductase, suggests a novel interdomain flavin-binding site. J Biol Chem 288:20668–20682. https://doi.org/ 10.1074/jbc.M113.479154 46. Kleven MD, Dlakić M, Lawrence CM (2015) Characterization of a single b-type heme, FAD, and metal binding sites in the transmembrane domain of sixtransmembrane epithelial antigen of the prostate (STEAP) family proteins. J Biol Chem 290:22558–22569. https:// doi.org/10.1074/jbc.M115.664565 47. Zhang X, Krause KH, Xenarios I et al (2013) Evolution of the Ferric Reductase Domain (FRD) superfamily: modularity, functional diversification, and signature motifs. PLoS One 8. https://doi.org/ 10.1371/journal.pone.0058126 48. Oosterheert W, Reis J, Gros P, Mattevi A (2020) An elegant fourhelical fold in NOX and STEAP enzymes facilitates electron transport across biomembranes – similar vehicle, different destination. Acc Chem Res 53:1969–1980. https://doi.org/10.1021/acs.accounts. 0c00400 49. Oosterheert W, Van Bezouwen LS, Rodenburg RNP et al (2018) Cryo-EM structures of human STEAP4 reveal mechanism of iron (III) reduction. Nat Commun 9. https://doi.org/10.1038/s41467-01806817-7 50. Chen K, Wang L, Shen J, et al (2021) Electron relay via diffusible and protein-bound flavin cofactor in human six-transmembrane epithelial antigen of the prostate (STEAP). bioRxiv. https://doi.org/10. 1101/2021.12.23.474010 51. Moser CC, Keske JM, Warncke K et al (1992) Nature of biological electron transfer. Nature 355:796–802. https://doi.org/10.1038/ 355796a0 52. Winkler JR, Gray HB (2014) Electron flow through metalloproteins. Chem Rev 114:3369–3380. https://doi.org/10.1021/cr4004715 53. Kim K, Mitra S, Wu G et al (2016) Six-Transmembrane Epithelial Antigen of Prostate 1 (STEAP1) has a single b heme and is capable of reducing metal ion complexes and oxygen. Biochemistry 55: 6673–6684. https://doi.org/10.1021/acs.biochem.6b00610 54. Oosterheert W, Gros P (2020) Cryo-EM structure and potential enzymatic function of human six-transmembrane epithelial antigen of the prostate 1. J Biol Chem jbc.RA120.013690. https://doi.org/ 10.1074/jbc.ra120.013690 55. Challita-Eid PM, Morrison K, Etessami S et al (2007) Monoclonal antibodies to six-transmembrane epithelial antigen of the prostate-1 inhibit intercellular communication in vitro and growth of human tumor xenografts in vivo. Cancer Res 67:5798–5805. https://doi.org/ 10.1158/0008-5472.CAN-06-3849 56. Yamamoto T, Tamura Y, Kobayashi JI et al (2013) Six-transmembrane epithelial antigen of the prostate-1 plays a role for in vivo tumor growth via intercellular communication. Exp Cell Res 319:2617–2626. https://doi.org/10.1016/j.yexcr.2013.07.025 57. Carrasquillo JA, Fine B, Pandit-Taskar N, et al (2019) Imaging metastatic castration-resistant prostate cancer patients with 89Zr-

534 DFO-MSTP2109A anti-STEAP1 antibody. J Nucl Med jnumed.118.222844. https://doi.org/10.2967/jnumed.118.222844 58. O’Donoghue JA, Danila DC, Pandit-Taskar N et al (2019) Pharmacokinetics and biodistribution of a [89 Zr]Zr-DFO-MSTP2109A anti-STEAP1 antibody in metastatic castration-resistant prostate cancer patients. Mol Pharm 16:3083–3090. https://doi.org/10. 1021/acs.molpharmaceut.9b00326 59. Rada B, Leto TL (2008) Oxidative innate immune defenses by Nox/Duox family NADPH oxidases. Trends Innate Immun Contrib Microbiol Basel, Karger 15:164–187

W. Oosterheert et al. 60. Lambeth JD (2004) NOX enzymes and the biology of reactive oxygen. Nat Rev Immunol 4:181–189. https://doi.org/10.1038/ nri1312 61. Magnani F, Nenci S, Millana Fananas E et al (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci 114:6764–6769. https://doi.org/10.1073/pnas.1702293114 62. Warren JJ, Ener ME, Vlček A et al (2012) Electron hopping through proteins. Coord Chem Rev 256:2478–2487. https://doi.org/10.1016/ j.ccr.2012.03.032 63. Mattevi A (2006) To be or not to be an oxidase: challenging the oxygen reactivity of flavoenzymes. Trends Biochem Sci 31:276– 283. https://doi.org/10.1016/j.tibs.2006.03.003

Part VII Pathology

Chronic Granulomatous Disease

32

Marie José Stasia and Dirk Roos

Abstract

Keywords

Chronic granulomatous disease (CGD) is an immunodeficiency that affects phagocytic leukocytes (neutrophils, monocytes, macrophages and eosinophils). These cells comprise our first-line defense against bacteria and fungi, and patients with CGD therefore suffer from recurrent, life-threatening infections caused by these pathogens. The primary defect in CGD concerns the enzyme that generates superoxide (O2-.), which is essential in the microbicidal mechanisms of the phagocytes. This enzyme is the leukocyte NADPH oxidase, dormant in resting cells, but it gains activity upon activation of the cells by phagocytosis or by attachment of certain soluble agents. Superoxide is subsequently converted into hydrogen peroxide (H2O2) and other reactive oxygen species (ROS). The NADPH oxidase enzyme consists of five subunits, two located in the plasma membrane and intracellular vesicles (gp91phox and p22phox, together called flavocytochrome b558) and three in the cytosol (p47phox, p67phox and p40phox). Upon cell activation, these five subunits combine into one large protein complex with enzymatic activity. This chapter describes the discovery of this enzyme (subunit composition, intracellular localization, activation mechanism, and genetic background), the clinical symptoms and treatment modalities of CGD, methods to diagnose CGD at the cellular and genetic level, and the prospect of gene replacement or repair in the future.

CGD · NADPH oxidase · Infections · Inflammation · Superoxide · Hydrogen peroxide

M. J. Stasia University Grenoble Alpes, CEA, CNRS, IBS, and Centre Hospitalier Universitaire Grenoble Alpes, Chronic Granulomatous Disease Diagnosis and Research Centre (CDiReC), Grenoble, France e-mail: [email protected] D. Roos (✉) Sanquin Research, and Landsteiner Laboratory, Amsterdam University Medical Center, Location AMC, Amsterdam, The Netherlands e-mail: [email protected]

1

Introduction

Chronic granulomatous disease (CGD) is a rare primary immunodeficiency. The first description of CGD was made in 1954, during a conference of the American Pediatric Society, by Charles Janeway Sr and Robert Good, describing cohorts of boys with life-threatening and recurrent infections. It is known as a disease entity since 1957, when it was described by Robert Berendes et al. and termed fatal granulomatous disease of childhood [1, 2]. Patients with CGD suffer from infections with bacteria and fungi, as well as from obstructions of ducts such as the biliary tract, the esophagus, airways and urinary tract, by granulomata (hence the name of the disease). Somewhat later, Quie et al. observed that neutrophils from CGD patients are unable to kill bacteria, in contrast to neutrophils from healthy donors [3]. A key observation in understanding the cause of CGD was made by Beulah Holmes et al. and by Robert Baehner and David Nathan, who noticed that the so-called respiratory burst (sudden increase in oxygen consumption) of leukocytes was absent in the leukocytes from CGD patients [4, 5]. This respiratory burst is observed when leukocytes are incubated in vitro with certain agents, such as small particles (e.g. latex beads, bacteria or urate crystals) in the presence of plasma or serum, or with soluble agents such as phorbol myristate acetate (PMA), the bacterial tripeptide formyl-methionylleucyl-phenylalanine (fMLF) or leukotriene B4 (LTB4). Already in 1933, Baldrige and Gerard described a marked increase in “cell respiration” (oxygen uptake) during phagocytosis of microorganisms by canine neutrophils [6]. In 1959 it was observed that this respiration is non-mitochondrial, is accompanied by glucose consumption via the hexose

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_32

537

538

monophosphate shunt [7], and results in the formation of hydrogen peroxide (H2O2) [8]. This H2O2 production from activated phagocytes was formally validated when Paul and Sbarra and later Root et al. quantified it [9, 10]. The connection between H2O2 formation and the defense against bacteria was made when Klebanoff showed a bactericidal effect of myeloperoxidase (MPO), iodide and H2O2 on E. coli during the respiratory burst of phagocytes [11]. The discovery of superoxide dismutase (SOD) in 1969, an enzyme catalyzing the dismutation of superoxide anion (O2-.) into H2O2 [12] and its role in the generation of toxic oxygen derivatives, allowed Babior et al. to show that the initial product of the respiratory burst oxidase is O2-. rather than H2O2 [13, 14]. Superoxide is a one-electron reduction product of oxygen, and as such a reactive oxygen radical. Thereafter, it became clear that other oxygen radicals are generated by leukocytes as well, as products from subsequent O2-.–H2O2 interactions. Even more important may be the generation of hypochlorous acid (HClO), derived from chloride and H2O2, catalyzed by MPO. This compound, also known as bleach, is a strong bactericidal and fungicidal agent [15–17]. Admittedly, there is still discussion about its relevance in antibacterial or antifungal defense, since MPO deficiency has been claimed to be a rather common phenomenon without clear clinical consequences [18]. Thus, the general idea is that leukocytes from CGD patients do not show a respiratory burst, therefore do not generate reactive oxygen species needed to kill invading pathogens, and that lack of this generation permits the outgrowth of pathogens in CGD patients. Over the following years, it has become clear that this idea is correct. The assumption was that leukocytes contain an enzyme that reduces molecular oxygen to O2-., which subsequently forms H2O2. Indeed, partial purification of such an oxidase enzyme from human leukocytes supported this notion. One subject of hot debate was the question which substrate provides the reducing equivalents for the formation of O2-.. Filippo Rossi produced evidence that this is NADPH, whereas Manfred Karnovsky defended NADH [19, 20]. In the end, Rossi’s idea proved to be correct when it was demonstrated that NADPH was the preferential substrate of the phagocyte oxidase and that this activity was coupled with glucose oxidation via the hexose monophosphate (HMP) pathway [21–23]. In 1978 Anthony Segal discovered a b-type cytochrome in human and animal leukocytes with an alpha heme absorbance peak at 558 nm that was absent in most male CGD patients [24]. This protein was also present in the phagocytic vacuoles derived from human granulocytes [25]. Later, this protein was further purified and characterized as a flavocytochrome, because it contains not only two heme groups but also a flavin

M. J. Stasia and D. Roos

moiety [26]. Indeed, many years ago already, Quastel’s group had observed that addition of flavin mononucleotide accelerates the rates of oxygen consumption in leukocyte homogenates [27]. This flavocytochrome is now called gp91phox, since it is a glycoprotein (gp) of about 91 kD molecular mass; phox is derived from phagocyte oxidase. A later given name is Nox2 (neutrophil oxidase 2). Thereafter, it was demonstrated that gp91phox is associated with a membrane protein p22phox that stabilizes this heterodimer [28, 29]. Shortly thereafter, Jesaitis’ team purified the dimer to homogeneity [30]. Together, this dimer is called flavocytochrome b558 (in the original publications indicated as (flavo)cytochrome b-245, as its midpoint potential was established at -245 mV). The dimer is an integral membrane protein with cytosolic regions, transmembrane regions and glycosylated regions on the outside of the cell (For review see [31] and Chap. 4 by A.W. Segal). For several years, researchers have been trying to identify the membrane reductase that produces toxic oxygen derivatives in close correlation with this cytochrome b. Then thanks to a structural analogy study with members of the ferredoxin-NADP+ reductase family, Segal and others formally demonstrated that cytochrome b558 itself is the oxygen reductase through which O2-. is produced, and which possesses a binding site for NADPH and for FAD [32–34]. The flavin moiety is able to accept electrons from NADPH and transmit these via the heme groups to the apical side of the cell’s outer membrane, or inside a phagosome, where the electrons can react with molecular oxygen. It is highly expressed in phagocytic leukocytes, i.e. in neutrophils, eosinophils, monocytes, macrophages and microglia, and to some extent in B-lymphocytes [35–37]. Yet, the generation of superoxide by leukocytes proved to be more complicated than being catalyzed by a simple heterodimeric protein. First, the heme signal was absent in many but not in all CGD patients, especially not in female patients, indicating that perhaps more than two proteins were involved in superoxide generation [38]. This notion was substantiated by so-called complementation studies, in which cell hybrids from different CGD patients showed recovery of oxidase activity, thus proving the existence of at least three different protein origins of CGD [39, 40]. A real breakthrough was the discovery of cell-free oxidase assays, in which purified leukocyte membrane fragments were incubated with leukocyte cytosolic fractions and activated with amphiphilic agents such as arachidonic acid or sodium dodecylsulfate (SDS) or [41–47]. In this system, two, and later three cytosolic proteins were recognized to be involved in leukocyte NADPH oxidase activity [48–51]. These proteins are now known as p47phox, p67phox and p40phox. (For details about these proteins the reader is referred to the

32

Chronic Granulomatous Disease

Chap. 15 by P.M-C. Dang and J. El-Benna; Chap. 16 by H. Sumimoto, and Chap. 17 by T.W. Kuijpers and D. Roos). Following cell activation, these three cytosolic NADPH oxidase components were found to translocate from the cytosol to the membrane fraction of these cells [47]. For this translocation and the subsequent oxidase activity, the presence of flavocytochrome b558 in the membrane is required, thus firmly establishing all five proteins as true NADPH oxidase components. Indeed, CGD patients are now known with deficiencies in either one of these five proteins, proving that each of these proteins is essential for defense against pathogens [49, 52–56]. Thus, at present, the accepted mechanism of reactive oxygen species (ROS) formation by phagocytic leukocytes is as follows. In resting cells, p40phox, p47phox and p67phox are present in the cytosol as a heterotrimer, in a 1:1:1 stoichiometry [57]. This resting state is characterized by absence of oxygen radical formation. After binding of certain agents to cell surface receptors, an intracellular train of signal transduction events is started that leads to phosphorylation of all five oxidase proteins, possibly thus inducing conformation changes in these proteins (for p47phox, this has been proven [58–60]). The triggers of this reaction can be antibodies on the surface of microorganisms binding to Fcγ receptors, complement fragments C3b or iC3b, also bound to microorganisms or to antigen-antibody complexes binding to β2 integrin CR3 receptors, uncoated microorganisms via bacterial or fungal β-glucans binding to dectin-1 receptors or binding to the CD36 scavenger receptor on the phagocytes. These triggers can lead to phagocytosis of microorganisms and generation of ROS in the phagosome, but as such, phagocytosis is not required for the start of ROS generation. As a result of component phosphorylation, the cytosolic factors p40phox, p47phox and p67phox move to the flavocytochrome in the membrane and form a fivecomponent complex [61–64, 234]. The formation of this large complex is enhanced by an SH3 region in p47phox binding to the proline-rich region in p22phox, and by PX domains in p47phox and p40phox binding to phosphatidyl inositol phosphates (PtdInsPs) induced in the membrane by the cell activation [PX in p47phox binds to PtdIns(3,4)P2, PtdIns(3,5)P2, phosphatidyl serine (PS) and phosphatidic acid (PA), and PX in p40phox binds to PtdIns(3)P]. (For more details, see [65]). Finally, an important extra drive for this large complex formation is provided by the Rac1/2 system. For more details, see Chap. 18 by Y. Lin and Y.Zheng in this book. The RAS-related C3 botulinum toxin substrate 2 (Rac2) is a member of the Rho subclass of Ras superfamily GTPases required for proper immune function. Although it is part of the family of small Rho G proteins that can be ADP-ribosylated by the C3 botulinum toxin, Rac2 is a poor substrate for this toxin [66–68]. Rac1 is present in all

539

phagocytes, and Rac2 especially in neutrophils [69–73]. In resting cells, Rac-GDP is bound to its RhoGDP dissociation inhibitor (RhoGDI) [74, 75]. During cell activation, this interaction is broken, enabling exchange of GDP by GTP on the surface of Rac; this exchange is catalyzed by Rac-specific guanine nucleotide exchange factors (GEFs) (for a review see [76]). In its GTP-bound state, Rac is able to associate with p67phox (binding to a region consisting of four tetratricopeptide [TPR] motifs) [77–81] and with the plasma membrane (via the polybasic C-terminus and the isoprenyl tail) [82–84]. Altogether, this pulls the cytosolic p40phox–p47phox–p67phox complex towards the flavocytochrome in the cell membrane and possibly enables contact of p67phox with gp91phox via the activation domain (AD) in p67phox [85]. As a result, NADPH can now interact with the flavin moiety of gp91phox and donate its electrons to the oxidase system, thus starting the formation of oxygen radicals. Again, the relevance of the Rac system to this ROS formation was proven by studies with patients with a dominant negative mutation in the RAC2 gene. The neutrophils from these patients did not generate ROS when treated with fMLF or opsonized zymosan (although they did when treated with PMA, proving intactness of their NADPH oxidase and a difference in activation mechanism between various oxidase-stimulating agents), and the patients developed CGD-like symptoms [86]. The great usefulness of the cell-free system in NADPH oxidase research will be described in other chapters of this book, in particular the small revolution made possible by the use of a trimer made up of the functional sequences of each cytosolic protein [87].

2

Clinical Aspects

2.1

Symptoms and Infecting Microorganisms

Chronic granulomatous disease is a rare disease, presenting once in 200,000–250,000 newborns [88]. CGD patients manifest their symptoms usually at an early age, in the first two years of life, but presentation at later age also occurs. Most CGD patients (about 80%) are males, because the most frequently encountered cause of CGD is located in the Xchromosome-linked CYBB gene for gp91phox. However, mutations in the autosomal genes for the other components of the NADPH oxidase also cause CGD, with equal presentation in both sexes. CGD patients suffer from recurrent infections with a variety of bacterial and fungal organisms [88, 89]. These infections are often found in the skin, the airways or the gastro-intestinal tract, infected by contact, airborne or food, and in the lymph nodes that drain these organs.

540

Hematogenous spread leads to secondary infection of other organs, such as liver, bones, brain and kidneys. These symptoms are usually noted already in the first year of life as dermatitis, gastric obstruction or bloody diarrhea and failure to thrive. Sometimes, the presenting symptoms can be mistaken for pyloric stenosis, milk or food allergy or iron deficiency anemia. The most frequently encountered type of infection in CGD patients of all ages is pneumonitis [88, 89]. Typical infecting pathogens are Staphylococcus aureus, Aspergillus species, Burkholderia cepacia and enteric gram-negative bacteria. Aspergillus and other fungal species infecting the lung are difficult to combat, requiring 3–6 months of treatment with antifungal drugs. Cutaneous abscesses and lymphadenitis, typically caused by S. aureus or various gram-negative organisms such as B. cepacia and Serratia marcescens, form the next common type of infection in CGD. Recurrent impetigo, usually in the perinasal area and caused by S. aureus, is frequently seen and requires prolonged courses of oral and topical antibiotics. More deep-seated infections such as hepatic and perihepatic abscesses are also frequently encountered, usually caused by S. aureus. In such cases, the patients typically present with fever, malaise and weight loss. Another deep-seated infection is osteomyelitis, in CGD caused by blood-borne spread of S. aureus, Salmonella species or S. marcescens or by contact spread of non-fumigatus Aspergillus, such as Aspergillus nidulans, from pneumonia to the ribs or vertebral bodies [90]. Yet another persistent problem seen in many CGD patients is the occurrence of perianal abscesses, caused by Escherichia coli species, Klebsiella species, Nocardia or Candida. (For a recent review see [91]). The presence of co-infection with more than one pathogen is not unusual in CGD and highlights the need for a culture-based diagnosis. In patients with CGD, case reports of disseminated infection following BCG vaccination point to an increased susceptibility to tuberculosis, most evident in TB endemic regions. BCG vaccination is contraindicated in CGD patients [92]. Indeed, in countries, where BCG vaccination is usually performed early after birth, mycobacterial disease is common [93, 94]. Thus, the diagnosis of CGD must be made taking into account the specificity of the germs related to the geographic variability and specific practices [95]. The median age at death has increased from 15.5 years before 1990 to >30 years in the last decade. Fungal infection was the most common cause of mortality. Multiple recurring infections, rather than persistence of prior infections, are typical in CGD, emphasizing the need to establish the correct diagnosis [96]. CGD is also characterized by disorders of dysregulated inflammation, including Crohn-like inflammatory bowel disease, obstructive inflammation of the genitourinary tract, and pneumonitis resembling sarcoidosis [97–100]. The origin of

M. J. Stasia and D. Roos

this inflammatory state is not well understood. Currently, several possible explanations are considered, one of which is related to the autophagy process and another to undue T-cell activation. Autophagy is the process to degrade and recycle intracellular constituents. For this purpose, the proteins or organelles to be recycled are surrounded by a double membrane, which is then fused with lysosomes. This fusion process requires binding of the ubiquitin-like protein LC3. As such, this resembles degradation of material in a phagosome, and indeed, LC3 can also attach to phagosomes and in this way enhance killing and degradation of ingested microorganisms. LC3 recruitment to phagosomes is a ROS-dependent process, and therefore disturbed in CGD cells [101, 102]. This has two serious consequences: incomplete clearance of apoptotic cells or cell debris, which may lead to inflammation or autoimmune phenomena in various ways [103–105]. It will also lead to reduced degradation and removal of bacteria and yeasts, which may activate inflammasomes, thus generating excessive amounts of IL-1β and IL-18 and hyperinflammatory symptoms as a result [106, 107]. For more details, see [108]. Finally, CGD neutrophils also show decreased myeloid-derived suppressor cell (MDSC) activity [109], which may cause a defect in the regulation of T-cell reactivity. This too may lead to autoimmune phenomena. A characteristic histological phenomenon in CGD patients is the presence of granulomata, i.e. macrophage-encapsulated inflammatory regions. The origin of these regions, which have given their name to the disease, is not clear, but probably non-infectious [110]. These granulomata may obstruct essential portal elements, such as airways, antral region of the stomach, biliary tract, intestines, urinary tract, or ureteropelvic junction causing all sorts of related clinical symptoms [99]. Heterozygous carriers of mutations that cause CGD are usually without clinical complaints, with the exception of female carriers of X91 CGD with a mutation in gp91phox. These women possess one copy of the mutated CYBB gene and one copy of the wild-type gene in every somatic cell. Early in fetal development, one X chromosome in each female cell is inactivated. This is a random process, independent of the presence of a mutation on one of the X chromosomes. As a result, X91 CGD carriers have two populations of phagocytes: those that are gp91phox-positive and those that are -negative. When the wild-type copy of CYBB is expressed, this results in gp91phox-positive cells, which do generate ROS; when the mutated copy of the gene is expressed, this results in gp91phox-negative cells, which do not generate ROS. X91 CGD carriers therefore, have overall reduced NADPH oxidase activity. X-chromosome inactivation should theoretically result in a 50:50 ratio of expression, but in reality, the ratio follows a Gaussian distribution. Most X91 CGD carriers fall within the

32

Chronic Granulomatous Disease

range of 20–80% reduction, with those in the higher range sometimes expressing the full spectrum of clinical symptoms of CGD [61, 111]. However, a far larger proportion (up to 60%) of X91 CGD carriers suffers from skin rashes resembling discoid lupus erythematosus (DLE) and other photosensitive skin rashes [112, 113]. Also, very common in these carriers is recurrent aphthous stomatitis. Gastrointestinal symptoms and disease, autoimmune phenomena and hyperinflammation, arthralgia, polyarthritis and Raynaud’s disease as well as chorioretinitis have also been described (For a full description of these findings see [112] and [113]). This hyperresponsiveness in X91 CGD carriers may be related to the role of TLR7 in B cells. Under normal conditions, TLR7 acts as a redox sensor relevant to endosomal ROS production, cytokine production and antigen-specific antibody responses [114] as also discussed in Chap. 17 by T. Kuijpers and D. Roos. The normal suppression of TLR7 activation by endosomal NADPH oxidase activity may be a mechanism to inhibit inflammatory responsiveness against self-RNA/antigens, which may otherwise result in autoimmunity. The TLR7 gene is located on the X-chromosome and has been identified as one of the rare alleles that are insensitive to X-chromosome inactivation [115]. Thus, in female X91 CGD carriers, part of the B cells will lack ROS production and hence lack TLR7 activity inhibition because CYBB is inactivated, but TLR7 is still active in all B cells. This may explain the fact that female carriers of X91 CGD show lupus-like autoimmune manifestations. Recently it was demonstrated that the residual level of ROS correlates with interferon γ (IFN-γ)expressing T cells, suggesting a role in promoting immune dysregulation in carriers [116].

2.2

Treatment of Infections and Hyperinflammation in CGD

Bacterial and fungal infections in CGD patients need to be treated as early as possible, and in an aggressive way. Even before the cause of the infection has been properly identified, treatment with broad-spectrum antibiotics or antifungals should be started. If necessary, deep-seated infections that do not respond to treatment should be removed by surgery. Thereafter, CGD patients should be put on maintenance prophylaxis. Standard prophylaxis for CGD includes an antibacterial agent (generally trimethoprim-sulfamethoxazole), a moldactive antifungal agent (e.g., itraconazole), and sometimes recombinant IFN-γ. Trimethoprim-sulfamethoxazole has been used for decades in CGD. This agent has been proven to be safe and effective in reducing bacterial infections

541

[117]. The protective benefit of trimethoprimsulfamethoxazole may relate to its intracellular accumulation within neutrophils, which augments staphylococcal killing [118]. Trimethoprim-sulfamethoxazole is active against the majority of bacterial pathogens that cause infection in CGD, including most strains of S. aureus, the majority of community-acquired oxacillin-resistant strains, Burkholderia species, and Nocardia species [119]. Prevention of invasive aspergillosis and other filamentous fungal diseases relies on avoiding environments where high levels of fungal spores are expected (e.g., gardening and building renovations) and mold-active antifungal prophylaxis. Itraconazole prophylaxis has been shown to be safe and effective in patients with CGD [120, 121]. Although not specifically evaluated as prophylaxis in CGD, other moldactive azoles (voriconazole or posaconazole), may be considered as antifungal prophylaxis. Recombinant IFN-γ is widely used as prophylaxis in patients with CGD. In a randomized trial, prophylactic recombinant IFN-γ significantly reduced the incidence of serious infections and was beneficial regardless of age, the use of prophylactic antibiotics, or the type of CGD (X-linked or autosomal recessive) [122]. The mechanism of action is still unknown, since IFN-γ does not enhance ROS production by CGD neutrophils either in vitro or in vivo. Therefore, it is likely that its beneficial action is derived from enhancement of ROS-independent mechanisms. Doubts as to its clinical effects have been raised in a small Italian study [123]. Allogeneic hematopoietic stem cell transplantation is usually curative in CGD, and is generally accepted as a standard of care. Reduced intensity bone marrow conditioning, i.e. a conditioning regimen that uses less chemotherapy and radiation than the standard regimen, which destroys the patient’s bone marrow cells, has improved the results considerably [124, 125]. The major risk is still graft-versus-host disease. If no suitable human pluripotent stem cell (HPSC) donor can be found, granulocyte transfusions can be given as an alternative. Such donor-derived transfusions are administered to critically ill patients with neutropenia or neutrophil dysfunction and infections that do not respond to antimicrobial therapy [126, 127]. Granulocyte-colony stimulating factor (G-CSF) and dexamethasone treatment of donors increases the yield of granulocytes for transfusion, but it also recruits a distinct pool of neutrophils from the bone marrow with an altered gene expression profile [128]. Certain genes known to be involved in the antifungal immune response are downregulated in G-CSF/dexamethasone-mobilized neutrophil, leading to impaired killing in vitro by such neutrophils [129]. Instead, pooled granulocyte concentrates can also be used [130]. However, the major disadvantage of this transfusion treatment is the induction of anti-HLA antibodies, precluding future transfusions or engraftments. To overcome

542

M. J. Stasia and D. Roos

this problem, De Ravin et al. have used autologous patient neutrophils functionally corrected by mRNA transfection [131]. The value of this treatment has been shown in non-human primates but not yet in humans.

3

Cellular Aspects

3.1

Intraphagosomal and Extracellular ROS Production Are Deficient in CGD Patients

As mentioned above, ROS formation (in amounts capable of killing pathogens) depends on the presence of flavocytochrome b558, and is limited mainly to phagocytic cells in humans [132]. During an infection, the killing mechanism is mainly concentrated in the phagosome, the intracellular vesicle that contains bacteria or fungi after their phagocytosis. The NADPH oxidase complex, when assembled, transfers electrons from cytoplasmic NADPH to molecular oxygen, leading to the formation of O2-. inside the phagosome. From this primary product different downstream ROS are generated, such as H2O2 (spontaneously), peroxynitrite (by reaction of with NO) and hypochlorous acid (HOCl) (by reaction with chloride ions, catalyzed by MPO released from the cytosolic azurophil granules into the phagosome). The pH of the phagosome is generally considered to be acidic, similar to the lysosomal compartments, to favor the killing of microorganisms. However, phagosomal pH regulation is complex. An H+-ATPase present in the phagosomal membrane is responsible for this acidification. At the same time, for two electrons needed to generate two molecules of O2-. by the phagosomal NADPH oxidase complex, two protons (H+ ions) are released into the cytoplasm. Then the H+ channel Hv1 removes these protons from the cytosol by passive influx into the phagosome, which contributes to the acidification of the phagosomal compartment. However, superoxide generated by the NADPH oxidase complex is a weak base that counteracts this acidification [133]. The most acidic pH is found in macrophages, whereas in neutrophils it is in a neutral to basic range. In addition, Hv1-dependent charge compensation supports a calcium influx into the cytosol required for calcium-dependent responses such as degranulation and migration [134]. Neutrophils isolated from CGD patients show a decrease in their phagosomal pH during NADPH oxidase activation at rates comparable to those observed in macrophages [135, 136]. Thus, the alkalinization of the phagosome is strictly dependent on the activity of the oxidase, and oxidase-deficient cells that do not produce superoxide fail to alkalinize the phagosome [137]. However, no defect in phagocytosis, migration or degranulation are observed in CGD patients’ neutrophils.

Elevated levels of ROS are able to poison bacteria and fungi with more or less equal efficiency [138]. The general idea is that ROS may destroy microorganisms by targeting DNA, proteins and lipids, but this process is only partially understood [139, 140]. Bacteria in turn, have developed a number of strategies to resist killing by ROS, including detoxification of these radical species and repair of damage to molecular and cellular targets [141]. For example, the kynurenine pathway of Pseudomonas aeruginosa is able to scavenge superoxide [142]. Catalase-positive germs such as Staphylococcus, Serratia, Nocardia and Burkholderia are predominantly found to cause severe infections in CGD patients. It has been proposed that H2O2 release by catalasenegative pathogens may compensate for the defective NADPH oxidase function during phagocytosis, whereas catalase-positive organisms lack this supportive mechanism because they rapidly metabolize H2O2. However, this idea has been challenged by the retained virulence of Aspergillus and staphylococci rendered genetically deficient for catalase production [143, 144]. A direct relationship between ROS formation and killing of pathogens by leukocytes is demonstrated in the in-vitro killing test. In this test, purified leukocytes are incubated with living microorganisms, such as bacteria or fungi. At intervals in the next two hours, samples are taken, centrifuged to remove the leukocytes, and dilutions from the supernatant are plated onto culture dishes. After overnight growth, colonies can be counted to assess the number of surviving microorganisms. This test clearly shows a defect in the killing of pathogens by leukocytes from CGD patients [3]. It also shows that for the killing of bacteria like S. aureus, ROS formation is absolutely required, whereas for e.g. E. coli, both oxygen-dependent and -independent killing mechanisms exist [39, 145]. The test has been modified to distinguish between extracellular and intracellular killing [146], and to measure hyphen and conidia killing of fungi [147, 148]. Pulmonary infection due to Aspergillus is the first cause of death of CGD patients [89, 149]. Indeed, neutrophils of CGD patients are unable to kill Aspergillus fumigatus under various test conditions [147]. The activation mechanism of the leukocyte NADPH oxidase is slightly different in case of phagocytosis as compared to soluble activators, since p40phox is involved in the first but not in the latter case [51, 53]. The difference originates from the generation of PI(3)P in the membranes of phagosomes but not in the plasma membrane of leukocytes treated with soluble agents. PI(3)P can bind to the PX domain in p40phox and to trigger NADPH oxidase complex assembly [150]. This is illustrated by the description of p40phox deficiency, in which intraphagosomal ROS production in neutrophils is strongly decreased whereas extracellular production is normal. p40phox-deficient patients do not suffer from overt infections

32

Chronic Granulomatous Disease

but present with granulomatous colitis, suggesting a role of intracellular ROS production to limit inflammatory reactions [54, 151]. More details can be found in Chap. 17 by Kuijpers and Roos.

3.2

NADPH Oxidase Activity Can Be Stimulated by Soluble Activators

Phagocytosis of particulate material (microorganisms, immune complexes, urate crystals) is not necessary for ROS formation. Indeed, in case of frustrated phagocytosis by inhibitors of phagocytosis or when the particles are too big to be engulfed, ROS formation takes place normally. Also, many soluble inducers of ROS formation are known, such as PMA, fMLF and LTB4. The signaling pathway of NADPH oxidase activation by opsonized microorganisms is mainly through solicitation of immunoglobulin receptors, complement receptors, dectin receptors or Toll-like receptors, whereas activation by soluble agents differs and depends on the nature of the stimulus used. For example, PMA is a direct activator of protein kinase C whereas the chemoattractant peptide fMLF binds to a G-protein-dependent receptor that triggers a signaling cascade involving other kinases (for review see [152]). Soluble stimuli are highly useful to detect a deficiency in ROS production by CGD neutrophils. Furthermore, a differential diagnosis between “classical CGD” and Rac2 or p40phox deficiency can be assessed by comparing the stimulation of the NADPH oxidase complex by soluble and particulate agents. More details can be found in Chap. 17 by Kuijpers and Roos.

3.3

CGD Diagnosis: ROS Measurement in CGD Neutrophils

The cellular diagnosis of CGD patients can be performed at the level of oxygen consumption, superoxide generation, hydrogen peroxide formation, hypochlorous acid production or metabolic activation by leukocytes [153]. In addition, the expression of NADPH oxidase proteins can be measured (see next section of this chapter). All of these assays show a defect by CGD leukocytes. Most commonly used are tests that measure either superoxide or hydrogen peroxide generation. A detailed description of these tests can be found in [153]. Some of these tests require purification of the leukocytes, others can be performed with diluted whole blood. Superoxide has both reducing and oxidizing properties, depending on an electron-accepting or electron-donating co-reactant. Its formation can be assayed by either nitro-blue tetrazolium (NBT) reduction, cytochrome c reduction, lucigenin chemiluminescence or isoluminol

543

chemiluminescence. NBT reduction is an old and reliable test that can be performed with simple equipment. It measures microscopically the formation of formazan deposits in individual cells, which is especially useful for carrier detection in X-linked CGD (Fig. 32.1). For the cytochrome c reduction assay, horse heart cytochrome c is added to the leukocytes, and its reduction by extracellular superoxide is followed spectroscopically at 550 nm. In a parallel sample, SOD is added to prevent this reaction; the cytochrome c reduction measured in this sample is superoxide independent and should be subtracted from the assay without SOD. However, both the NBT test and the cytochrome c reduction assay are not able to confidentially analyze CGD neutrophils with hypomorphic mutations, such as those from X91- CGD patients with decreased expression of gp91phox, which show less NADPH oxidase activity than normal cells but distinctly more than cells from “classical” CGD [154, 155]. In contrast, chemiluminescence assays are very sensitive and therefore require very few cells. Lucigenin and isoluminol can react with superoxide and become excited by this process; thereafter they can release this energy in the form of light, which can be measured by luminometry. Isoluminol does not enter into cells, and therefore detects only extracellular superoxide. Hydrogen peroxide (H2O2) has mainly “oxidizing” properties. Catalyzed by a peroxidase, it can react with many substrates, of which dihydrorhodamine-1,2,3 (DHR), luminol and di-acetoxymethylester-2′7′-dichlorodihydrofluorescein diacetate (DCFH-DA) are commonly used for CGD diagnostics. DHR freely enters the cells and is oxidized intracellularly to rhodamine-1,2,3, which emits a bright fluorescent signal at 585 nm when excited by light with a wavelength of 488 nm [156]. In contrast, DCFH-DA needs to be hydrolyzed by nonspecific intracellular esterases to liberate the polar molecule 2′7′-dichlorofluorescein (DCFH), which can be oxidized by various ROS to the fluorescent molecule DCF, catalyzed by peroxidases [157] . Both DHR and DCFH oxidation is peroxidase dependent and thus relies on the activity of MPO or eosinophil peroxidase in the phagocytes. In case of MPO deficiency, a not uncommon condition, both assays will give a negative result with neutrophils, which may be misinterpreted as an NADPH oxidase deficiency, i.e., as CGD [158]. The assays are carried out in a flow cytometer and thus measures the fluorescent signal from each separate cell, which can again be used for detection of carriers of X91 CGD (Fig. 32.1A). Amplex Red (10-acetyl-3,7-dihydroxyphenoxazine) does not enter cells and therefore detects only H2O2 excreted by the phagocytes. For this reason, a peroxidase is added to the assay mixture. Amplex Red is oxidized to the brightly fluorescent resorufin, which can be detected at 580 nm after excitation at 530 nm. The assay can be carried out in a microtiter plate on a plate reader with a fluorescence detector. Also, this assay is highly sensitive, reliable and easy to

544

XCGD neutrophils

Control neutrophils

500

400

400

400

300

300

300

200

Count

500

100

200 100

0

200 100

0 0

102

103

104

105

0 0

102

103

DHR

400 350

Control

300 250 200

XCGD carrier

150 100

XCGD patient

50

Control neutrophils

35,0 30,0 25,0 20,0 15,0 10,0 5,0 0,0

8. 5 9. 5 10 .5 11 .5 12 .5 13 .5 14 .5

5. 5 6. 5 7. 5

3. 5 4. 5

0 0. 5 1. 5 2. 5

104

2 0 10

105

DHR

Amplex RedR

μM H2O2/min/106 PMN

B

C

XCGD carrier’s neutrophils

500

Count

Count

A

M. J. Stasia and D. Roos

XCGD neutrophils

3 10

4 10

5 10

DHR

T

T

Zymosan 1mg/mL

PMA 10ng/mL

PAF-fMLP 1,25μM

XCGD carrier’s neutrophils NBT positive

NBT negative

Fig. 32.1 X910 CGD diagnosis in neutrophils. (a) DHR-loaded control, X910 CGD without expression of gp91phox and X910 CGD carrier neutrophils were stimulated with PMA (grey curve) or not (red curve). (b) On the left side, kinetics of H2O2 production by control (green), X910 CGD (yellow) and X910 CGD carrier neutrophils (dark blue) stimulated with PMA. On the right side, the specific activity of

NADPH oxidase of each type of neutrophil stimulated with PMA, opsonized zymosan or platelet-activating factor (PAF)/fMLF has been plotted. (c) The NBT reduction test in control, X910 CGD and X910 CGD carrier neutrophils stimulated with opsonized latex beads. Note the dark-blue formazan deposits

perform. Both DHR and Amplex Red oxidation offer the advantage of testing cells for their response to a number of stimuli (PMA, zymosan, serum-treated zymosan, fMLF) over a prolonged time period (Fig. 32.1a and b). This is a distinct advantage when testing cells from X- CGD patients with atypical forms such as X91+ with normal expression of inactive gp91phox and X91- CGD [159–161]. Measuring this “remaining activity” is essential for predicting disease severity and survival of the patients [162, 163]. Also, in case of p40phox deficiency it can distinguish H2O2 formation induced by particles (disturbed) from that induced by soluble agents (normal) [54, 151]. One of the advantages of the DHR flow cytometric analysis of neutrophils from female relatives of X-linked CGD patients is that one can distinguish between the oxidase-positive and oxidase-negative neutrophil

populations and thus assess the percentage of these two populations. This percentage depends on the degree of X chromosome inactivation (see Sect. 5.2 of this chapter) (Fig. 32.1a). This can also be assessed by the NBT test read under the microscope but with less sensitivity (Fig. 32.1c). Luminol (5-amino-2,3-dihydro-1,4-phtalazinedione) is a ROS probe with chemiluminescent properties. It enters into cells and therefore detects both intra- and extracellular H2O2. By adding SOD and catalase, to remove extracellular O2and H2O2, the reaction can be made specific for intracellular ROS [164]. The luminol assay relies again on the availability of intracellular peroxidase and thus carries the danger again of misdiagnosing MPO deficiency for CGD. In the phagocytes from CGD patients with a deficiency of p47phox, low superoxide generation can be measured (about

32

Chronic Granulomatous Disease

545

1–5% of normal). Perhaps p67phox can translocate to cytochrome b558 in the membrane even in the absence of p47phox, as demonstrated by Dusi et al. [165] to form a non-optimal NADPH oxidase complex assembly able to produce small amounts of superoxide, as also seen in a simplified cell-freesystem that needs only Rac and p67phox for superoxide production [87]. Another possible explanation assumes a role for p47phox in the transfer of electrons within gp91phox from FAD to the heme groups and then to oxygen, leading to superoxide generation. In the absence of p47phox, FAD can only be reduced by NADPH but cannot transmit its electrons to heme and oxygen. However, if some direct electron transfer from FAD in gp91phox to oxygen in the cells from these patients is possible, this will lead to direct H2O2 production without O2- as an intermediate. Some evidence for this idea has been presented [166]. Probably as a result of this low superoxide release, CGD patients with p47phox deficiency, in general suffer from a rather mild form of CGD [162, 163, 167, 168].

for the NADPH oxidase researcher community [172]. First, it permitted to test new retrovirus constructs for gene replacement therapy in X-linked CGD [173]. Second, it highlighted functional and structurally important regions of gp91phox. Indeed, mutations lying at the origin of the X91+ and X91CGD variants were reproduced in gp91phox-KO PLB-985 cells, to study their impact on the activation process and on the structural stability of the NADPH oxidase complex (for review see [154]). Finally, the last cell model used that is closest to the neutrophil CGD model is certainly the induced pluripotent stem cells (iPSCs), thanks to Dr. Yamanaka who received the Nobel Prize for Medicine in 2012 for this work [174]. This model is very relevant for testing new genome-editing technologies as discussed in section “Future Developments” of this chapter and in Chap. 33 by G. Santilli and A. Thrasher. It also permitted for the first time to demonstrate proof of concept of protein therapy efficacy for X910 CGD [175].

3.4

4

Genetic Aspects

4.1

Genes and Mutations

Cellular Models Used to Improve Our Knowledge of NADPH Oxidase Functioning and for the Development of New Therapies for CGD

Since the early 1990s, cellular models have greatly increased our knowledge of the structure and function of the NADPH oxidase complex. In particular, cellular models from CGD patients with a deficiency in one of the subunits of the complex have been instrumental in this respect. In 1994, Leto et al. demonstrated the binding of one of the p47phox Src homology 3 (SH3) domains to the polyproline region of p22phox, thanks to the model of Epstein Barr virusimmortalized B lymphocytes (EBV-B lymphocytes) from a CGD patient with a mutation in the polyproline region of p22phox [169]. EBV-B lymphocytes from CGD patients were also widely used to test the restoration of superoxide generation in a CGD-derived B-cell line by retrovirus-mediated gene transfer [170]. A few years earlier, Tucker et al. isolated a new human diploid cell line, designated PLB-985, from a patient with acute nonlymphocytic leukemia (ANLL) in which myelomonoblasts were predominantly present [171]. PLB-985 cells possess the ability to undergo granulocytic differentiation in the presence of inducing agents such as dimethyl sulfoxide (DMSO), dimethyl formamide (DMF) or dibutyryl cyclic adenosine monophosphate (dbcAMP). Monocytic maturation is induced with PMA. The availability of pseudo-neutrophils in culture has greatly facilitated the study of the role of these cells in innate immunity. PLB-985 cells in which the CYBB gene had been knocked out (91phox-KO PLB-985 cells) is the most useful cell model

Genetic work on CGD was initiated in 1986, when RoyerPokora et al. succeeded by positional cloning in identifying the gene on the X chromosome at Xp21.1 that encodes a protein that was absent in X-linked CGD patients (i.e. no heme signal at 558 nm in the reduced-minus-oxidized spectrum) [55]. This gene was termed CYBB, cytochrome b beta chain, encoding gp91phox. Subsequent work identified CYBA at 16q21 encoding p22phox, NCF1 (neutrophil cytosolic factor 1) at 7q11.23 encoding p47phox, NCF2 at 1q25 encoding p67phox and NCF4 at 22q13.1 encoding p40phox [51, 176– 178]. The cDNA sequences from these genes revealed the amino acid sequences of the encoded proteins, which enabled the generation of (monoclonal) antibodies against these proteins or protein domains [179–184]. This has greatly facilitated research about the 3-D composition, internal interactions and working mechanism of the leukocyte NADPH oxidase. In addition, it was now possible to determine the expression of NADPH oxidase components in the leukocytes from CGD patients by Western blotting and more easily by flow cytometry (Fig. 32.2) [185, 186]. Since gp91phox and p22phox stabilize each other’s expression in neutrophils [187], absence of one will induce simultaneous low expression or absence of the other component. From then on, mutations in these genes in the genomic DNA from CGD patients could be detected. This has provided not only definite diagnosis of CGD patients, but has also enabled carrier detection and prenatal diagnosis.

546

M. J. Stasia and D. Roos

p22phox

500

p47phox

400

200 100

300 200

103

104

500

500

400

400

300 200

200

100

100

0

0 0 102

105

103

104

0 0 102

105

103

104

105

300

200 100

103

104

105

103

104

105

0 102

103

104

105

400

200

100 0

0

0 0 102

300 200

100

0

Count

400

300 Count

Count

Count

200

0 102 600

500

400

X910 CGD patient

300

100

0 0 102

Count

300

Count

Control

Count

400

p67phox

Count

gp91phox

0 102

103

104

105

0 102

103

104

105

Fig. 32.2 gp91phox, p22phox, p47phox, and p67phox expression in human neutrophils measured by FACS analysis. The red curves represent the expression of gp91phox, p22phox, p47phox, and p67phox in control and X910 CGD neutrophils labeled directly with the monoclonal antibody (mAb) 7D5 directed against external epitopes of gp91phox, or labeled

after permeabilisation with (mAb) anti-p22phox, (mAb) anti-p47phox and (mAb) anti-p67phox. The blue curves represent the labeling with irrelevant antibodies of the same subclass as the primary antibodies used (for technical details see [185])

Soon it became evident that most CGD patients (about 60–65%, but less in countries with a high incidence of consanguineous offspring) had mutations in CYBB, about 25% in NCF1 and about 6% in CYBA and 6% in NCF2. Very few patients have been found with a mutation in NCF4 [151]. In the latest survey of 2021, more than 1000 different mutations in CYBB have been collected from almost 3000 X91 CGD patients [188]. These mutations involved deletions in almost 40% of the mutations, missense and splice site mutations in about 20% each, nonsense mutations in 10%, and insertions, indels and promoter mutations in the remaining 10% of all mutations. Deletions were found not only of small size within CYBB but also of large size, removing (part of) CYBB and (part of) one or more nearby genes, such as DMD (Duchenne muscular dystrophy), DYNLT3 (Dynein, light chain, Tctextype 3), XK (X-linked Kx blood group), RP3 (X-linked retinitis pigmentosa) and OTC (ornithine transcarbamylase). Patients with such large deletions suffer not only from CGD but also from additional diseases. In the autosomal genes involved in CGD (CYBA, NCF1, NCF2 and NCF4), another 240 different mutations were found up to now (year 2021), in about 1800 patients [189]. Here too, all different kinds of mutations have been observed in these genes. Remarkably, of the estimated 1200 patients with mutations in NCF1, about 1000 carry the same mutation. This is a deletion of the NCF1 gene or parts thereof, caused by exchange of genetic material between NCF1 and one of the two pseudo-NCF1 genes, located near NCF1, one on each side [190, 191]. These pseudogenes

render genetic analysis of p47phox deficiency difficult, but not impossible (see next section of this chapter). In addition, a 1.6–1.8-kb deletion on one allele of NCF1, causing the Williams-Beuren Syndrome (WBS), occurs sometimes in combination with another mutation on the other NCF1 allele, leading to symptoms of both WBS and CGD in the patients [189, 192, 193]. NCF4 mutations give rise to a different expression of CGD, with much less bacterial and fungal infections but more auto-immune and hyperinflammatory phenomena [54, 151] (see also Chap. 17 by Kuijpers and Roos). In general, but especially evident in CYBB, mutations in these genes can either totally destroy mRNA and protein expression (X910), partly destroy both (X91-), or leave the expression intact but destroy the enzymatic activity (X91+). In case of X91-, generally originating from point mutations located in the coding sequence or in the promoter of the CYBB gene, the consequences for oxidase activity and clinical expression are usually mild [194–196]. In X910 and X91+, the oxidase activity is absent and the patients suffer from the full spectrum of CGD. The consequences of splice site mutations can be variable: sometimes some correct pre-mRNA splicing is still intact, leading to some wild-type protein expression and mild expression of the disease [197]. Finally, recent information has resulted in the recognition of some genes which, when mutated, give rise to CGD-like symptoms. First, for NADPH oxidase activity, sufficient substrate (NADPH) must be present. In neutrophils, NADPH is largely generated in the HMP pathway, with

32

Chronic Granulomatous Disease

glucose-6-phosphate dehydrogenase (G6PD) as a key enzyme. Thus, some G6PD mutations, resulting in very low NADPH levels in neutrophils, lead to low NADPH oxidase activity and, as a consequence, to CGD-like symptoms [189, 198, 199]. We have to realize, however, that not all variations in G6PD will lead to problems in the defense against pathogens: only those variations that cause a strong decrease in G6PD activity in neutrophils will have that effect, and possibly variations in other genes may also be required for that effect. Further, CYBC1 is a gene that encodes cytochrome b558 chaperone-1, a protein that is needed for maturation and full expression of gp91phox [200]. Two mutations in this gene have been discovered that cause a low expression of gp91phox and CGD as a result [201, 202]. Moreover, as mentioned above, the process of NADPH activation during phagocytosis of microorganisms by neutrophils is under control of the small GTPase protein Rac. Since 2000, several mutations have been found in the RAC2 gene [203, 204]. One of these is a dominant “loss of function” mutation that disturbs the activation process of the NADPH oxidase, thus causing CGD-like symptoms. Since Rac2 is involved in a number of different processes in leukocytes, the neutrophils from these patients show additional problems in migration and release of proteins from azurophil granules.

4.2

Genetic Diagnosis

Until recently, cellular investigations were necessary before adequate genetic analysis was possible. To make a choice which gene to analyze, one needed to know the expression of the oxidase component proteins and the indications of a possible X-linked nature of the disease or an autosomal recessive form of CGD in the family. For instance, no expression of gp91phox and a faint expression of p22phox can be seen in X910 CGD neutrophils, whereas p47phox and p67phox expression are then comparable to that of the control neutrophils (Fig. 32.2). Carrier detection for autosomal forms of CGD can best be performed at the DNA level, by searching for the familyspecific mutation, because NADPH oxidase protein expression and oxidase activity are close to normal in these carriers. Only in the X-linked form of CGD can the (female) carriers usually, but not always, be detected by a mosaic pattern of gp91phox-positive and -negative phagocytes, correlating with NADPH oxidase-positive and -negative cells. This is performed by searching for a mosaic pattern of oxidasepositive and -negative neutrophils in the NBT slide test or in the DHR flow-cytometric assay (Fig. 32.1). Alternatively, one can perform flow cytometry to detect gp91phox protein expression on the neutrophil surface with an anti-gp91phox

547

monoclonal antibody (Fig. 32.2). However, it must be kept in mind that up to one-third of all X-linked defects may arise from new mutations in germ-line cells and will therefore not always be present in the somatic cells of the mother. Thus, failure to define the mother as an X-linked carrier does not disprove the X-linked origin of the disease, or even the possibility of the mother having another child with X91 CGD. If a mosaic in oxidase activity is found in the mother but no mutation is detectable in CYBB from the patient, the X-linked G6PD gene may carry a mutation. Once the familyspecific mutation is known, it is more reliable to perform carrier detection for any of the CGD subtypes at the DNA level. Nowadays, small gene panels, sequenced on benchtop next generation sequencing (NGS) platforms, are used that can detect mutations in multiple genes in one run [192, 205]. If an entire exon is deleted without the presence of a mutation in the bordering exons, a splice-site mutation may be present in the bordering introns in the genomic DNA. Therefore, exons and exon-intron boundaries of all CGD genes are now routinely investigated, including the G6PD, CYBC1 and RAC2 genes. If no mutations are found in any of these genes, it may be that missplicing of mRNA occurs due to deep intronic mutations. Therefore, mRNA size and sequence should then be analyzed. Also, for NCF1, it is sometimes necessary to analyze the mRNA. Genetic analysis of NCF1 is more difficult, because it is accompanied on each side by one pseudo-NCF1 gene [206]. These pseudo-NCF1 genes are >99% homologous to NCF1 but lack a GT sequence at the start of exon 2, which induces a frameshift and a premature termination of protein synthesis. Although detection of only the GT deletion sequence in NCF1 accounts for over 85% of all p47phoxdeficient CGD patients, the molecular basis is most likely due to partial cross-over events between NCF1 and pseudogenes of p47phox at different recombination sites [190, 191, 207]. Therefore, NCF1-specific PCR is difficult, because the primers have to contain NCF1-specific sequences at the segregating points between NCF1 and its pseudogenes. It is recommended, therefore, to first perform a gene scan or a PCR-RFLP analysis of NCF1 loci [208, 209] to determine whether only GT deletion-containing pseudogenes are present or whether one or two NCF1 genes are present in the patient’s DNA (Fig. 32.3). In the latter case, sequencing of NCF1 cDNA usually reveals another mutation than the GT deletion in NCF1, which must then be confirmed with Sanger sequencing of genomic DNA after NCF1-specific amplification of the relevant part of NCF1 [210]. An additional advantage of NGS is the large number of DNA sequence reads that are obtained. In this way, not only germline mutations but also somatic mutations come to light. For instance, in two mothers with X91 CGD sons, we found

548

M. J. Stasia and D. Roos

Fig. 32.3 Gene scan of NCF1. The patient (top panel) has only pseudogenes, the control (middle panel) has a 2:1 ratio of pseudogenes to NCF1 genes, and the carrier (bottom panel) has a 5:1 ratio. For

technical details see [208]. Figure reproduced from reference [153] with permission from NOVA Science Publishers

3% and 16% of their DNA, respectively, to contain the mutation that was passed on to their sons (unpublished). Results obtained with NGS platforms should be confirmed by Sanger sequencing. Splice-site mutations found in genomic DNA should also be confirmed for their effect on mRNA splicing by analyzing the lack of one or more exons in the cDNA of the patient. Also, the presence of large deletions, usually based on the lack of PCR product formation, should be confirmed by an independent assay, such as multiplex ligase-dependent probe amplification (MLPA). MLPA analyzes with a set of probes annealing at different positions in the DNA which parts of a gene or gene-surrounding sequences are still present [191, 211]. This can also be done by CGD-array analysis [212]. Family studies are also very

helpful to distinguish between homozygous mutations in autosomal genes and compound heterozygous combinations of a deletion with a smaller mutation. Prenatal diagnosis of CGD can be performed by analysis of the NADPH oxidase activity of fetal blood neutrophils [213], but fetal blood sampling cannot be undertaken before 16–18 weeks of gestation. Instead, analysis of DNA from amniotic fluid cells or chorionic villi provides an earlier and more reliable diagnosis for families at risk. In cases where the family-specific mutation is known, this analysis can be performed by PCR amplification and sequencing of the relevant genomic DNA area [214]. Care must be taken to avoid contamination of fetal DNA with maternal DNA; detection of such contamination can be performed by short tandem repeat

32

Chronic Granulomatous Disease

(STR) analysis. The same strategy as for prenatal diagnosis of X91 CGD can be used for prenatal diagnosis of other CGD subtypes [215], although this may be more complicated if the parents carry different mutations.

4.3

Relation Between Mutations and Disease Severity

The relation between mutations in CGD genes and disease severity is unclear. In general, one can safely assume that mutations that destroy oxidase component expression or activity induce severe illness. However, when some expression and activity is left, it is hard to predict what the clinical consequences will be. As mentioned above, mutations in NCF1 in general lead to a milder form of CGD, possibly because some H2O2 formation is still possible. However, exceptions to this rule are frequently seen. The same applies to splice site mutations: some patients with such a mutation do quite well, possibly because some normal pre-mRNA splicing still occurs, but others are severely ill (For example, see patients P4a and P4b who are brothers in [161]). The correlation between the degree of normal splicing and the clinical expression of CGD has not been investigated as far as we know. What has been proven is the value of some remaining NADPH oxidase activity. First, Kuhns et al. [163] and later Köker et al. [162] have shown that residual NADPH oxidase activity in neutrophils from CGD patients is associated with less severe illness and a greater likelihood of long-term survival than observed for patients with little or no NADPH oxidase function. These effects were observed in both X-linked and autosomal recessive forms of CGD. It demonstrates that even low residual levels of ROS formation in neutrophils are protective, and supports the notion that a small proportion of NADPH oxidase-competent neutrophils, such as that which may be achievable with gene therapy, will reduce infection risk in CGD patients. Again, caution is needed: female carriers of X91 CGD mutations vary in their susceptibility for infections. In our own experience, some of these carriers, with up to 30% of oxidase-competent neutrophils show the full spectrum of clinical CGD symptoms, whereas others, with as few as 3% oxidasecompetent neutrophils, are completely healthy. Apparently, additional factors besides the percentage of oxidase-positive neutrophils are involved in the bactericidal and fungicidal competence.

5

Future Developments

Allogeneic hematopoietic stem cell (HSC) transplantation can cure X91 CGD, but many patients lack a suitable donor, and graft-versus-host-disease remains a significant

549

risk of allogeneic transplants [216–218]. Gene transfer correction of autologous HSCs lacks these barriers, but approaches using retroviral vectors have shown the risk of oncogenic insertional mutagenesis due to random vector insertion [219, 220]. For more details, see Chap. 33 by G. Santilli and A. Thrasher. Moreover, methylation of CpG dinucleotide promoter elements have led to silencing of transgene expression. However, a newly developed self-inactivating lentiviral vector G1XCGD has been developed that minimizes these complications [221]. In addition, a novel internal promoter sequence was used that binds transcription factors present primarily in mature myeloid cells, thus driving sufficient gp91phox expression to reconstitute NADPH oxidase activity in blood granulocytes and monocytes [222]. This vector has been used to transduce HSC from nine X91 CGD patients, which were subsequently given back to these patients. Two patients died within three months after treatment, probably from pre-existing comorbidities. One year after treatment, six of the seven surviving patients demonstrated stable vector copy numbers (0.4–1.8 copies per neutrophil) and persistence of 16–46% oxidase-positive neutrophils, without evidence of clonal dysregulation or transgene silencing. The surviving patients had no new CGD-related infections, and six of them were able to discontinue antibiotic prophylaxis [222]. Several additional patients have been treated likewise and are now under surveillance. A similar lentiviral vector has been developed for treatment of p47phox-deficient CGD patients and successfully tested in vivo in a mouse model of this disease and in vitro in human HSC [223]. Malech et al. have developed a “safe-harbor” gene therapy approach for correction of iPS cells from X91 CGD patients using zinc-finger nucleases (ZFNs) to obtain targeted insertion of a codon-optimized CYBB minigene into the AAVS1 locus under the control of a constitutive CAG promoter [224]. This resulted in constitutive expression of gp91phox, which restored ROS production upon in vitro differentiation of iPS cells into granulocytes. However, in clinical gene therapy approaches for X91 CGD, constitutive ectopic expression of the gp91phox protein or a consequent aberrant production of ROS in stem cells prior to myeloid differentiation could potentially impair HSC engraftment or otherwise alter stem cell function in an unexpected manner [225, 226]. Alternative approaches to address this issue might include targeted correction or gene transfer at the CYBB locus, resulting in normal regulation of gp91phox expression by the endogenous CYBB promoter. Sweeney et al. [227] have described several targeted gene transfer strategies for seamless exon replacement of CYBB exon 5 mutations or for transfer of CYBB minigenes to the start sites of exon 1 or exon 2 of the CYBB locus, using site-specific TAL effector nucleases (TALENs) [228, 229] or clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9

550

nuclease [8, 230, 231] targeting these sites. The varying effectiveness of these approaches for restoring gene expression and granulocyte function have highlighted the necessity of intronic elements in the design of targeted gene transfer strategies and donor constructs. Indeed, inclusion of CYBB intron 1 in the corrected sequence and transient inhibition of p53-binding protein to inhibit non-homologous end-joining proved to be essential for efficient restoration of gp91phox expression and ROS production in differentiated human HPSCs [232, 233]. For NCF1 gene therapy, an attractive target is resurrection of one of the two NCF1 pseudogenes, since these are almost always still present in the genome of the p47phox-deficient CGD patients. This has been performed by zinc finger nucleases targeting the exon 2 GT deletion in iPS cells and in CD34-positive hematopoietic stem cells derived from such patients [234]. In addition, recent studies pave the way to a first-in-man clinical trial of a new lentiviral gene therapy for the treatment of p47phox-deficient CGD patients [223]. Acknowledgments Dr. Marie José Stasia would like to express her sincere thanks to all the staff who work at the CDiReC with such dynamism and dedication, as well as to the University Grenoble Alpes (UGA) and the University Hospital Grenoble Alpes (CHUGA) for their constant support.

References 1. Berendes H, Bridges RA, Good RA (1957) A fatal granulomatosus of childhood: the clinical study of a new syndrome. Minn Med 40: 309–312 2. Bridges RA, Berendes H, Good RA (1959) A fatal granulomatous disease of childhood; the clinical, pathological, and laboratory features of a new syndrome. AMA J Dis Child 97:387–408 3. Quie PG, White JG, Holmes B et al (1967) In vitro bactericidal capacity of human polymorphonuclear leukocytes: diminished activity in chronic granulomatous disease of childhood. J Clin Invest 46:668–679 4. Baehner RL, Nathan DG (1967) Leukocyte oxidase: defective activity in chronic granulomatous disease. Science 155:835–836 5. Holmes B, Page AR, Good RA (1967) Studies of the metabolic activity of leukocytes from patients with a genetic abnormality of phagocytic function. J Clin Invest 46:1422–1432 6. Baldridge CW, Gerard RW (1932) The extra respiration of phagocytosis. Am J Physiol Legacy Cont 103:235–236 7. Sbarra AJ, Karnovsky ML (1959) The biochemical basis of phagocytosis. I. Metabolic changes during the ingestion of particles by polymorphonuclear leukocytes. J Biol Chem 234:1355–1362 8. Jinek M, Chylinski K, Fonfara I et al (2012) A programmable dualRNA-guided DNA endonuclease in adaptive bacterial immunity. Science (New York, NY) 337:816–821 9. Paul B, Sbarra AJ (1968) The role of the phagocyte in host-parasite interactions. 13. The direct quantitative estimation of H2O2 in phagocytizing cells. Biochim Biophys Acta 156:168–178 10. Root RK, Metcalf J, Oshino N et al (1975) H2O2 release from human granulocytes during phagocytosis. I. Documentation, quantitation, and some regulating factors. J Clin Invest 55:945–955 11. Klebanoff SJ (1967) Iodination of bacteria: a bactericidal mechanism. J Exp Med 126:1063–1078

M. J. Stasia and D. Roos 12. Mccord JM, Fridovich I (1969) The utility of superoxide dismutase in studying free radical reactions. I. Radicals generated by the interaction of sulfite, dimethyl sulfoxide, and oxygen. J Biol Chem 244:6056–6063 13. Babior BM, Kipnes RS, Curnutte JT (1973) Biological defense mechanisms. The production by leukocytes of superoxide, a potential bactericidal agent. J Clin Invest 52:741–744 14. Curnutte JT, Whitten DM, Babior BM (1974) Defective superoxide production by granulocytes from patients with chronic granulomatous disease. N Engl J Med 290:593–597 15. Klebanoff SJ (1980) Oxygen metabolism and the toxic properties of phagocytes. Ann Intern Med 93:480–489 16. Nauseef WM (2014) Myeloperoxidase in human neutrophil host defence. Cell Microbiol 16:1146–1155 17. Test ST, Lampert MB, Ossanna PJ et al (1984) Generation of nitrogen-chlorine oxidants by human phagocytes. J Clin Invest 74:1341–1349 18. Nauseef WM (1988) Myeloperoxidase deficiency. Hematol Oncol Clin North Am 2:135–158 19. Karnovsky ML, Shafer AW, Cagan RH et al (1966) Membrane function and metabolism in phagocytic cells. Trans N Y Acad Sci 28:778–787 20. Rossi F, Zatti M (1964) Biochemical aspects of phagocytosis in polymorphonuclear leucocytes. NADH and NADPH oxidation by the granules of resting and phagocytizing cells. Experientia 20:21– 23 21. Iverson D, Dechatelet LR, Spitznagel JK et al (1977) Comparison of NADH and NADPH oxidase activities in granules isolated from human polymorphonuclear leukocytes with a fluorometric assay. J Clin Invest 59:282–290 22. Suzuki H, Kakinuma K (1983) Evidence that NADPH is the actual substrate of the oxidase responsible for the “respiratory burst” of phagocytosing polymorphonuclear leukocytes. J Biochem 93:709– 715 23. Suzuki Y, Lehrer RI (1980) NAD(P)H oxidase activity in human neutrophils stimulated by phorbol myristate acetate. J Clin Invest 66:1409–1418 24. Segal AW, Jones OT, Webster D et al (1978) Absence of a newly described cytochrome b from neutrophils of patients with chronic granulomatous disease. Lancet (London, England) 2:446–449 25. Segal AW, Jones OT (1978) Novel cytochrome b system in phagocytic vacuoles of human granulocytes. Nature 276:515–517 26. Cross AR, Jones OT, Garcia R et al (1982) The association of FAD with the cytochrome b-245 of human neutrophils. Biochem J 208: 759–763 27. Gyn I, Islam MF, Quastel JH (1961) Biochemical aspects of phagocytosis. Nature 192:535–541 28. Dinauer MC, Orkin SH, Brown R et al (1987) The glycoprotein encoded by the X-linked chronic granulomatous disease locus is a component of the neutrophil cytochrome b complex. Nature 327: 717–720 29. Segal AW (1987) Absence of both cytochrome b-245 subunits from neutrophils in X-linked chronic granulomatous disease. Nature 326:88–91 30. Parkos CA, Allen RA, Cochrane CG et al (1987) Purified cytochrome b from human granulocyte plasma membrane is comprised of two polypeptides with relative molecular weights of 91,000 and 22,000. J Clin Invest 80:732–742 31. Vermot A, Petit-Härtlein I, Smith SME et al. (2021) NADPH Oxidases (NOX): an overview from discovery, molecular mechanisms to physiology and pathology. Antioxidants (Basel, Switzerland) 10 32. Rotrosen D, Yeung CL, Leto TL et al (1992) Cytochrome b558: the flavin-binding component of the phagocyte NADPH oxidase. Science (New York, NY) 256:1459–1462 33. Segal AW, West I, Wientjes F et al (1992) Cytochrome b-245 is a flavocytochrome containing FAD and the NADPH-binding site of

32

Chronic Granulomatous Disease

the microbicidal oxidase of phagocytes. Biochem J 284(Pt 3): 781–788 34. Sumimoto H, Sakamoto N, Nozaki M et al (1992) Cytochrome b558, a component of the phagocyte NADPH oxidase, is a flavoprotein. Biochem Biophys Res Commun 186:1368–1375 35. Batot G, Paclet MH, Doussière J et al (1998) Biochemical and immunochemical properties of B lymphocyte cytochrome b558. Biochim Biophys Acta 1406:188–202 36. Doussiere J, Brandolin G, Derrien V et al (1993) Critical assessment of the presence of an NADPH binding site on neutrophil cytochrome b558 by photoaffinity and immunochemical labeling. Biochemistry 32:8880–8887 37. Taylor WR, Jones DT, Segal AW (1993) A structural model for the nucleotide binding domains of the flavocytochrome b-245 betachain. Protein Sci 2:1675–1685 38. Segal AW, Heyworth PG, Cockcroft S et al (1985) Stimulated neutrophils from patients with autosomal recessive chronic granulomatous disease fail to phosphorylate a Mr-44,000 protein. Nature 316:547–549 39. Hamers MN, De Boer M, Meerhof LJ et al (1984) Complementation in monocyte hybrids revealing genetic heterogeneity in chronic granulomatous disease. Nature 307:553–555 40. Weening RS, Corbeel L, De Boer M et al (1985) Cytochrome b deficiency in an autosomal form of chronic granulomatous disease. A third form of chronic granulomatous disease recognized by monocyte hybridization. J Clin Invest 75:915–920 41. Bromberg Y, Pick E (1984) Unsaturated fatty acids stimulate NADPH-dependent superoxide production by cell-free system derived from macrophages. Cell Immunol 88:213–221 42. Bromberg Y, Pick E (1985) Activation of NADPH-dependent superoxide production in a cell-free system by sodium dodecyl sulfate. J Biol Chem 260:13539–13545 43. Curnutte JT (1985) Activation of human neutrophil nicotinamide adenine dinucleotide phosphate, reduced (triphosphopyridine nucleotide, reduced) oxidase by arachidonic acid in a cell-free system. J Clin Invest 75:1740–1743 44. Curnutte JT, Kuver R, Scott PJ (1987) Activation of neutrophil NADPH oxidase in a cell-free system. Partial purification of components and characterization of the activation process. J Biol Chem 262:5563–5569 45. Heyneman RA, Vercauteren RE (1984) Activation of a NADPH oxidase from horse polymorphonuclear leukocytes in a cell-free system. J Leukoc Biol 36:751–759 46. Ligeti E, Doussiere J, Vignais PV (1988) Activation of the O2(.-)generating oxidase in plasma membrane from bovine polymorphonuclear neutrophils by arachidonic acid, a cytosolic factor of protein nature, and nonhydrolyzable analogues of GTP. Biochemistry 27:193–200 47. Mcphail LC, Shirley PS, Clayton CC et al (1985) Activation of the respiratory burst enzyme from human neutrophils in a cell-free system. Evidence for a soluble cofactor. J Clin Invest 75:1735– 1739 48. Curnutte JT, Scott PJ, Mayo LA (1989) Cytosolic components of the respiratory burst oxidase: resolution of four components, two of which are missing in complementing types of chronic granulomatous disease. Proc Natl Acad Sci USA 86:825–829 49. Nunoi H, Rotrosen D, Gallin JI et al (1988) Two forms of autosomal chronic granulomatous disease lack distinct neutrophil cytosol factors. Science (New York, NY) 242:1298–1301 50. Volpp BD, Nauseef WM, Clark RA (1988) Two cytosolic neutrophil oxidase components absent in autosomal chronic granulomatous disease. Science (New York, NY) 242:1295–1297 51. Wientjes FB, Hsuan JJ, Totty NF et al (1993) p40phox, a third cytosolic component of the activation complex of the NADPH

551 oxidase to contain src homology 3 domains. Biochem J 296 (Pt 3):557–561 52. Clark RA, Malech HL, Gallin JI et al (1989) Genetic variants of chronic granulomatous disease: prevalence of deficiencies of two cytosolic components of the NADPH oxidase system. N Engl J Med 321:647–652 53. Matute JD, Arias AA, Dinauer MC et al (2005) p40phox: the last NADPH oxidase subunit. Blood Cells Mol Dis 35:291–302 54. Matute JD, Arias AA, Wright NA et al (2009) A new genetic subgroup of chronic granulomatous disease with autosomal recessive mutations in p40 phox and selective defects in neutrophil NADPH oxidase activity. Blood 114:3309–3315 55. Royer-Pokora B, Kunkel LM, Monaco AP et al (1986) Cloning the gene for an inherited human disorder – chronic granulomatous disease – on the basis of its chromosomal location. Nature 322: 32–38 56. Volpp BD, Nauseef WM, Donelson JE et al (1989) Cloning of the cDNA and functional expression of the 47-kilodalton cytosolic component of human neutrophil respiratory burst oxidase. Proc Natl Acad Sci USA 86:7195–7199 57. Ziegler CS, Bouchab L, Tramier M et al (2019) Quantitative livecell imaging and 3D modeling reveal critical functional features in the cytosolic complex of phagocyte NADPH oxidase. J Biol Chem 294:3824–3836 58. Ago T, Kuribayashi F, Hiroaki H et al (2003) Phosphorylation of p47phox directs phox homology domain from SH3 domain toward phosphoinositides, leading to phagocyte NADPH oxidase activation. Proc Natl Acad Sci USA 100:4474–4479 59. Boussetta T, Gougerot-Pocidalo MA, Hayem G et al (2010) The prolyl isomerase Pin1 acts as a novel molecular switch for TNFalpha-induced priming of the NADPH oxidase in human neutrophils. Blood 116:5795–5802 60. Marcoux J, Man P, Castellan M et al (2009) Conformational changes in p47(phox) upon activation highlighted by mass spectrometry coupled to hydrogen/deuterium exchange and limited proteolysis. FEBS Lett 583:835–840 61. Anderson-Cohen M, Holland SM, Kuhns DB et al (2003) Severe phenotype of chronic granulomatous disease presenting in a female with a de novo mutation in gp91-phox and a non familial, extremely skewed X chromosome inactivation. Clin Immunol 109:308–317 62. Belambri SA, Rolas L, Raad H et al (2018) NADPH oxidase activation in neutrophils: role of the phosphorylation of its subunits. Eur J Clin Investig 48(Suppl 2):e12951 63. Groemping Y, Lapouge K, Smerdon SJ et al (2003) Molecular basis of phosphorylation-induced activation of the NADPH oxidase. Cell 113:343–355 64. Fontayne A, Dang PM, Gougerot-Pocidalo MA et al (2002) Phosphorylation of p47phox sites by PKC alpha, beta II, delta, and zeta: effect on binding to p22phox and on NADPH oxidase activation. Biochemistry 41:7743–7750 65. Nauseef WM (2004) Assembly of the phagocyte NADPH oxidase. Histochem Cell Biol 122:277–291 66. Hall A (1994) Small GTP-binding proteins and the regulation of the actin cytoskeleton. Annu Rev Cell Biol 10:31–54 67. Just I, Mohr C, Schallehn G et al (1992) Purification and characterization of an ADP-ribosyltransferase produced by Clostridium limosum. J Biol Chem 267:10274–10280 68. Stasia MJ, Jouan A, Bourmeyster N et al (1991) ADP-ribosylation of a small size GTP-binding protein in bovine neutrophils by the C3 exoenzyme of Clostridium botulinum and effect on the cell motility. Biochem Biophys Res Commun 180: 615–622

552 69. Abo A, Pick E (1991) Purification and characterization of a third cytosolic component of the superoxide-generating NADPH oxidase of macrophages. J Biol Chem 266:23577–23585 70. Abo A, Pick E, Hall A et al (1991) Activation of the NADPH oxidase involves the small GTP-binding protein p21rac1. Nature 353:668–670 71. Didsbury J, Weber RF, Bokoch GM et al (1989) rac, a novel ras-related family of proteins that are botulinum toxin substrates. J Biol Chem 264:16378–16382 72. Heyworth PG, Bohl BP, Bokoch GM et al (1994) Rac translocates independently of the neutrophil NADPH oxidase components p47phox and p67phox. Evidence for its interaction with flavocytochrome b558. J Biol Chem 269:30749–30752 73. Quinn MT, Evans T, Loetterle LR et al (1993) Translocation of Rac correlates with NADPH oxidase activation. Evidence for equimolar translocation of oxidase components. J Biol Chem 268:20983– 20987 74. Abo A, Webb MR, Grogan A et al (1994) Activation of NADPH oxidase involves the dissociation of p21rac from its inhibitory GDP/GTP exchange protein (rhoGDI) followed by its translocation to the plasma membrane. Biochem J 298(Pt 3):585–591 75. Grizot S, Fauré J, Fieschi F et al (2001) Crystal structure of the Rac1-RhoGDI complex involved in nadph oxidase activation. Biochemistry 40:10007–10013 76. Pick E (2014) Role of the Rho GTPase Rac in the activation of the phagocyte NADPH oxidase: outsourcing a key task. Small GTPases 5:e27952 77. Diebold BA, Bokoch GM (2001) Molecular basis for Rac2 regulation of phagocyte NADPH oxidase. Nat Immunol 2:211–215 78. Diekmann D, Abo A, Johnston C et al (1994) Interaction of Rac with p67phox and regulation of phagocytic NADPH oxidase activity. Science (New York, NY) 265:531–533 79. Kao YY, Gianni D, Bohl B et al (2008) Identification of a conserved Rac-binding site on NADPH oxidases supports a direct GTPase regulatory mechanism. J Biol Chem 283:12736–12746 80. Lapouge K, Smith SJ, Walker PA et al (2000) Structure of the TPR domain of p67phox in complex with Rac.GTP. Mol Cell 6:899– 907 81. Leusen JH, De Klein A, Hilarius PM et al (1996) Disturbed interaction of p21-rac with mutated p67-phox causes chronic granulomatous disease. J Exp Med 184:1243–1249 82. Ando S, Kaibuchi K, Sasaki T et al (1992) Post-translational processing of rac p21s is important both for their interaction with the GDP/GTP exchange proteins and for their activation of NADPH oxidase. J Biol Chem 267:25709–25713 83. Gorzalczany Y, Sigal N, Itan M et al (2000) Targeting of Rac1 to the phagocyte membrane is sufficient for the induction of NADPH oxidase assembly. J Biol Chem 275:40073–40081 84. Mizrahi A, Berdichevsky Y, Casey PJ et al (2010) A prenylated p47phox-p67phox-Rac1 chimera is a Quintessential NADPH oxidase activator: membrane association and functional capacity. J Biol Chem 285:25485–25499 85. Han CH, Freeman JL, Lee T et al (1998) Regulation of the neutrophil respiratory burst oxidase. Identification of an activation domain in p67(phox). J Biol Chem 273:16663–16668 86. Ambruso DR, Knall C, Abell AN et al (2000) Human neutrophil immunodeficiency syndrome is associated with an inhibitory Rac2 mutation. Proc Natl Acad Sci USA 97:4654–4659 87. Pick E (2020) Cell-free NADPH oxidase activation assays: a triumph of reductionism. Methods Mol Biol (Clifton, NJ) 2087:325– 411 88. Winkelstein JA, Marino MC, Johnston RB Jr et al (2000) Chronic granulomatous disease. Report on a national registry of 368 patients. Medicine 79:155–169

M. J. Stasia and D. Roos 89. Van Den Berg JM, Van Koppen E, Ahlin A et al (2009) Chronic granulomatous disease: the European experience. PLoS One 4: e5234 90. Segal BH, Decarlo ES, Kwon-Chung KJ et al (1998) Aspergillus nidulans infection in chronic granulomatous disease. Medicine 77: 345–354 91. Segal BH, Holland SM (2017) Bacterial and fungal infections in chronic granulomatous disease. In: Seger R, Roos D, Segal BH, Kuijpers T (eds) Genetics, biology and clinical management. NOVA Sci. Publ., New York, pp 91–123 92. Lee PP, Chan KW, Jiang L et al (2008) Susceptibility to mycobacterial infections in children with X-linked chronic granulomatous disease: a review of 17 patients living in a region endemic for tuberculosis. Pediatr Infect Dis J 27:224–230 93. Blancas-Galicia L, Santos-Chávez E, Deswarte C et al (2020) Genetic, immunological, and clinical features of the first Mexican Cohort of patients with chronic granulomatous disease. J Clin Immunol 40:475–493 94. Zhou Q, Hui X, Ying W et al (2018) A cohort of 169 chronic granulomatous disease patients exposed to BCG vaccination: a retrospective study from a Single Center in Shanghai, China (2004–2017). J Clin Immunol 38:260–272 95. Prince BT, Thielen BK, Williams KW et al (2020) Geographic variability and pathogen-specific considerations in the diagnosis and management of chronic granulomatous disease. Pediatric Health Med Ther 11:257–268 96. Guide SV, Stock F, Gill VJ et al (2003) Reinfection, rather than persistent infection, in patients with chronic granulomatous disease. J Infect Dis 187:845–853 97. Cathébras P, Sauron C, Morel F et al (2001) An unusual case of sarcoidosis. Lancet (London, England) 358:294 98. De Ravin SS, Naumann N, Robinson MR et al (2006) Sarcoidosis in chronic granulomatous disease. Pediatrics 117:e590–e595 99. Magnani A, Brosselin P, Beauté J et al (2014) Inflammatory manifestations in a single-center cohort of patients with chronic granulomatous disease. J Allergy Clin Immunol 134:655–662.e658 100. Marciano BE, Rosenzweig SD, Kleiner DE et al (2004) Gastrointestinal involvement in chronic granulomatous disease. Pediatrics 114:462–468 101. Huang J, Canadien V, Lam GY et al (2009) Activation of antibacterial autophagy by NADPH oxidases. Proc Natl Acad Sci USA 106:6226–6231 102. Martinez J, Almendinger J, Oberst A et al (2011) Microtubuleassociated protein 1 light chain 3 alpha (LC3)-associated phagocytosis is required for the efficient clearance of dead cells. Proc Natl Acad Sci USA 108:17396–17401 103. Fernandez-Boyanapalli R, Mcphillips KA, Frasch SC et al (2010) Impaired phagocytosis of apoptotic cells by macrophages in chronic granulomatous disease is reversed by IFN-γ in a nitric oxide-dependent manner. J Immunol 185:4030–4041 104. Fernandez-Boyanapalli RF, Frasch SC, Mcphillips K et al (2009) Impaired apoptotic cell clearance in CGD due to altered macrophage programming is reversed by phosphatidylserine-dependent production of IL-4. Blood 113:2047–2055 105. Greenlee-Wacker MC, Rigby KM, Kobayashi SD et al (2014) Phagocytosis of Staphylococcus aureus by human neutrophils prevents macrophage efferocytosis and induces programmed necrosis. J Immunol 192:4709–4717 106. Martinez J, Cunha LD, Park S et al (2016) Noncanonical autophagy inhibits the autoinflammatory, lupus-like response to dying cells. Nature 533:115–119 107. Meissner F, Seger RA, Moshous D et al (2010) Inflammasome activation in NADPH oxidase defective mononuclear phagocytes from patients with chronic granulomatous disease. Blood 116: 1570–1573

32

Chronic Granulomatous Disease

108. Thomas DC (2018) How the phagocyte NADPH oxidase regulates innate immunity. Free Radic Biol Med 125:44–52 109. Aarts CEM, Hiemstra IH, Béguin EP et al (2019) Activated neutrophils exert myeloid-derived suppressor cell activity damaging T cells beyond repair. Blood Adv 3:3562–3574 110. Chin TW, Stiehm ER, Falloon J et al (1987) Corticosteroids in treatment of obstructive lesions of chronic granulomatous disease. J Pediatr 111:349–352 111. Rösen-Wolff A, Soldan W, Heyne K et al (2001) Increased susceptibility of a carrier of X-linked chronic granulomatous disease (CGD) to Aspergillus fumigatus infection associated with age-related skewing of lyonization. Ann Hematol 80:113–115 112. Battersby AC, Braggins H, Pearce MS et al (2017) Inflammatory and autoimmune manifestations in X-linked carriers of chronic granulomatous disease in the United Kingdom. J Allergy Clin Immunol 140:628–630.e626 113. Marciano BE, Zerbe CS, Falcone EL et al (2018) X-linked carriers of chronic granulomatous disease: illness, lyonization, and stability. J Allergy Clin Immunol 141:365–371 114. To EE, Vlahos R, Luong R et al (2017) Endosomal NOX2 oxidase exacerbates virus pathogenicity and is a target for antiviral therapy. Nat Commun 8:69 115. Souyris M, Cenac C, Azar P et al (2018) TLR7 escapes X chromosome inactivation in immune cells. Sci Immunol 3 116. Chiriaco M, Salfa I, Ursu GM et al. (2021) Immunological aspects of X-linked chronic granulomatous disease female carriers. Antioxidants (Basel, Switzerland) 10 117. Margolis DM, Melnick DA, Alling DW et al (1990) Trimethoprimsulfamethoxazole prophylaxis in the management of chronic granulomatous disease. J Infect Dis 162:723–726 118. Seger RA, Baumgartner S, Tiefenauer LX et al (1981) Chronic granulomatous disease: effect of sulfamethoxazole/trimethoprim on neutrophil microbicidal function. Helv Paediatr Acta 36:579– 588 119. Lafont E, Marciano BE, Mahlaoui N et al (2020) Nocardiosis associated with primary immunodeficiencies (nocar-DIP): an international retrospective study and literature review. J Clin Immunol 40:1144–1155 120. Gallin JI, Alling DW, Malech HL et al (2003) Itraconazole to prevent fungal infections in chronic granulomatous disease. N Engl J Med 348:2416–2422 121. Mouy R, Veber F, Blanche S et al (1994) Long-term itraconazole prophylaxis against Aspergillus infections in thirty-two patients with chronic granulomatous disease. J Pediatr 125:998–1003 122. Group ICCS (1991) A controlled trial of interferon gamma to prevent infection in chronic granulomatous disease. N Engl J Med 324:509–516 123. Martire B, Rondelli R, Soresina A et al (2008) Clinical features, long-term follow-up and outcome of a large cohort of patients with Chronic Granulomatous Disease: an Italian multicenter study. Clin Immunol (Orlando, FL) 126:155–164 124. Güngör T, Teira P, Slatter M et al (2014) Reduced-intensity conditioning and HLA-matched haemopoietic stem-cell transplantation in patients with chronic granulomatous disease: a prospective multicentre study. Lancet (London, England) 383:436–448 125. Parta M, Kelly C, Kwatemaa N et al (2017) Allogeneic reducedintensity hematopoietic stem cell transplantation for chronic granulomatous disease: a single-center prospective trial. J Clin Immunol 37:548–558 126. Marciano BE, Allen ES, Conry-Cantilena C et al (2017) Granulocyte transfusions in patients with chronic granulomatous disease and refractory infections: the NIH experience. J Allergy Clin Immunol 140:622–625 127. Marfin AA, Price TH (2015) Granulocyte transfusion therapy. J Intensive Care Med 30:79–88

553 128. Drewniak A, Van Raam BJ, Geissler J et al (2009) Changes in gene expression of granulocytes during in vivo granulocyte colonystimulating factor/dexamethasone mobilization for transfusion purposes. Blood 113:5979–5998 129. Gazendam RP, Van De Geer A, Van Hamme JL et al (2016) Impaired killing of Candida albicans by granulocytes mobilized for transfusion purposes: a role for granule components. Haematologica 101:587–596 130. Van De Geer A, Gazendam RP, Tool AT et al (2017) Characterization of buffy coat-derived granulocytes for clinical use: a comparison with granulocyte colony-stimulating factor/ dexamethasone-pretreated donor-derived products. Vox Sang 112:173–182 131. De Ravin SS, Brault J, Meis RJ et al (2020) NADPH oxidase correction by mRNA transfection of apheresis granulocytes in chronic granulomatous disease. Blood Adv 4:5976–5987 132. Chetty M, Thrasher AJ, Abo A et al (1995) Low NADPH oxidase activity in Epstein-Barr-virus-immortalized B-lymphocytes is due to a post-transcriptional block in expression of cytochrome b558. Biochem J 306(Pt 1):141–145 133. Okochi Y, Okamura Y (2021) Regulation of neutrophil functions by Hv1/VSOP voltage-gated proton channels. Int J Mol Sci 22 134. El Chemaly A, Okochi Y, Sasaki M et al (2010) VSOP/Hv1 proton channels sustain calcium entry, neutrophil migration, and superoxide production by limiting cell depolarization and acidification. J Exp Med 207:129–139 135. Hackam DJ, Rotstein OD, Zhang WJ et al (1997) Regulation of phagosomal acidification. Differential targeting of Na+/H+ exchangers, Na+/K+-ATPases, and vacuolar-type H+-atpases. J Biol Chem 272:29810–29820 136. Segal AW, Geisow M, Garcia R et al (1981) The respiratory burst of phagocytic cells is associated with a rise in vacuolar pH. Nature 290:406–409 137. Mantegazza AR, Savina A, Vermeulen M et al (2008) NADPH oxidase controls phagosomal pH and antigen cross-presentation in human dendritic cells. Blood 112:4712–4722 138. Winterbourn CC, Hampton MB, Livesey JH et al (2006) Modeling the reactions of superoxide and myeloperoxidase in the neutrophil phagosome: implications for microbial killing. J Biol Chem 281: 39860–39869 139. Buvelot H, Posfay-Barbe KM, Linder P et al (2017) Staphylococcus aureus, phagocyte NADPH oxidase and chronic granulomatous disease. FEMS Microbiol Rev 41:139–157 140. Cabiscol E, Tamarit J, Ros J (2000) Oxidative stress in bacteria and protein damage by reactive oxygen species. Int Microbiol 3:3–8 141. De Jong NWM, Van Kessel KPM, Van Strijp, JAG (2019) Immune evasion by Staphylococcus aureus. Microbiol Spectr 7 142. Genestet C, Le Gouellec A, Chaker H et al (2014) Scavenging of reactive oxygen species by tryptophan metabolites helps Pseudomonas aeruginosa escape neutrophil killing. Free Radic Biol Med 73:400–410 143. Chang YC, Segal BH, Holland SM et al (1998) Virulence of catalase-deficient aspergillus nidulans in p47(phox)-/- mice. Implications for fungal pathogenicity and host defense in chronic granulomatous disease. J Clin Invest 101:1843–1850 144. Messina CG, Reeves EP, Roes J et al (2002) Catalase negative Staphylococcus aureus retain virulence in mouse model of chronic granulomatous disease. FEBS Lett 518:107–110 145. Rosen H, Michel BR (1997) Redundant contribution of myeloperoxidase-dependent systems to neutrophil-mediated killing of Escherichia coli. Infect Immun 65:4173–4178 146. Solberg CO (1972) Protection of phagocytized bacteria against antibiotics. A new method for the evaluation of neutrophil granulocyte functions. Acta Med Scand 191:383–387 147. Gazendam RP, Van Hamme JL, Tool AT et al (2016) Human neutrophils use different mechanisms to kill Aspergillus fumigatus

554 Conidia and Hyphae: evidence from phagocyte defects. J Immunol 196:1272–1283 148. Gazendam RP, Van Hamme JL, Tool AT et al (2014) Two independent killing mechanisms of Candida albicans by human neutrophils: evidence from innate immunity defects. Blood 124: 590–597 149. Marciano BE, Spalding C, Fitzgerald A et al (2015) Common severe infections in chronic granulomatous disease. Clin Infect Dis 60:1176–1183 150. Tian W, Li XJ, Stull ND et al (2008) Fc gamma R-stimulated activation of the NADPH oxidase: phosphoinositide-binding protein p40phox regulates NADPH oxidase activity after enzyme assembly on the phagosome. Blood 112:3867–3877 151. Van De Geer A, Nieto-Patlán A, Kuhns DB et al (2018) Inherited p40phox deficiency differs from classic chronic granulomatous disease. J Clin Invest 128:3957–3975 152. Nguyen GT, Green ER, Mecsas J (2017) Neutrophils to the ROScue: mechanisms of NADPH oxidase activation and bacterial resistance. Front Cell Infect Microbiol 7:373 153. Roos D, Tool ATJ, Van Leeuwen K et al (2017) Biochemical and genetic diagnosis of chronic granulomatous disease. In: Seger R, Roos D, Segal BH, Kuijpers T (eds) Genetics, biology and clinical management. NOVA Sci. Publ., New York, pp 231–300 154. O’neill S, Brault J, Stasia MJ et al (2015) Genetic disorders coupled to ROS deficiency. Redox Biol 6:135–156 155. Stasia MJ, Li XJ (2008) Genetics and immunopathology of chronic granulomatous disease. Semin Immunopathol 30:209–235 156. Emmendörffer A, Nakamura M, Rothe G et al (1994) Evaluation of flow cytometric methods for diagnosis of chronic granulomatous disease variants under routine laboratory conditions. Cytometry 18: 147–155 157. Elloumi HZ, Holland SM (2014) Diagnostic assays for chronic granulomatous disease and other neutrophil disorders. Methods Mol Biol (Clifton, NJ) 1124:517–535 158. Mauch L, Lun A, O’gorman MR et al (2007) Chronic granulomatous disease (CGD) and complete myeloperoxidase deficiency both yield strongly reduced dihydrorhodamine 123 test signals but can be easily discerned in routine testing for CGD. Clin Chem 53:890– 896 159. Beaumel S, Picciocchi A, Debeurme F et al (2017) Downregulation of NOX2 activity in phagocytes mediated by ATM-kinase dependent phosphorylation. Free Radic Biol Med 113:1–15 160. Jirapongsananuruk O, Malech HL, Kuhns DB et al (2003) Diagnostic paradigm for evaluation of male patients with chronic granulomatous disease, based on the dihydrorhodamine 123 assay. J Allergy Clin Immunol 111:374–379 161. Mollin M, Beaumel S, Vigne B et al (2021) Clinical, functional and genetic characterization of 16 patients suffering from chronic granulomatous disease variants – identification of 11 novel mutations in CYBB. Clin Exp Immunol 203:247–266 162. Köker MY, Camc{oğlu Y, Van Leeuwen K et al (2013) Clinical, functional, and genetic characterization of chronic granulomatous disease in 89 Turkish patients. J Allergy Clin Immunol 132:1156– 1163.e1155 163. Kuhns DB, Alvord WG, Heller T et al (2010) Residual NADPH oxidase and survival in chronic granulomatous disease. N Engl J Med 363:2600–2610 164. Bylund J, Björnsdottir H, Sundqvist M et al (2014) Measurement of respiratory burst products, released or retained, during activation of professional phagocytes. Methods Mol Biol (Clifton, NJ) 1124: 321–338 165. Dusi S, Donini M, Rossi F (1996) Mechanisms of NADPH oxidase activation: translocation of p40phox, Rac1 and Rac2 from the cytosol to the membranes in human neutrophils lacking p47phox or p67phox. Biochem J 314(Pt 2):409–412

M. J. Stasia and D. Roos 166. Cross AR, Curnutte JT (1995) The cytosolic activating factors p47phox and p67phox have distinct roles in the regulation of electron flow in NADPH oxidase. J Biol Chem 270:6543–6548 167. Aygun D, Koker MY, Nepesov S et al (2020) Genetic characteristics, infectious, and noninfectious manifestations of 32 patients with chronic granulomatous disease. Int Arch Allergy Immunol 181:540–550 168. Wolach B, Gavrieli R, De Boer M et al (2017) Chronic granulomatous disease: clinical, functional, molecular, and genetic studies. The Israeli experience with 84 patients. Am J Hematol 92:28–36 169. Leto TL, Adams AG, De Mendez I (1994) Assembly of the phagocyte NADPH oxidase: binding of Src homology 3 domains to proline-rich targets. Proc Natl Acad Sci USA 91:10650–10654 170. Thrasher A, Chetty M, Casimir C et al (1992) Restoration of superoxide generation to a chronic granulomatous disease-derived B-cell line by retrovirus mediated gene transfer. Blood 80:1125– 1129 171. Tucker KA, Lilly MB, Heck L Jr et al (1987) Characterization of a new human diploid myeloid leukemia cell line (PLB-985) with granulocytic and monocytic differentiating capacity. Blood 70: 372–378 172. Zhen L, King AA, Xiao Y et al (1993) Gene targeting of X chromosome-linked chronic granulomatous disease locus in a human myeloid leukemia cell line and rescue by expression of recombinant gp91phox. Proc Natl Acad Sci USA 90:9832–9836 173. Ding C, Kume A, Björgvinsdóttir H et al (1996) High-level reconstitution of respiratory burst activity in a human X-linked chronic granulomatous disease (X-CGD) cell line and correction of murine X-CGD bone marrow cells by retroviral-mediated gene transfer of human gp91phox. Blood 88:1834–1840 174. Takahashi K, Yamanaka S (2006) Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126:663–676 175. Brault J, Vaganay G, Le Roy A et al (2017) Therapeutic effects of proteoliposomes on X-linked chronic granulomatous disease: proof of concept using macrophages differentiated from patient-specific induced pluripotent stem cells. Int J Nanomedicine 12:2161–2177 176. Dinauer MC, Pierce EA, Bruns GA et al (1990) Human neutrophil cytochrome b light chain (p22-phox). Gene structure, chromosomal location, and mutations in cytochrome-negative autosomal recessive chronic granulomatous disease. J Clin Invest 86:1729– 1737 177. Leto TL, Lomax KJ, Volpp BD et al (1990) Cloning of a 67-kD neutrophil oxidase factor with similarity to a noncatalytic region of p60c-src, vol 248, Science (New York, NY), pp 727–730 178. Parkos CA, Dinauer MC, Walker LE et al (1988) Primary structure and unique expression of the 22-kilodalton light chain of human neutrophil cytochrome b. Proc Natl Acad Sci USA 85:3319–3323 179. Burritt JB, Deleo FR, Mcdonald CL et al (2001) Phage display epitope mapping of human neutrophil flavocytochrome b558. Identification of two juxtaposed extracellular domains. J Biol Chem 276:2053–2061 180. Burritt JB, Foubert TR, Baniulis D et al (2003) Functional epitope on human neutrophil flavocytochrome b558. J Immunol 170:6082– 6089 181. Campion Y, Jesaitis AJ, Nguyen MV et al (2009) New p22-phox monoclonal antibodies: identification of a conformational probe for cytochrome b 558. J Innate Immun 1:556–569 182. Jesaitis AJ, Riesselman M, Taylor RM et al (2019) Enhanced immunoaffinity purification of human neutrophil flavocytochrome B for structure determination by electron microscopy. Methods Mol Biol (Clifton, NJ) 1982:39–59 183. Verhoeven AJ, Bolscher BG, Meerhof LJ et al (1989) Characterization of two monoclonal antibodies against cytochrome b558 of human neutrophils. Blood 73:1686–1694

32

Chronic Granulomatous Disease

184. Yamauchi A, Yu L, Pötgens AJ et al (2001) Location of the epitope for 7D5, a monoclonal antibody raised against human flavocytochrome b558, to the extracellular peptide portion of primate gp91phox. Microbiol Immunol 45:249–257 185. Bakri FG, Mollin M, Beaumel S et al (2021) Second report of chronic granulomatous disease in Jordan: clinical and genetic description of 31 patients from 21 different families, including families from Lybia and Iraq. Front Immunol 12:639226 186. Köker MY, Sanal O, Van Leeuwen K et al (2009) Four different NCF2 mutations in six families from Turkey and an overview of NCF2 gene mutations. Eur J Clin Investig 39:942–951 187. Yu L, Zhen L, Dinauer MC (1997) Biosynthesis of the phagocyte NADPH oxidase cytochrome b558. Role of heme incorporation and heterodimer formation in maturation and stability of gp91phox and p22phox subunits. J Biol Chem 272:27288–27294 188. Roos D, Van Leeuwen K, Hsu AP et al (2021) Hematologically important mutations: X-linked chronic granulomatous disease (fourth update). Blood Cells Mol Dis 90:102587 189. Roos D, Van Leeuwen K, Hsu AP et al (2021) Hematologically important mutations: the autosomal forms of chronic granulomatous disease (third update). Blood Cells Mol Dis 92:102596 190. Görlach A, Lee PL, Roesler J et al (1997) A p47-phox pseudogene carries the most common mutation causing p47-phox-deficient chronic granulomatous disease. J Clin Invest 100:1907–1918 191. Hayrapetyan A, Dencher PC, Van Leeuwen K et al (2013) Different unequal cross-over events between NCF1 and its pseudogenes in autosomal p47(phox)-deficient chronic granulomatous disease. Biochim Biophys Acta 1832:1662–1672 192. Ripen AM, Chiow MY, Rama Rao PR et al (2021) Revealing chronic granulomatous disease in a patient with Williams-Beuren syndrome using whole exome sequencing. Front Immunol 12: 778133 193. Stasia MJ, Mollin M, Martel C et al (2013) Functional and genetic characterization of two extremely rare cases of Williams-Beuren syndrome associated with chronic granulomatous disease. Eur J Hum Genet 21:1079–1084 194. Defendi F, Decleva E, Martel C et al (2009) A novel point mutation in the CYBB gene promoter leading to a rare X minus chronic granulomatous disease variant–impact on the microbicidal activity of neutrophils. Biochim Biophys Acta 1792:201–210 195. Roos D, De Boer M, Borregard N et al (1992) Chronic granulomatous disease with partial deficiency of cytochrome b558 and incomplete respiratory burst: variants of the X-linked, cytochrome b558negative form of the disease. J Leukoc Biol 51:164–171 196. Weening RS, De Boer M, Kuijpers TW et al (2000) Point mutations in the promoter region of the CYBB gene leading to mild chronic granulomatous disease. Clin Exp Immunol 122:410– 417 197. Roos D, De Boer M (2021) Mutations in cis that affect mRNA synthesis, processing and translation. Biochim Biophys Acta Mol Basis Dis 1867:166166 198. Roos D, Van Zwieten R, Wijnen JT et al (1999) Molecular basis and enzymatic properties of glucose 6-phosphate dehydrogenase volendam, leading to chronic nonspherocytic anemia, granulocyte dysfunction, and increased susceptibility to infections. Blood 94: 2955–2962 199. Van Bruggen R, Bautista JM, Petropoulou T et al (2002) Deletion of leucine 61 in glucose-6-phosphate dehydrogenase leads to chronic nonspherocytic anemia, granulocyte dysfunction, and increased susceptibility to infections. Blood 100:1026–1030 200. Thomas DC, Clare S, Sowerby JM et al (2017) Eros is a novel transmembrane protein that controls the phagocyte respiratory burst and is essential for innate immunity. J Exp Med 214:1111– 1128

555 201. Arnadottir GA, Norddahl GL, Gudmundsdottir S et al (2018) A homozygous loss-of-function mutation leading to CYBC1 deficiency causes chronic granulomatous disease. Nat Commun 9: 4447 202. Thomas DC, Charbonnier LM, Schejtman A et al (2019) EROS/ CYBC1 mutations: decreased NADPH oxidase function and chronic granulomatous disease. J Allergy Clin Immunol 143: 782–785.e781 203. Lagresle-Peyrou C, Olichon A, Sadek H et al (2021) A gain-offunction RAC2 mutation is associated with bone-marrow hypoplasia and an autosomal dominant form of severe combined immunodeficiency. Haematologica 106:404–411 204. Lougaris V, Baronio M, Gazzurelli L et al (2020) RAC2 and primary human immune deficiencies. J Leukoc Biol 108:687–696 205. Kanagal-Shamanna R, Portier BP, Singh RR et al (2014) Nextgeneration sequencing-based multi-gene mutation profiling of solid tumors using fine needle aspiration samples: promises and challenges for routine clinical diagnostics. Modern Pathol 27: 314–327 206. Chanock SJ, Roesler J, Zhan S et al (2000) Genomic structure of the human p47-phox (NCF1) gene. Blood Cells Mol Dis 26:37–46 207. Vázquez N, Lehrnbecher T, Chen R et al (2001) Mutational analysis of patients with p47-phox-deficient chronic granulomatous disease: the significance of recombination events between the p47-phox gene (NCF1) and its highly homologous pseudogenes. Exp Hematol 29:234–243 208. Dekker J, De Boer M, Roos D (2001) Gene-scan method for the recognition of carriers and patients with p47(phox)-deficient autosomal recessive chronic granulomatous disease. Exp Hematol 29: 1319–1325 209. Wrona D, Siler U, Reichenbach J (2019) Novel diagnostic tool for p47 (phox)-deficient chronic granulomatous disease patient and carrier detection. Mol Ther Methods Clin Dev 13:274–278 210. Roos D, De Boer M, Köker MY et al (2006) Chronic granulomatous disease caused by mutations other than the common GT deletion in NCF1, the gene encoding the p47phox component of the phagocyte NADPH oxidase. Hum Mutat 27:1218–1229 211. Stasia MJ, Van Leeuwen K, De Boer M et al (2012) Rare duplication or deletion of exons 6, 7 and 8 in CYBB leading to X-linked chronic granulomatous disease in two patients from different families. J Clin Immunol 32:653–662 212. Martel C, Mollin M, Beaumel S et al (2012) Clinical, functional and genetic analysis of twenty-four patients with chronic granulomatous disease – identification of eight novel mutations in CYBB and NCF2 genes. J Clin Immunol 32:942–958 213. Newburger PE, Cohen HJ, Rothchild SB et al (1979) Prenatal diagnosis of chronic granulomatous disease. N Engl J Med 300: 178–181 214. De Boer M, Bolscher BG, Sijmons RH et al (1992) Prenatal diagnosis in a family with X-linked chronic granulomatous disease with the use of the polymerase chain reaction. Prenat Diagn 12: 773–777 215. De Boer M, Singh V, Dekker J et al (2002) Prenatal diagnosis in two families with autosomal, p47(phox)-deficient chronic granulomatous disease due to a novel point mutation in NCF1. Prenat Diagn 22:235–240 216. Chiesa R, Wang J, Blok HJ et al (2020) Hematopoietic cell transplantation in chronic granulomatous disease: a study of 712 children and adults. Blood 136:1201–1211 217. Dedieu C, Albert MH, Mahlaoui N et al (2021) Outcome of chronic granulomatous disease – Conventional treatment vs stem cell transplantation. Pediatr Allergy Immunol 32:576–585 218. Fox TA, Chakraverty R, Burns S et al (2018) Successful outcome following allogeneic hematopoietic stem cell transplantation in adults with primary immunodeficiency. Blood 131:917–931

556 219. Siler U, Paruzynski A, Holtgreve-Grez H et al (2015) Successful combination of sequential gene therapy and rescue Allo-HSCT in two children with X-CGD – importance of timing. Curr Gene Ther 15:416–427 220. Stein S, Ott MG, Schultze-Strasser S et al (2010) Genomic instability and myelodysplasia with monosomy 7 consequent to EVI1 activation after gene therapy for chronic granulomatous disease. Nat Med 16:198–204 221. Santilli G, Almarza E, Brendel C et al (2011) Biochemical correction of X-CGD by a novel chimeric promoter regulating high levels of transgene expression in myeloid cells. Mol Ther 19:122–132 222. Kohn DB, Booth C, Kang EM et al (2020) Lentiviral gene therapy for X-linked chronic granulomatous disease. Nat Med 26: 200–206 223. Schejtman A, Vetharoy W, Choi U et al (2021) Preclinical optimization and safety studies of a new lentiviral gene therapy for p47 (phox)-deficient chronic granulomatous disease. Hum Gene Ther 32:949–958 224. Zou J, Sweeney CL, Chou BK et al (2011) Oxidase-deficient neutrophils from X-linked chronic granulomatous disease iPS cells: functional correction by zinc finger nuclease-mediated safe harbor targeting. Blood 117:5561–5572 225. Grez M, Reichenbach J, Schwäble J et al (2011) Gene therapy of chronic granulomatous disease: the engraftment dilemma. Mol Ther 19:28–35

M. J. Stasia and D. Roos 226. Yahata T, Takanashi T, Muguruma Y et al (2011) Accumulation of oxidative DNA damage restricts the self-renewal capacity of human hematopoietic stem cells. Blood 118:2941–2950 227. Sweeney CL, Zou J, Choi U et al (2017) Targeted repair of CYBB in X-CGD iPSCs requires retention of intronic sequences for expression and functional correction. Mol Ther 25:321–330 228. Miller JC, Tan S, Qiao G et al (2011) A TALE nuclease architecture for efficient genome editing. Nat Biotechnol 29:143–148 229. Zhang F, Cong L, Lodato S et al (2011) Efficient construction of sequence-specific TAL effectors for modulating mammalian transcription. Nat Biotechnol 29:149–153 230. Cong L, Ran FA, Cox D et al (2013) Multiplex genome engineering using CRISPR/Cas systems. Science (New York, NY) 339: 819–823 231. Mali P, Yang L, Esvelt KM et al (2013) RNA-guided human genome engineering via Cas9. Science (New York, NY) 339: 823–826 232. De Ravin SS, Brault J, Meis RJ et al (2021) Enhanced homologydirected repair for highly efficient gene editing in hematopoietic stem/progenitor cells. Blood 137:2598–2608 233. Sweeney CL, Pavel-Dinu M, Choi U et al (2021) Correction of X-CGD patient HSPCs by targeted CYBB cDNA insertion using CRISPR/Cas9 with 53BP1 inhibition for enhanced homologydirected repair. Gene Ther 28:373–390 234. Merling RK, Kuhns DB, Sweeney CL et al (2017) Gene-edited pseudogene resurrection corrects p47(phox)-deficient chronic granulomatous disease. Blood Adv 1:270–278

Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy

33

Giorgia Santilli and Adrian J. Thrasher

Abstract

Keywords

The past 20 years have witnessed massive progress in the early diagnosis of Chronic Granulomatous Disease (CGD), leading to prompt initiation of prophylactic and immunosuppressive treatments with a substantial improvement in life expectancy. Still, only ~50% of CGD patients live into their 40s, depending on the residual NADPH oxidase activity, and the overall quality of life is often poor for patients. Allogeneic hematopoietic stem cell transplantation as definitive therapy has shown excellent results when offered to pediatric patients free from active infections. The outcome is worse if patients suffer from inflammatory complications and fungal infections at the time of transplant. In those instances, the risks associated with the procedure and graft versus host disease remain considerable. Ex vivo correction of autologous hematopoietic stem and progenitor cells through gene therapy holds great promise, but it is more complex than anticipated. This chapter will cover definitive treatments for CGD and discuss the problems encountered in early gene therapy trials focusing on new platforms to overcome pitfalls and maintain the promise of an alternative and safe lifelong treatment for this disease.

Hematopoietic stem cell transplantation · Gene therapy · Gene editing · Lentiviral vectors · Endonucleases

G. Santilli (✉) Molecular and Cellular Immunology Section, UCL Great Ormond Street Institute of Child Health, University College London, London, UK e-mail: [email protected] A. J. Thrasher Molecular and Cellular Immunology Section, UCL Great Ormond Street Institute of Child Health, University College London, London, UK NIHR Great Ormond Street Hospital Biomedical Research Centre, London, UK e-mail: [email protected]

List of Abbreviations AAV ADA-SCID CGD CRISPR/ Cas9 DHR DSB EFS G-CSF GvHD HDR HLA HSCs HSCT HSPCs INDELs iPSCs LTR MFD MUD NHEJ NIH OS PIDs ROS SCID SCID-X1

Adeno-associated virus Adenosine deaminase deficiency-severe combined immunodeficiency Chronic granulomatous disease Clustered regularly interspersed palindromic repeats/associated protein 9 Dihydrorhodamine DNA Double strand break Event free survival Granulocyte colony-stimulating factor Graft versus host disease Homology directed repair Human leukocyte antigen Hematopoietic stem cells Hematopoietic stem cell transplantation Hematopoietic stem and progenitor cells Insertions and deletions Induced pluripotent stem cells Long terminal repeats Matched family donor Matched unrelated donor Non-Homologous end joining National institute of health Overall survival Primary immunodeficiency disorders Reactive oxygen species Severe combined immunodeficiency X-Linked severe combined immunodeficiency

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_33

557

558

G. Santilli and A. J. Thrasher

SFFV SIN TALENs WAS ZFN γRV

Spleen focus forming virus Self-inactivating vector Transcription activator-like effector nucleases Wiskott-aldrich syndrome Zinc finger nucleases Gammaretrovirus/Gammaretroviral vector

1

Introduction

1.1

Hematopoietic Stem Cell Transplantation

The concept of hematopoietic stem cell transplantation (HSCT) was seeded during World War II when scientists frantically searched for humoral or cellular factors that could revert the effects of radiation in survivors of the atomic bomb [1]. Years of studies on irradiated mice led to the discovery that bone marrow could be destroyed through radio- or chemotherapy and replaced with one from a donor [2, 3]. In the late 1950s, clinicians trialled the procedure in humans as a potential cure for blood malignancies. Leukemic patients underwent total body irradiation and infusion of bone marrow from one or more relatives. A few patients went into remission but eventually succumbed to an uncontrolled immune reaction, providing the first evidence of what was later called “graft versus tumour effect” and “graft versus host disease (GvHD)” [4]. During this period, other patients suffering from non-malignant aplastic anaemia and monogenic blood disorders underwent experimental bone marrow transplantation but without success. At that time, the genetic determinants of matching were still undiscovered. The concept of “conditioning” to empty the host bone marrow niche and prevent rejection with minimal toxicity was not refined. The extensive works of the geneticists Peter Gorer and George Snell, who mapped the Major Histocompatibility Locus in mice [5, 6], inspired the French immunologist Jean Dausset who, in 1958, provided the first genetic evidence of the leukocyte alloantigen in humans, later called human leukocyte antigen-A2 (HLA-A2) [7]. He reported that antibodies from patients who had received blood transfusions agglutinated white blood cells from specific sources but not others and subsequently won the Nobel prize for this discovery in 1980 (shared with Baruj Benaceraff and George Snell). Further concerted efforts unveiled the relationship between different leukocyte antigens and their genetics and allowed rapid advances in the field of organ transplantation. As early as 1968, there were the first reports of successful bone marrow transplantation using matched sibling donors for patients with primary immunodeficiency disorders (PIDs). PIDs are

rare monogenic blood disorders that affect children and generally present early in life with failure to thrive, frequent infections, and death in the most severe cases. The reports described the transplant story of two boys affected by X-linked severe combined immunodeficiency (SCID) and one boy affected by Wiskott-Aldrich syndrome (WAS) [8– 10]. SCID patients were, and are still today, regarded as a medical emergency as children are born without a functional adaptive immune system. X-linked SCID is caused by mutations in the gene encoding the common gamma chain unit, shared by multiple cytokine receptors (IL2RG gene) and involved in developing T and NK cells and functional B lymphocytes [11]. Wiskott-Aldrich syndrome (WAS) is an X-linked genetic disorder caused by mutations in a cytoskeleton remodelling protein (the WAS protein), resulting in thrombocytopenia, primary immune deficiency, autoimmunity, and predisposition to lymphoid malignancy [12]. In the following years, scientists refined HLA typing and matching, discovered new, less toxic ways of conditioning, and optimized new protocols to minimize and treat GvHD. As early as the 1980s, thanks to the work of pioneers such as Jon van Rood, Rose Payne, Walter and Julia Bodmer, Bernard Amos, Paul Terasaki, and Ruggero Ceppellini, to name a few, the HLA complex had been mapped on chromosome 6 and found to encode HLA-A, B, C (class I) and DR, DP, DQ (class II) determinants [6]. The current gold standard for matching is high-resolution typing at HLA-A, B, C, DRB1, and DQB1. The development of sensitive molecular typing technologies, the establishment of the Bone Marrow Donors Worldwide (BMDW) database, and recruitment campaigns through social media have greatly improved the identification of well-matched donors. Another crucial step towards enhanced success for allogeneic HSCT has been the introduction of newborn screening [13], unveiling the importance of transplanting patients as early as possible before the establishment of severe infections or inflammatory complications. Scientists have reported, for example, a better outcome when SCID patients undergo transplantation before the onset of infections [14] or WAS patients before splenectomy and at a younger age [15]. Given the quantitative cellular deficiency in most SCID sub-types and the absence of productive thymopoiesis, scientists initially performed bone marrow transplants without pre-emptying the niche to avoid long-term effects such as infertility, cancer, and other endocrine complications. In this setting and using HLA-matched sibling donors, matched unrelated donors, and T cell-depleted grafts from a parent, HSCT resulted in the recovery of T cell output in certain types of SCID, such as SCID-X1. However, in this context, B cells, which are present though only partially functional in SCID-X1, remain largely of recipient origin, indicating a

33

Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy

failure to engraft significant numbers of true multipotent hematopoietic stem cells (HSCs) [16]. On the contrary, unconditioned transplants in other forms of SCID, such as adenosine deaminase (ADA)-SCID (in which T, B and NK cells are all deficient), had better outcomes. ADA-SCID is caused by genetic defects in the adenosine-deaminase enzyme involved in purine metabolism and controlling deoxyATP-mediated toxicity. The complete donor chimerism observed in ADA-SCID transplanted patients is possibly due to the intrinsic genetic defect; indeed, loss of the ADA detoxifying activity precludes the development of all immune cells, dramatically decreasing the cellularity of the bone marrow [17]. Nowadays, clinicians tend to adopt reduced-toxicity (RTC) and reduced-intensity conditioning (RIC) for patients affected by non-ADA forms of SCID. In contrast, patients generally receive more myeloablative conditioning, when full donor chimerism is deemed necessary to avoid later complications related to persisting immunodeficiency or autoimmunity [18]. The transplant protocol for CGD sits somewhere between those two extremes. Whether or not a clinician should offer a transplant, which conditioning regimen to adopt, and which donor source to use very much depends on the history of the individual patient. While it is impossible to generalize on transplant protocols for PIDs, we can safely generalize on the increasingly positive outcomes of recent protocols, with survival rates >80% and long-term disease control and improved quality of life. The high survival rate partly reflects the refined ability to find good HLA matches. Nowadays, Caucasians have a ~>70% chance of finding a fully matched donor (related or unrelated), but this percentage falls to ~ 18 years) who underwent HSCT transplantation between March 1993 and December 2018. The primary endpoints of the study were overall survival (OS) and event (graft failure or death) free survival (EFS).

33

Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy

Most transplants (87%) were performed between 2006 and 2018 when the reduced-intensity conditioning was established, busulfan with fludarabine being the first choice, followed by busulfan and cyclophosphamide, treosulfan and fludarabine, treosulfan fludarabine and thiotepa. The stem cell source was mainly bone marrow, followed by peripheral blood and cord blood for younger patients, and most of the donors were matched at 10/10 HLA-loci (or 6/6 when using a cord blood source), family or unrelated donors. About 20% of patients received mismatched unrelated or family donors. All patients had comorbidities at the time of transplant, mostly gastrointestinal (GI) inflammatory complications. In contrast with previous reports [64], patients with colitis generally had the worst outcome and the highest occurrence of post-transplant complications, as observed by others [65]. Overall survival data were promising, suggesting how the reduced-intensity conditioning with busulfan/treosulfan in the presence or absence of serotherapy is generally well tolerated. The combination with cyclophosphamide seemed to reduce the risk of graft failure but, at the same time, favour GvHD if compared with fludarabine, although this effect could be explained by the lack of serotherapy when using cyclophosphamide. As expected, the use of fully matched family donors (MFD) and fully matched unrelated donors (MUD) resulted in the best outcome in terms of OS (between 87.7 and 89.4%), and EFS was more significant when using a matched donor compared with one or more than one-antigen mismatched donor, with MFD outperforming MUD (85.3 vs 74.4%). Studies on the use of haploidentical donors for HSCT in CGD have also reported favourable outcomes when using myeloablative conditioning in combination with a T-cell receptor (TCR) alpha/beta T-cell depleted grafts [66, 67], although associated with fatal GvHD in some patients [68]. Recent investigations have demonstrated that many female X-CGD carriers suffer from autoimmune syndromes preventing them from acting as haploidentical donors [69]. While the inheritance pattern does not seem to influence the transplant outcome, age (>18) showed a trend of poorer OS, EFS, and increased late occurrence of GvHD. Patients transplanted before 5 years of age have superior outcomes with 100% OS versus 81% of older patients [70]. A recent French/German study compared the long-term outcome of 54 patients under conventional treatment and 50 who opted for HSCT after being on conventional therapy for several years [71]. In contrast to a Swedish study that pointed out a better OS in HSCT vs conventional treatment [72], the survival rate in the French/German study ranged between 76 and 90% in both cohorts. However, transplanted patients regained average growth and had fewer infective episodes than conventional therapy treated ones, leading to the recommendation of transplants for patients younger than 8 years of age and conventional treatment for older patients

563

who seem to manage infections better. The report is in line with previous observations in a more extensive US study [73] and advocates for the use of transplantation at an early age without waiting for general medical conditions to worsen, particularly when a MFD is available. One of the most severe infections in CGD is invasive aspergillosis (IA) which is often fatal and has precluded the use of HSCT as salvage treatment. A recent report of three patients suffering from IA showed that HSCT with reduced-intensity conditioning is now a feasible option when pulmonary aspergillosis cannot be dealt with with aggressive antifungals within 3 months after the initiation of the therapy [71, 74]. At present, a significant other contraindication to transplant is the coexistence of X-CGD and McLeod syndrome. McLeod patients not only suffer from classical CGD but also develop an immune response following red blood cells and granulocyte infusion because of the large deletion in the X-chromosome, leading to a diminished expression of the Kell glycoprotein antigen on erythrocytes [75].

3

Gene Therapy for CGD

3.1

Early Trials

CGD is a monogenic disorder, therefore, amenable to gene therapy. The observation that female carriers of X-CGD with as low as 10–20% functional neutrophils do not suffer from life-threatening infections [76] had fuelled the enthusiasm of clinicians who decided to apply gene therapy protocols previously established for SCID-X1 and ADA-SCID to the treatment of X-linked and autosomal recessive CGD. In 1995, five patients affected by p47phox deficient chronic granulomatous disease (p47CGD) aged between 18 and 37 enrolled in a gene therapy trial led by Harry Malech at the National Institute of Health (NIH), USA [77]. HSPCs were isolated from G-CSF mobilized blood through CD34+ marker selection and transduced in the presence of interleukin 3, granulocytes-macrophage colony-stimulating factor and G-CSF. Patients received the cell product without conditioning and tolerated the gene therapy procedure well. Despite high levels of transduction in the infused product (ranging between 21 and 90%), the number of functional neutrophils, as assessed by the dihydrorhodamine (DHR) test (see Chap. 32 by M.J. Stasia and D. Roos), in the blood of transplanted patients never surpassed 0.051% at a peak around day 22–39 and only in exceptional cases lasted for up to 6 months. In subsequent trials for X-CGD, performed again at NIH and Indiana University [78, 79], scientists improved transduction by using a repeated dosage of viral vector over 4 days rather than 3 days as before and by adopting retronectin to

564

increase viral uptake. Retronectin is a recombinant human fibronectin fragment that colocalizes viral particles and cells by binding the viral envelope via its heparin-binding domain and the cell membrane via interacting with integrin receptors [80]. Despite good correction levels (up to 92%) in the graft, the percentage of functional neutrophils in patients’ blood ranged between 0.06 and 2%, 3–4 weeks after the procedure, to decline further in the following months. Scientists had no reason to believe that the NADPH oxidase subunits confer any selective advantage to hematopoietic lineages, given the lack of revertant mutations or skewed lyonization in favour of the normal allele in female carriers of X-CGD. A new clinical trial at NIH introduced non-suppressive myeloablative conditioning with busulfan to enhance engraftment. Non-myeloablative busulfan conditioning improved initial gene marking and resulted in some (although low) persistence of corrected cells in two of the three patients enrolled in the trial, young adults aged 28, 28, and 19 suffering from active infections at the time of transplant [81]. Interestingly, patient one enrolled in this trial had already been treated by gene therapy in the previous NIH trial with minimal correction. This time, after a peak of correction of 26% at 3 weeks post gene therapy, the patient’s functional neutrophils stabilized at around 1% from 7 months until 2 years (the last time point reported in [81]), suggesting correction and engraftment of a long-term progenitor. A subsequent clinical trial in Seoul showed a similar trend. Correction levels peaked at 6.4% in one patient and 14.5% in another to sharply decrease at 2 months and stabilize at 0.05 and 0.21% after 3 years [82]. Overall, these early clinical trials reported clinical benefits. However, despite optimized transduction protocols and good levels of CD34+ cells in the graft, patients had less than 3% functional neutrophils 3 months after transplantation, insufficient for robust correction of the immunodeficiency. In 2000, a German trial enrolled two X-CGD patients, young adults with a history of infections. Patient one suffered from a liver abscess caused by Staphylococcus and patient two from a pulmonary Aspergillus infection at the time of the gene therapy procedure. The only modification in the transduction protocol was the use of another murine retroviral vector called Spleen Focus Forming Virus (SFFV) to deliver CYBB cDNA encoding gp91phox. Still, the outcome of this trial was surprisingly different from previous ones [83]. Instead of tailing off, corrected cells started to accumulate and peaked at unprecedented levels of 60–80% in both patients around 8 months post gene therapy. This expansion of functional cells, although clonal, was initially welcomed by scientists who saw for the first time a sustained effect of gene therapy. Clear clinical benefits were reported with the resolution of liver abscesses and lung aspergillosis, a common trait of all gene therapy trials for CGD, regardless of the correction level. Unfortunately, and in line with other trials

G. Santilli and A. J. Thrasher

using gammaretroviruses, patients experienced adverse events due to insertional mutagenesis: a myelodysplastic syndrome with consequent loss of chromosome 7 (monosomy 7) [84]. The retroviral integration site distribution revealed activating insertions in both individuals in the MECOM (MDS1-Evi1-complex) locus. The increase in gene-marked cells was restricted to myeloid cells, probably due to the action of the activated Evi1 gene in those clones. Gene marking in B cells was much lower (5–10%) but stable, suggesting that stem cells with long-term repopulating ability were engrafted. Unexpectedly, while neutrophils containing the provirus remained high, up to 24 and 45 months for patient one and two respectively, DHR levels and gp91phox expression declined over time due to the occurrence of CpG methylation in the SFFV LTR promoter resulting in gene silencing even though the mutagenic enhancer effects were unaffected. The first patient died 27 months after gene therapy from sepsis following a dental procedure, and the second patient underwent allogeneic stem cell transplantation 45 months after gene therapy but eventually succumbed of acute myeloid leukaemia (AML). Two pediatric patients were treated in Switzerland with a similar protocol and experienced adverse events but survived following allogeneic HSCT [85]. When a similar procedure was performed on three pediatric patients the correction was minimal and transient, confirming that the clonal expansion of functional cells was vital for the initial clinical success of the German trial.

3.2

New, Myeloid-Specific Vectors for the Gene Therapy of CGD

After the initial enthusiasm, scientists realized that gene therapy for CGD was not an easy task [86]. Firstly, the therapeutic genes in CGD, at odds with those in SCID-X1 or ADA, do not confer any proliferative drive, and virally transduced cells lack a selective advantage over non-corrected ones. Secondly, there were concerns about the safety of full-length γRVs and their ability to confer long-term clinical benefits for this disorder. Among the hypotheses for the failure of the therapy in CGD patients, was the immune response against the gp91phox antigen in individuals with an intact adaptive system. However, using fludarabine in the conditioning regimen to dampen the immune response did not change the outcome in the Seoul trial [82]. Scientists focused their attention on HSCs instead, postulating that viral transduction and ectopic expression of NADPH oxidase subunits could impair stem cells’ ability to home back in the bone marrow and persist in a quiescent state in their niche. Reactive oxygen species (ROS) homeostasis is key in regulating the stem cell niche [87]. Quiescent,

33

Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy

proliferating, or differentiating stem cells exhibit different amounts of intracellular ROS due to their metabolism. Low ROS concentration characterizes self-renewing cells in the osteoblastic niche, while high ROS prompts stem cells of the perivascular niche towards differentiation. ROS has a beneficial effect on stem cells if maintained in a strict concentration range. Fears that overexpression of a single NADPH oxidase subunit could drive ROS production seemed unjustified, given that the active enzyme complex does not form in stem cells. However, ectopic expression of gp91phox in induced pluripotent stem cells (iPSCs)-derived CD34+ cells impedes full neutrophil maturation [88]. To avoid overexpression of gp91phox in HSCs and to limit the toxicity of viral vectors, scientists replaced viral promoters with myeloid-specific ones using safer, SIN lentiviral vector backbones. Studies on the native CYBB promoter revealed a series of repressor and enhancer elements that needed to be included in the lentiviral vector to confer physiological expression to the transgene. A minimal CYBB promoter region combined with a synthetic enhancer rescued high levels of gp91phox expression in myeloid cells [89] and was later combined with target sites for the micro RNA126 (miR-126) in a new lentiviral vector [90]. The miR-126 is highly expressed in CD34+ cells but diminishes with myeloid differentiation to disappear entirely in neutrophils and monocytes [91]. The presence of two miR-126 target sequences in the expression cassette promotes the degradation of the therapeutic RNA only in cells that express miR-126 while sparing myeloid cells, adding an extra layer of specificity to gp91phox expression. A new myeloid promoter was developed in a collaborative effort between Manuel Grez in Frankfurt and Adrian Thrasher in London [92]. This promoter consists of minimal 5′ flanking regions of two myeloid genes, Cathepsin G and C-Fes, hence its name “chimeric,” and contains binding sites for transcription factors highly expressed in neutrophils and monocytes. As a result, the expression of gp91phox driven by the chimeric promoter in the context of a SIN lentiviral vector is high in neutrophils and monocytes, lower in HSCs and B cells, and almost absent in T cells.

3.3

Gene Therapy for CGD: The Present

Extensive preclinical studies with the SIN lentiviral vector containing the chimeric promoter (G1XCGD) were carried out by a European consortium, and highlighted the excellent performance of the vector and its safety profile. A multicentre clinical trial of gene therapy for X-CGD commenced in early 2015. Nine patients enrolled in this trial (Trial registry numbers: NCT02234934, NCT01855685); five were from the USA

565

and three plus one compassionate-used patient from the UK [93]. The trial adopted for the first time myeloablative conditioning to create enough space in the bone marrow for the HSCs graft. The busulfan dose was adjusted based on pharmacokinetics to obtain a total net area under the curve (AUC) of 70,000–75,000 ng/ml*h. The cell dose was also carefully considered; all nine patients received >6 × 106 CD34+ cells/ kg ranging from 6.5 to 32.6 × 106 cells/kg to maximize the engraftment of corrected HSCs in a disease with no selective advantage. HSPCs were mobilized in the blood circulation by treatment with G-CSF and plerixafor, CD34+ cells were isolated, allowed to recover overnight, and then transduced twice with G1XCGD at 1 × 108 infectious genomes/ml to reach between 0.7 and 5.5 average vector copies in the infused product. While gene marking in T cells remained low among all the patients (no T cell/thymic immunodepletion was utilised), B cell and myeloid marking were relatively high and stable over time for several patients. Six young adult patients between 18 and 39 years of age retained DHR values between 16 and 46%, 1-year post gene therapy, while the average vector copies ranged between 0.6 and 0.9. More importantly, and for the first time in the history of CGD gene therapy, patients experienced long-term benefits and stable NADPH oxidase correction over the observation period of 24–36 months in the absence of mutagenesis and clonal expansion. Methylation of the chimeric promoter was ruled out by CpG bisulphite sequencing at 9 months and 15 months-time points, and vector integration studies did not reveal clones (> than 3% abundance) with integration sites near genes of concern. Six patients achieved significant clinical benefits, and all but one (a patient who had to remain on prophylaxis antibiotics due to a low number of lymphocytes and lung disease) stopped prophylactic antibiotics or antifungal treatment. Two pediatric patients succumbed shortly after the procedure for causes most likely unrelated to the procedure. One patient developed idiopathic pneumonia, aggravated by the absence of one lung due to aspergillosis-related pneumonectomy. The other patient experienced bleeding in the brain following platelet destruction due to an autoimmune reaction consequent to several granulocyte infusions for his deepseated infections. One other pediatric patient only had transient engraftment of corrected cells. Since 2020 four patients have been treated with the same gene therapy protocol, three pediatric (ranging from 3 to 11 years of age) and one adult, 31 years old. While the adult patient had sustained high levels of DHR (77% of DHR positive neutrophils at 6 months after the procedure), the pediatric patients showed a quick decline in the number of corrected cells, a similar outcome observed in early trials.

566

G. Santilli and A. J. Thrasher

Table 33.1 summarizes the outcomes of gene therapy trials for CGD.

3.4

Clinical Success: Are We There Yet?

Despite the undoubted success of this trial, the fact that some patients, surprisingly the pediatric ones (although numbers remain small), did not experience long-term benefits remains puzzling. There was no noticeable difference in the graft or the state of active infection at the time of the procedure to explain this outcome. One consideration is that HSCs from pediatric and adult individuals are intrinsically different; lineage-balanced the first ones and lineage-biased the second ones, as they preferentially develop into myeloid cells concomitantly with thymic regression. One also needs to take into-account that the HSCs pool in CGD patients is affected by chronic inflammation. Adult patients likely have compensatory mechanisms that enable them to survive infections and adjust to chronic inflammation. Studies on the X-CGD mouse model have revealed an intrinsic inflammatory background. These mice are kept in a sterile environment. Therefore, they are free from infections, yet their bone marrow contains elevated levels of interleukin 1 beta (IL1-β) [94]. High amounts of the proinflammatory IL1-β affect the quality of HSCs that are less in the G0 phase of the cell cycle. When transplanting Lineage negative, Sca+, c-kit+; LSK cells (the murine counterpart of human HSCs) taken from X-CGD mice and those from wild-type mice, the output in the early weeks is mainly from CGD cells. However, at later time points, the contribution of CGD cells is outperformed by the wild-type ones as a consequence of the ‘active’ state of CGD HSCs that primes them to exhaustion after a few months [94]. CGD mice do not suffer from abnormal haematopoiesis, leading to the conclusion that the transplant procedure/in vitro manipulation has a deleterious impact on the activation of the HSCs pool, unveiling its compromised quality only after a secondary challenge. CGD individuals suffer from chronic inflammation resulting in defective autophagy leading to release of IL1-β [95]. Regardless of the presence of active infections, IL1-β, tumour necrosis factor alpha (TNF-α), interleukin 6 (IL6), and inflammatory monocytes/macrophages are high in the blood and bone marrow of CGD patients [96, 97]. TNF-α and IL1-β, for example, trigger the overexpression of interleukin 27 receptor alpha (IL27RA) in HSCs, leading to myeloid skewing and reduced hematopoietic potential [98]. It is now widely accepted that inflammation activates HSCs and promotes myelopoiesis [99]. Therefore, it is plausible that gene therapy in CGD targets an already compromised population. One possibility is to

treat CGD patients with a combination of anti-inflammatory drugs such as anakinra, an IL1-β receptor antagonist, ustekinumab, an antibody targeting interleukin 12 (IL12) and 23 (IL23) and rapamycin, before cell harvest to enhance the quality of the HSPCs pool for ex vivo gene therapy [94, 100, 101]. Short transduction protocols could also preserve such a fragile pool’s self-renewal and repopulation potential. Of note, gene therapy protocols for CGD have always relied on a high dosage of vector and ex vivo culturing of cells for 3–4 days. The past 10 years have experienced a global upsurge in new protocols designed to shorten ex vivo manipulation of CD34+ cells. For example, the use of transduction enhancers can lead to a significant reduction of viral dosage and cycles of transduction, impacting both production costs and the quality of the graft [102]. Withdrawal of activating cytokines from culture media and the addition of small drugs have proved effective in preserving the stem cell phenotype of HSCs from different sources [103–105]. Some of those protocols have already made it into the clinic for the gene therapy of metabolic disorders, and they could hold the key to success of autologous gene therapy for diseases like CGD, where there is a strong inflammatory component, and HSCs cells seem primed to exhaustion and less able to tolerate in vitro manipulation.

3.5

New Platforms: Gene Editing for CGD

Despite progress in obtaining physiological levels of gp91phox expression, ectopic expression of gp91phox is still a concern for the longevity of the therapy. Moreover, despite the creation of safer lentiviral vectors, reports of severe adverse events observed in the gene therapy trials for X-linked adrenoleukodystrophy highlight the potential risks of insertional oncogenesis, although the lentiviral vectors used in this study have an internal gammaretroviral LTR for gene expression (unpublished data). A way to overcome the issue of unregulated expression of gp91phox or insertional mutagenesis is by using gene editing techniques. Gene editing can be used to target the integration of a gene of interest in a safe harbour in the genome. Targeted insertion into the Adeno-Associated Virus integration site 1 (AAVS1) safe harbour of a DNA cassette containing the coding sequences for gp91phox, p47phox, p67phox, p22phox and p40phox has been achieved in iPSCs using ZFN nucleases [106]. While this is an attractive approach, the use of an ectopic promoter to drive the expression of the therapeutic genes potentially nullifies some of the advantages of a targeted approach. TALENs and CRISPR/ Cas9 systems have been exploited to mediate targeted integration of the whole gp91phox coding sequence next to its regulatory sequence in iPSCs cells from X-CGD patients

Busulfan (AUC: 70,000–75,000 ng/ml*h; 12 mg/kg) Busulfan (AUC: 70,000–75,000 ng/ml*h; 12 mg/kg) N/A

4

9

London NCT01381003

Multicentre (USA, EU) NCT02234934 NCT01855685 France NCT02757911

10

39.5–45%

25.3–32.7%

25–73% 10.8–28.9%

5–20%

49.1–79.2%

N/A

N/A

3.6–5.1 × 106 6–19 × 106 18.9–71 × 106 5.4–5.8 × 106 0.2–10 × 106 6.5-32.6 × 106

N/A

N/A

MFGS-gp91phox MT-gp91phox

MFGS-gp91phox SF71-gp91phox (SFFV-LTR) G1XCGD (pCCLChimgp91/ VSVg-LV) G1XCGD (pCCLChimgp91/ VSVg-LV) LV-CGD p47phox/ gp91phox

SF71-gp91phox (SFFV-LTR)

SF71-gp91phox (SFFV-LTR)

N/A

45–92%

2–44 × 106 N/A

MSCV-gp91phox

MFGS-gp91phox

Vector used MFGS-p47phox

% transduction efficiency 21–90%

Dose of infused CD34+ cells/ kg 0.1–4.7 × 106

N/A

N/A

Yes

Yes

Yes

Yes

Yes

N/A

N/A

16–46% (six out of nine)

No

No

Myeloproliferation and MDS with monosomy 7 Myeloproliferation and MDS with monosomy 7 No

>15% gene marking in CD15+ cells ~20% gene marking in CD15+ cells No

Yes

No

No

–

N/A

N/A

No

No

No

No

No

–

Genotoxicity No

Engraftment (>2%) >3months No

Clinical benefits –

No data yet

No data yet

Thrasher AJ, personal comm Kohn DB et al. [94]

Kang EM et al. [81] Kang HJ et al. [82]

References Malech HL et al. [77] Malech HL et al. [78] Barese CN et al. [79] Ott MG et al. [83], Stein et al. [84] Siler et al. [85]

Transduction efficiency is expressed as percentage of vector positive colonies (in the multicentre trial) and as percentage of p47phox or gp91phox positive cells (other trials) This table is adapted from Grez et al. [86] LTR long terminal repeats, MSCV murine stem cell virus, MT, MFGS vectors based on the moloney leukaemia virus, SFFV spleen focus forming virus, Chim chimeric promoter (made by the fusion of 5′ minimal flanking regions of Cathepsin G and c-Fes), VSVg vesicular stomatitis virus, LV lentiviral vector, AUC area under the curve, N/A not available

Shenzhen, China NCT03645486

Busulfan (6.4 mg/kg) + Fludarabine (120 mg/ m2) Melphalan (140 mg/ m2)

2

4

Busulfan (10 mg/kg)

3

NIH, USA NCT00394316 Seoul NCT00778882

Liposomal Busulfan (8.8 mg/kg)

2

Zurich NCT00927134

No conditioning

2 Liposomal Busulfan (8 mg/kg)

No conditioning

4

2

Total conditioning No conditioning

Frankfurt NCT00564759

Center NIH, USA NCT00001476 NIH, USA NCT00001476 Indiana, USA

Patients treated 5

Table 33.1 Summary of gene therapy trials for chronic granulomatous disease

33 Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy 567

568

[107]. This strategy has unveiled the importance of introns, particularly early introns when relying on endogenous regulatory sequences to drive the expression of the therapeutic gene. Indeed, the first CYBB intron is required to express gp91phox, and only Cas9 systems targeting exon two and mediating the insertion of a cDNA from exon two onwards were able to achieve normal gp91phox expression, and the rescue of ROS production in iPSCs derived myeloid cells in vitro. Given the high frequencies of mutations in CYBB exon 7, a strategy has been developed to target exon 7 and thereby repair the mutation C676T with a single stranded oligonucleotide [108]. Targeted DNA insertion achieved more than 20% functional neutrophils in vitro. More importantly, manipulated HSPCs were able to engraft immunodeficient mice and provide ~10% correction levels over time. Currently, AAV6 is considered the elective tool to deliver entire coding sequences as donor template for efficient HSPCs gene editing, given its minimal rates of random integration and high recombination frequencies even when small homology regions are provided. However, achieving complete disease correction through gene editing remains cumbersome for diseases like CGD without a selective advantage. Moreover, the combined effects of p53 activation in response to a DSB and AAV6 transduction impact the repopulating/homing ability of HSPCs, and often high frequencies HDR translate into poor engraftment. Scientists are searching new ways to dampen the p53-mediated DNA damaging response while increasing HDR rates [109, 110]. While much effort has been put into developing gene therapy strategies for the X-linked form of CGD, given its prevalence and the “general” notion that this is the most severe form of the disease, autosomal recessive CGD has been neglected for some time [111]. However, recent reports have highlighted how regardless of the subunit affected, the amount of residual NADPH oxidase activity dictates the severity of the disease and probably other still unknown genetic components [112]. Also, the frequency of the autosomal recessive disorders varies with the degree of consanguineous marriages, so p47phox deficient CGD (p47CGD) is the most prevalent in Turkey [113], and p67phox deficient CGD (p67CGD) is the most prevalent type recorded in Jordan [114]. The London group has recently developed a gene therapy strategy for p47CGD using a conventional lentiviral gene therapy approach [115]. A similar lentiviral vector to that used in the X-CGD trial was able to correct >40% of murine p47CGD myeloid cells and reduce the bacterial burden in p47CGD mice challenged with Salmonella typhimurium. Extensive genotoxicity studies have highlighted the excellent safety profile of this lentiviral vector that is about to enter the clinic in the UK and USA [116]. The hotspot mutation in NCF1 encoding p47phox, a deletion of the GT dinucleotide at the beginning of exon 2 (see Chap. 32 by M.J. Stasia and D. Roos), has fuelled the interest

G. Santilli and A. J. Thrasher

in a gene editing approaches. The GT deletion has probably arisen from a homologous recombination event between the NCF1 gene and its pseudogenes that contain the dinucleotide deletion but are identical to NFC1 for the remaining part. Therefore, it is tricky to create a double-strand break that targets the NCF1 gene while sparing the pseudogenes. One could instead use this promiscuity to achieve correction of the faulty gene alongside the functional ‘resurrection’ of the pseudogenes [117]. However, HDR-mediated insertion of the correct DNA sequence remains cumbersome for p47CGD, and many events end up with translocations between genes and pseudogenes that have so far impeded the clinical progression of such an approach [118].

4

Concluding Remarks

From the early application of gamma retroviral gene therapy for CGD without patient conditioning remarkable progress has been made. Protocols now exist that have achieved clinical efficacy in a number of patients using improved vectors and cell processing, along with conditioning regimens that mirror this in allogeneic HSCT. Gene editing approaches, although in their infancy, are an exciting prospect. There is, however, still much to learn about the biology of the stem cell compartment in CGD, and irrespective of the technology used, suppression of a pro-inflammatory environment may well be necessary to ensure successful and reliable engraftment of gene modified cells. Alternatively, there is considerable interest in developing in vivo approaches whereby HSCs can be targeted without the need for harmful ex vivo manipulation, although it may be some time before the technology for this is ready for clinical application. Acknowledgements A.J.T. is supported by the Wellcome Trust. G.S and A.J.T. are supported by the UK National Institute for Health Research Biomedical Research Centre at Great Ormond Street Hospital for Children NHS Foundation Trust and University College London.

References 1. Granot N, Storb R (2020) History of hematopoietic cell transplantation: challenges and progress. Haematologica 105(12): 2716–2729. https://doi.org/10.3324/haematol.2019.245688 2. Jacobson LO, Simmons EL, Marks EK, Robson MJ, Bethard WF, Gaston EO (1950) The role of the spleen in radiation injury and recovery. J Lab Clin Med 35(5):746–770 3. Lorenz E, Uphoff D, Reid TR, Shelton E (1951) Modification of irradiation injury in mice and Guinea pigs by bone marrow injections. J Natl Cancer Inst 12(1):197–201 4. Mathe G, Amiel JL, Schwarzenberg L, Cattan A, Schneider M (1965) Adoptive immunotherapy of acute leukemia: experimental and clinical results. Cancer Res 25(9):1525–1531 5. Gorer PA, Lyman S, Snell GD (1948) Studies on the genetic and antigenic basis of tumour transplantation linkage between a

33

Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy

histocompatibility gene and ‘fused’ in mice. Proc R Soc B Biol Sci 135:499–505. https://doi.org/10.1098/rspb.1948.0026 6. Thorsby E (2009) A short history of HLA. Tissue Antigens 74(2): 101–116. https://doi.org/10.1111/j.1399-0039.2009.01291.x 7. Carosella ED (2009) From MAC to HLA: professor Jean Dausset, the pioneer. Hum Immunol 70(9):661–662. https://doi.org/10. 1016/j.humimm.2009.07.010 8. Bach FH, Albertini RJ, Joo P, Anderson JL, Bortin MM (1968) Bone-marrow transplantation in a patient with the Wiskott-Aldrich syndrome. Lancet 2(7583):1364–1366. https://doi.org/10.1016/ s0140-6736(68)92672-x 9. De Koning J, Van Bekkum DW, Dicke KA, Dooren LJ, Radl J, Van Rood JJ (1969) Transplantation of bone-marrow cells and fetal thymus in an infant with lymphopenic immunological deficiency. Lancet 1(7608):1223–1227. https://doi.org/10.1016/s0140-6736 (69)92112-6 10. Gatti RA, Meuwissen HJ, Allen HD, Hong R, Good RA (1968) Immunological reconstitution of sex-linked lymphopenic immunological deficiency. Lancet 2(7583):1366–1369. https://doi.org/10. 1016/s0140-6736(68)92673-1 11. Sponzilli I, Notarangelo LD (2011) Severe combined immunodeficiency (SCID): from molecular basis to clinical management. Acta Biomed 82(1):5–13 12. Ochs HD, Thrasher AJ (2006) The Wiskott-Aldrich syndrome. J Allergy Clin Immunol 117(4):725–738. https://doi.org/10.1016/j. jaci.2006.02.005 13. King JR, Hammarstrom L (2018) Newborn screening for primary immunodeficiency diseases: history, current and future practice. J Clin Immunol 38(1):56–66. https://doi.org/10.1007/s10875-0170455-x 14. Pai SY, Logan BR, Griffith LM et al (2014) Transplantation outcomes for severe combined immunodeficiency, 2000–2009. N Engl J Med 371(5):434–446. https://doi.org/10.1056/ NEJMoa1401177 15. Ozsahin H, Cavazzana-Calvo M, Notarangelo LD et al (2008) Long-term outcome following hematopoietic stem-cell transplantation in Wiskott-Aldrich syndrome: collaborative study of the European Society for Immunodeficiencies and European Group for Blood and Marrow Transplantation. Blood 111(1):439–445. https://doi.org/10.1182/blood-2007-03-076679 16. Slatter MA, Gennery AR (2018) Hematopoietic cell transplantation in primary immunodeficiency—conventional and emerging indications. Expert Rev Clin Immunol 14(2):103–114. https://doi. org/10.1080/1744666X.2018.1424627 17. Hassan A, Booth C, Brightwell A et al (2012) Outcome of hematopoietic stem cell transplantation for adenosine deaminasedeficient severe combined immunodeficiency. Blood 120(17): 3615–3624; quiz 3626. https://doi.org/10.1182/blood-201112-396879 18. Moratto D, Giliani S, Bonfim C et al (2011) Long-term outcome and lineage-specific chimerism in 194 patients with WiskottAldrich syndrome treated by hematopoietic cell transplantation in the period 1980–2009: an international collaborative study. Blood 118(6):1675–1684. https://doi.org/10.1182/blood-201011-319376 19. Wirth T, Parker N, Yla-Herttuala S (2013) History of gene therapy. Gene 525(2):162–169. https://doi.org/10.1016/j.gene.2013.03.137 20. Sambrook J, Westphal H, Srinivasan PR, Dulbecco R (1968) The integrated state of viral DNA in SV40-transformed cells. Proc Natl Acad Sci USA 60(4):1288–1295. https://doi.org/10.1073/pnas.60. 4.1288 21. Rogers S, Pfuderer P (1968) Use of viruses as carriers of added genetic information. Nature 219(5155):749–751. https://doi.org/ 10.1038/219749a0 22. Blaese RM, Culver KW, Miller AD et al (1995) T lymphocytedirected gene therapy for ADA-SCID: initial trial results after

569

4 years. Science 270(5235):475–480. https://doi.org/10.1126/ science.270.5235.475 23. Bordignon C, Notarangelo LD, Nobili N et al (1995) Gene therapy in peripheral blood lymphocytes and bone marrow for ADAimmunodeficient patients. Science 270(5235):470–475. https:// doi.org/10.1126/science.270.5235.470 24. Raper SE, Chirmule N, Lee FS et al (2003) Fatal systemic inflammatory response syndrome in a ornithine transcarbamylase deficient patient following adenoviral gene transfer. Mol Genet Metab 80(1–2):148–158. https://doi.org/10.1016/j.ymgme.2003.08.016 25. Vandenberghe LH, Wilson JM, Gao G (2009) Tailoring the AAV vector capsid for gene therapy. Gene Ther 16(3):311–319. https:// doi.org/10.1038/gt.2008.170 26. Zincarelli C, Soltys S, Rengo G, Rabinowitz JE (2008) Analysis of AAV serotypes 1-9 mediated gene expression and tropism in mice after systemic injection. Mol Ther 16(6):1073–1080. https://doi. org/10.1038/mt.2008.76 27. Salganik M, Hirsch ML, Samulski RJ (2015) Adeno-associated virus as a mammalian DNA vector. Microbiol Spectr 3(4). https:// doi.org/10.1128/microbiolspec.MDNA3-0052-2014 28. Gaudet D, Methot J, Kastelein J (2012) Gene therapy for lipoprotein lipase deficiency. Curr Opin Lipidol 23(4):310–320. https:// doi.org/10.1097/MOL.0b013e3283555a7e 29. Cideciyan AV, Aleman TS, Boye SL et al (2008) Human gene therapy for RPE65 isomerase deficiency activates the retinoid cycle of vision but with slow rod kinetics. Proc Natl Acad Sci USA 105(39):15112–15117. https://doi.org/10.1073/pnas.0807027105 30. Russell S, Bennett J, Wellman JA et al (2017) Efficacy and safety of voretigene neparvovec (AAV2-hRPE65v2) in patients with RPE65-mediated inherited retinal dystrophy: a randomised, controlled, open-label, phase 3 trial. Lancet 390(10097):849–860. https://doi.org/10.1016/S0140-6736(17)31868-8 31. Mendell JR, Al-Zaidy S, Shell R et al (2017) Single-dose genereplacement therapy for spinal muscular atrophy. N Engl J Med 377(18):1713–1722. https://doi.org/10.1056/NEJMoa1706198 32. Ferrari G, Thrasher AJ, Aiuti A (2021) Gene therapy using haematopoietic stem and progenitor cells. Nat Rev Genet 22(4): 216–234. https://doi.org/10.1038/s41576-020-00298-5 33. Cavazzana-Calvo M, Hacein-Bey S, de Saint BG et al (2000) Gene therapy of human severe combined immunodeficiency (SCID)-X1 disease. Science 288(5466):669–672. https://doi.org/10.1126/sci ence.288.5466.669 34. Gaspar HB, Parsley KL, Howe S et al (2004) Gene therapy of X-linked severe combined immunodeficiency by use of a pseudotyped gammaretroviral vector. Lancet 364(9452): 2181–2187. https://doi.org/10.1016/S0140-6736(04)17590-9 35. Bousso P, Wahn V, Douagi I et al (2000) Diversity, functionality, and stability of the T cell repertoire derived in vivo from a single human T cell precursor. Proc Natl Acad Sci USA 97(1):274–278. https://doi.org/10.1073/pnas.97.1.274 36. Hacein-Bey-Abina S, Garrigue A, Wang GP et al (2008) Insertional oncogenesis in 4 patients after retrovirus-mediated gene therapy of SCID-X1. J Clin Invest 118(9):3132–3142. https://doi. org/10.1172/JCI35700 37. Hacein-Bey-Abina S, von Kalle C, Schmidt M et al (2003) A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N Engl J Med 348(3): 255–256. https://doi.org/10.1056/NEJM200301163480314 38. Schwarzwaelder K, Howe SJ, Schmidt M et al (2007) Gammaretrovirus-mediated correction of SCID-X1 is associated with skewed vector integration site distribution in vivo. J Clin Investig 117(8):2241–2249. https://doi.org/10.1172/Jci31661 39. Aiuti A, Cattaneo F, Galimberti S et al (2009) Gene therapy for immunodeficiency due to adenosine deaminase deficiency. New Engl J Med 360(5):447–458. https://doi.org/10.1056/ NEJMoa0805817

570 40. Aiuti A, Slavin S, Aker M et al (2002) Correction of ADA-SCID by stem cell gene therapy combined with nonmyeloablative conditioning. Science 296(5577):2410–2413. https://doi.org/10.1126/ science.1070104 41. Aiuti A, Roncarolo MG, Naldini L (2017) Gene therapy for ADA-SCID, the first marketing approval of an ex vivo gene therapy in Europe: paving the road for the next generation of advanced therapy medicinal products. EMBO Mol Med 9(6):737–740. https://doi.org/10.15252/emmm.201707573 42. Cicalese MP, Ferrua F, Castagnaro L et al (2016) Update on the safety and efficacy of retroviral gene therapy for immunodeficiency due to adenosine deaminase deficiency. Blood 128(1):45–54. https://doi.org/10.1182/blood-2016-01-688226 43. Cattoglio C, Facchini G, Sartori D et al (2007) Hot spots of retroviral integration in human CD34+ hematopoietic cells. Blood 110(6):1770–1778. https://doi.org/10.1182/blood-200701-068759 44. Deichmann A, Brugman MH, Bartholomae CC et al (2011) Insertion sites in engrafted cells cluster within a limited repertoire of genomic areas after Gammaretroviral vector gene therapy. Mol Ther 19(11):2031–2039. https://doi.org/10.1038/mt.2011.178 45. Derse D, Crise B, Li Y et al (2007) Human T-cell leukemia virus type 1 integration target sites in the human genome: comparison with those of other retroviruses. J Virol 81(12):6731–6741. https:// doi.org/10.1128/Jvi.02752-06 46. Cooray S, Howe SJ, Thrasher AJ (2012) Retrovirus and lentivirus vector design and methods of cell conditioning. Method Enzymol 507:29–57. https://doi.org/10.1016/B978-0-12-386509-0.00003-X 47. Aiuti A, Biasco L, Scaramuzza S et al (2013) Lentiviral hematopoietic stem cell gene therapy in patients with WiskottAldrich syndrome. Science 341(6148):865–U871. https://doi.org/ 10.1126/science.1233151 48. Ferrua F, Cicalese MP, Galimberti S et al (2019) Lentiviral haemopoietic stem/progenitor cell gene therapy for treatment of Wiskott-Aldrich syndrome: interim results of a non-randomised, open-label, phase 1/2 clinical study. Lancet Haematol 6(5):E239– E253. https://doi.org/10.1016/S2352-3026(19)30021-3 49. Kohn DB, Booth C, Shaw KL et al (2021) Autologous ex vivo lentiviral gene therapy for adenosine deaminase deficiency. New Engl J Med 384(21):2002–2013. https://doi.org/10.1056/ NEJMoa2027675 50. Mamcarz E, Zhou S, Lockey T et al (2019) Lentiviral gene therapy combined with low-dose Busulfan in infants with SCID-X1. New Engl J Med 380(16):1525–1534. https://doi.org/10.1056/ NEJMoa1815408 51. Tucci F, Scaramuzza S, Aiuti A, Mortellaro A (2021) Update on clinical ex vivo hematopoietic stem cell gene therapy for inherited monogenic diseases. Mol Ther 29(2):489–504. https://doi.org/10. 1016/j.ymthe.2020.11.020 52. Doudna JA (2020) The promise and challenge of therapeutic genome editing. Nature 578(7794):229–236. https://doi.org/10. 1038/s41586-020-1978-5 53. Porter SN, Levine RM, Pruett-Miller SM (2019) A practical guide to genome editing using targeted nuclease technologies. Compr Physiol 9(2):665–714. https://doi.org/10.1002/cphy.c180022 54. Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna JA, Charpentier E (2012) A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science 337(6096): 816–821. https://doi.org/10.1126/science.1225829 55. Xu L, Yang H, Gao Y et al (2017) CRISPR/Cas9-mediated CCR5 ablation in human hematopoietic stem/progenitor cells confers HIV-1 resistance in vivo. Mol Ther 25(8):1782–1789. https://doi. org/10.1016/j.ymthe.2017.04.027 56. Frangoul H, Ho TW, Corbacioglu S (2021) CRISPR-Cas9 gene editing for sickle cell disease and beta-thalassemia. N Engl J Med 384(23):e91. https://doi.org/10.1056/NEJMc2103481

G. Santilli and A. J. Thrasher 57. Schiroli G, Ferrari S, Conway A et al (2017) Preclinical modeling highlights the therapeutic potential of hematopoietic stem cell gene editing for correction of SCID-X1. Sci Transl Med 9(411): eaan0820. https://doi.org/10.1126/scitranslmed.aan0820 58. Schiroli G, Conti A, Ferrari S et al (2019) Precise gene editing preserves hematopoietic stem cell function following transient p53-mediated DNA damage response. Cell Stem Cell 24(4): 551–565 e558. https://doi.org/10.1016/j.stem.2019.02.019 59. Anzalone AV, Randolph PB, Davis JR et al (2019) Search-andreplace genome editing without double-strand breaks or donor DNA. Nature 576(7785):149–157. https://doi.org/10.1038/ s41586-019-1711-4 60. Wang J, Zhou H, Wang G et al (2022) Efficient targeted insertion of large DNA fragments without DNA donors. Nat Methods 19(3): 331–340. https://doi.org/10.1038/s41592-022-01399-1 61. Gillmore JD, Maitland ML, Lebwohl D (2021) CRISPR-Cas9 in vivo gene editing for transthyretin amyloidosis. Reply. N Engl J Med 38 5(18 ):1722– 1723. https://d oi.org/10.1056 / NEJMc2114592 62. Gungor T, Teira P, Slatter M et al (2014) Reduced-intensity conditioning and HLA-matched haemopoietic stem-cell transplantation in patients with chronic granulomatous disease: a prospective multicentre study. Lancet 383(9915):436–448. https://doi.org/10. 1016/S0140-6736(13)62069-3 63. Chiesa R, Wang J, Blok HJ et al (2020) Hematopoietic cell transplantation in chronic granulomatous disease: a study of 712 children and adults. Blood 136(10):1201–1211. https://doi.org/10. 1182/blood.2020005590 64. Marsh RA, Leiding JW, Logan BR et al (2019) Chronic granulomatous disease-associated IBD resolves and does not adversely impact survival following allogeneic HCT. J Clin Immunol 39(7): 653–667. https://doi.org/10.1007/s10875-019-00659-8 65. Connelly JA, Marsh R, Parikh S, Talano JA (2018) Allogeneic hematopoietic cell transplantation for chronic granulomatous disease: controversies and state of the art. J Pediatric Infect Dis Soc 7 (suppl_1):S31–S39. https://doi.org/10.1093/jpids/piy015 66. Hoenig M, Niehues T, Siepermann K et al (2014) Successful HLA haploidentical hematopoietic SCT in chronic granulomatous disease. Bone Marrow Transplant 49(10):1337–1338. https://doi.org/ 10.1038/bmt.2014.125 67. Parta M, Hilligoss D, Kelly C et al (2015) Haploidentical hematopoietic cell transplantation with post-transplant cyclophosphamide in a patient with chronic granulomatous disease and active infection: a first report. J Clin Immunol 35(7):675–680. https://doi. org/10.1007/s10875-015-0204-y 68. Parta M, Hilligoss D, Kelly C et al (2020) Failure to prevent severe graft-versus-host disease in Haploidentical hematopoietic cell transplantation with post-transplant cyclophosphamide in chronic granulomatous disease. J Clin Immunol 40(4):619–624. https://doi. org/10.1007/s10875-020-00772-z 69. Chiriaco M, Salfa I, Ursu GM et al (2021) Immunological aspects of X-linked chronic granulomatous disease female carriers. Antioxidants 10(6):891. https://doi.org/10.3390/antiox10060891 70. Lum SH, Flood T, Hambleton S et al (2019) Two decades of excellent transplant survival for chronic granulomatous disease: a supraregional immunology transplant center report. Blood 133(23):2546–2549. https://doi.org/10.1182/blood.2019000021 71. Dedieu C, Albert MH, Mahlaoui N et al (2020) Outcome of chronic granulomatous disease – conventional treatment vs stem cell transplantation. Pediatr Allergy Immunol 32(3):576–585. https://doi. org/10.1111/pai.13402 72. Ahlin A, Fugelang J, de Boer M, Ringden O, Fasth A, Winiarski J (2013) Chronic granulomatous disease-haematopoietic stem cell transplantation versus conventional treatment. Acta Paediatr 102(11):1087–1094. https://doi.org/10.1111/apa.12384

33

Definitive Treatments for Chronic Granulomatous Disease with a Focus on Gene Therapy

73. Yonkof JR, Gupta A, Fu PF, Garabedian E, Dalal J, Network UI (2019) Role of allogeneic hematopoietic stem cell transplant for chronic granulomatous disease (CGD): a report of the United States Immunodeficiency Network. J Clin Immunol 39(4):448–458. https://doi.org/10.1007/s10875-019-00635-2 74. Marsh RA, Leiding JW, Logan BR et al (2020) Chronic granulomatous disease-associated IBD resolves and does not adversely impact survival following allogeneic HCT. J Clin Immunol 39(7): 653–667. https://doi.org/10.1007/s10875-020-00852-0 75. Lhomme F, Peyrard T, Babinet J et al (2020) Chronic granulomatous disease with the McLeod phenotype: a French National Retrospective Case Series. J Clin Immunol 40(5):752–762. https://doi. org/10.1007/s10875-020-00791-w 76. Marciano BE, Zerbe CS, Falcone EL et al (2018) X-linked carriers of chronic granulomatous disease: illness, lyonization, and stability. J Allergy Clin Immunol 141(1):365–371. https://doi.org/10. 1016/j.jaci.2017.04.035 77. Malech HL, Maples PB, Whiting-Theobald N et al (1997) Prolonged production of NADPH oxidase-corrected granulocytes after gene therapy of chronic granulomatous disease. Proc Natl Acad Sci USA 94(22):12133–12138. https://doi.org/10.1073/ pnas.94.22.12133 78. Malech HHM, Whiting-Theobald N, Linton G, Miller J, Holland S et al (2000) Multiple cycles of ex-vivo gene therapy for X-linked chronic granulomatous disease (CGD) sustain production of oxidase-normal peripheral blood neutrophils. Mol Ther 1(5 Suppl):S146 79. Barese CN, Goebel WS, Dinauer MC (2004) Gene therapy for chronic granulomatous disease. Expert Opin Biol Ther 4(9): 1423–1434. https://doi.org/10.1517/14712598.4.9.1423 80. van der Loo JCM, Xiao XL, McMillin D, Hashino K, Kato I, Williams DA (1998) VLA-5 is expressed by mouse and human long-term repopulating hematopoietic cells and mediates adhesion to extracellular matrix protein fibronectin. J Clin Investig 102(5): 1051–1061. https://doi.org/10.1172/Jci3687 81. Kang EM, Choi U, Theobald N et al (2010) Retrovirus gene therapy for X-linked chronic granulomatous disease can achieve stable long-term correction of oxidase activity in peripheral blood neutrophils. Blood 115(4):783–791. https://doi.org/10.1182/blood2009-05-222760 82. Kang HJ, Bartholomae CC, Paruzynski A et al (2011) Retroviral gene therapy for X-linked chronic granulomatous disease: results from phase I/II trial. Mol Ther 19(11):2092–2101. https://doi.org/ 10.1038/mt.2011.166 83. Ott MG, Schmidt M, Schwarzwaelder K et al (2006) Correction of X-linked chronic granulomatous disease by gene therapy, augmented by insertional activation of MDS1-EVI1, PRDM16 or SETBP1. Nat Med 12(4):401–409. https://doi.org/10.1038/ nm1393 84. Stein S, Ott MG, Schultze-Strasser S et al (2010) Genomic instability and myelodysplasia with monosomy 7 consequent to EVI1 activation after gene therapy for chronic granulomatous disease. Nat Med 16(2):198–204. https://doi.org/10.1038/nm.2088 85. Siler U, Paruzynski A, Holtgreve-Grez H et al (2015) Successful combination of sequential gene therapy and rescue Allo-HSCT in two children with X-CGD – importance of timing. Curr Gene Ther 15(4):416–427. https://doi.org/10.2174/ 1566523215666150515145255 86. Grez M, Reichenbach J, Schwable J, Seger R, Dinauer MC, Thrasher AJ (2011) Gene therapy of chronic granulomatous disease: the engraftment dilemma. Mol Ther 19(1):28–35. https://doi. org/10.1038/mt.2010.232 87. Kumar S, Geiger H (2017) HSC niche biology and HSC expansion ex vivo. Trends Mol Med 23(9):799–819. https://doi.org/10.1016/ j.molmed.2017.07.003

571

88. Lin HT, Masaki H, Yamaguchi T et al (2015) An assessment of the effects of ectopic gp91phox expression in XCGD iPSC-derived neutrophils. Mol Ther Methods Clin Dev 2:15046. https://doi.org/ 10.1038/mtm.2015.46 89. Barde I, Laurenti E, Verp S et al (2011) Lineage- and stagerestricted lentiviral vectors for the gene therapy of chronic granulomatous disease. Gene Ther 18(11):1087–1097. https://doi.org/ 10.1038/gt.2011.65 90. Chiriaco M, Farinelli G, Capo V et al (2014) Dual-regulated lentiviral vector for gene therapy of X-linked chronic granulomatosis. Mol Ther 22(8):1472–1483. https://doi.org/10. 1038/mt.2014.87 91. Gentner B, Visigalli I, Hiramatsu H et al (2010) Identification of hematopoietic stem cell-specific miRNAs enables gene therapy of globoid cell leukodystrophy. Sci Transl Med 2(58):58ra84. https:// doi.org/10.1126/scitranslmed.3001522 92. Santilli G, Almarza E, Brendel C et al (2011) Biochemical correction of X-CGD by a novel chimeric promoter regulating high levels of transgene expression in myeloid cells. Mol Ther 19(1):122–132. https://doi.org/10.1038/mt.2010.226 93. Kohn DB, Booth C, Kang EM et al (2020) Lentiviral gene therapy for X-linked chronic granulomatous disease. Nat Med 26(2): 200–206. https://doi.org/10.1038/s41591-019-0735-5 94. Weisser M, Demel UM, Stein S et al (2016) Hyperinflammation in patients with chronic granulomatous disease leads to impairment of hematopoietic stem cell functions. J Allergy Clin Immunol 138(1): 219–228 e219. https://doi.org/10.1016/j.jaci.2015.11.028 95. de Luca A, Smeekens SP, Casagrande A et al (2014) IL-1 receptor blockade restores autophagy and reduces inflammation in chronic granulomatous disease in mice and in humans. Proc Natl Acad Sci USA 111(9):3526–3531. https://doi.org/10.1073/pnas. 1322831111 96. Gibbings SL, Haist KC, Nick H et al (2022) Heightened turnover and failed maturation of monocyte-derived macrophages in murine chronic granulomatous disease. Blood 139(11):1707–1721. https:// doi.org/10.1182/blood.2021011798 97. Labrosse R, Abou-Diab J, Blincoe A et al (2017) Very early-onset inflammatory manifestations of X-linked chronic granulomatous disease. Front Immunol 26(6):820–842. https://doi.org/10.3389/ fimmu.2017.01167 98. He HQ, Xu PL, Zhang XF et al (2020) Aging-induced IL27Ra signaling impairs hematopoietic stem cells. Blood 136(2):183–198. https://doi.org/10.1182/blood.2019003910 99. Yang D, de Haan G (2021) Inflammation and aging of hematopoietic stem cells in their niche. Cell 10(8):1849. https:// doi.org/10.3390/cells10081849 100. Bhattacharya S, Marciano B, Malech H et al (2021) Ustekinumab for chronic granulomatous disease-associated inflammatory bowel disease. Gastroenterology 160(3):S80–S81 101. Gabrion A, Hmitou I, Moshous D et al (2017) Mammalian target of rapamycin inhibition counterbalances the inflammatory status of immune cells in patients with chronic granulomatous disease. J Allergy Clin Immun 139(5):1641. https://doi.org/10.1016/j.jaci. 2016.08.033 102. Schott JW, Leon-Rico D, Ferreira CB et al (2019) Enhancing lentiviral and Alpharetroviral transduction of human hematopoietic stem cells for clinical application. Mol Ther Methods Clin Dev 14: 134–147. https://doi.org/10.1016/j.omtm.2019.05.015 103. De Felice L, Tatarelli C, Mascolo MG et al (2005) Histone deacetylase inhibitor valproic acid enhances the cytokine-induced expansion of human hematopoietic stem cells. Cancer Res 65(4): 1505–1513. https://doi.org/10.1158/0008-5472.CAN-04-3063 104. Zimran E, Papa L, Djedaini M, Patel A, Iancu-Rubin C, Hoffman R (2020) Expansion and preservation of the functional activity of adult hematopoietic stem cells cultured ex vivo with a histone

572 deacetylase inhibitor. Stem Cells Transl Med 9(4):531–542. https:// doi.org/10.1002/sctm.19-0199 105. Zonari E, Desantis G, Petrillo C et al (2017) Efficient ex vivo engineering and expansion of highly purified human hematopoietic stem and progenitor cell populations for gene therapy. Stem Cell Rep 8(4):977–990. https://doi.org/10.1016/j.stemcr.2017.02.010 106. Merling RK, Sweeney CL, Chu J et al (2015) An AAVS1-targeted Minigene platform for correction of iPSCs from all five types of chronic granulomatous disease. Mol Ther 23(1):147–157. https:// doi.org/10.1038/mt.2014.195 107. Sweeney CL, Zou JZ, Choi U et al (2017) Targeted repair of CYBB in X-CGD iPSCs requires retention of Intronic sequences for expression and functional correction. Mol Ther 25(2):321–330. https://doi.org/10.1016/j.ymthe.2016.11.012 108. De Ravin SS, Li L, Wu X et al (2017) CRISPR-Cas9 gene repair of hematopoietic stem cells from patients with X-linked chronic granulomatous disease. Sci Transl Med 9(372):eaah3480. https://doi. org/10.1126/scitranslmed.aah3480 109. Sweeney CL, Pavel-Dinu M, Choi U et al (2021) Correction of X-CGD patient HSPCs by targeted CYBB cDNA insertion using CRISPR/Cas9 with 53BP1 inhibition for enhanced homologydirected repair. Gene Ther 28(6):373–390. https://doi.org/10. 1038/s41434-021-00251-z 110. De Ravin SS, Brault J, Meis RJ et al (2021) Enhanced homologydirected repair for highly efficient gene editing in hematopoietic stem/progenitor cells. Blood 137(19):2598–2608. https://doi.org/ 10.1182/blood.2020008503 111. Battersby AC, Braggins H, Pearce MS et al (2017) Inflammatory and autoimmune manifestations in X-linked carriers of chronic granulomatous disease in the United Kingdom. J Allergy Clin

G. Santilli and A. J. Thrasher Immunol 140(2):628–630 e626. https://doi.org/10.1016/j.jaci. 2017.02.029 112. Roos D, van Leeuwen K, Hsu AP et al (2021) Hematologically important mutations: X-linked chronic granulomatous disease (fourth update). Blood Cells Mol Dis 90:102587. https://doi.org/ 10.1016/j.bcmd.2021.102587 113. Koker MY, Camcioglu Y, van Leeuwen K et al (2013) Clinical, functional, and genetic characterization of chronic granulomatous disease in 89 Turkish patients. J Allergy Clin Immunol 132(5): 1156–1163 e1155. https://doi.org/10.1016/j.jaci.2013.05.039 114. Bakri FG, Mollin M, Beaumel S et al (2021) Second report of chronic granulomatous disease in Jordan: clinical and genetic description of 31 patients from 21 different families, including families from Lybia and Iraq. Front Immunol 12:639226. https:// doi.org/10.3389/fimmu.2021.639226 115. Schejtman A, Aragao WC, Clare S et al (2020) Lentiviral gene therapy rescues p47(phox) chronic granulomatous disease and the ability to fight salmonella infection in mice. Gene Ther 27(9): 459–469. https://doi.org/10.1038/s41434-020-0164-6 116. Schejtman A, Vetharoy W, Choi U et al (2021) Preclinical optimization and safety studies of a new lentiviral gene therapy for p47 (phox)-deficient chronic granulomatous disease. Hum Gene Ther 32(17–18):949–958. https://doi.org/10.1089/hum.2020.276 117. Merling RK, Kuhns DB, Sweeney CL et al (2017) Gene-edited pseudogene resurrection corrects p47(phox)-deficient chronic granulomatous disease. Blood Adv 1(4):270–278. https://doi.org/ 10.1182/bloodadvances.2016001214 118. Wrona D, Pastukhov O, Pritchard RS et al (2020) CRISPR-directed therapeutic correction at the NCF1 locus is challenged by frequent incidence of chromosomal deletions. Mol Ther-Methods Clin Dev 17:936–943. https://doi.org/10.1016/j.omtm.2020.04.015

Part VIII Future

Quo Vadis NADPH Oxidases: Perspectives on Clinical Translation

34

Ulla G. Knaus, Ajay M. Shah, and Victor J. Thannickal

Abstract

Since the discovery of several new homologs of the NADPH oxidases (NOXs) just over two decades ago, significant progress has been made in our understanding of the structural and biochemical features of the seven members of the NOX/DUOX family enzymes. Identification of physiological and pathophysiological roles of these reactive oxygen species (ROS) generating enzymes continues to evolve. There is growing recognition that the development of therapeutic strategies to target these enzymes in specific diseases has to account for contextuality of their biological actions, with particular attention to the phase of the disease process—early versus late stage, cellular and subcellular localizations, and confounding variables such as the environment, metagenomics/microbiome, and aging. The authors of this Perspective provide guidance on how pre-clinical studies may be optimized to improve success in clinical translation; these recommendations include validation of NOX/DUOX expression in human tissues/organs, recognition of caveats in the use of animal models, and application of alternative models such as human multi-cellular organoids. Therapeutic opportunities to target

Ulla G. Knaus, Ajay M. Shah and Victor J. Thannickal contributed equally to writing this chapter. U. G. Knaus (✉) Conway Institute, School of Medicine, University College Dublin, Dublin, Ireland e-mail: [email protected] A. M. Shah King’s College London British Heart Foundation Centre, School of Cardiovascular & Metabolic Medicine and Sciences, London, UK e-mail: [email protected] V. J. Thannickal John W. Deming Department of Medicine, Tulane University School of Medicine, New Orleans, LA, USA e-mail: [email protected]

NOX/DUOX enzymes in immunity, organ fibrosis, and cardiovascular disease are also discussed. Keywords

NADPH oxidase · NOX · DUOX · Immunity · Inflammation · Fibrosis · Cardiovascular disease · Aging · Animal models · Therapeutic target

1

Introduction

In the last decades NADPH oxidase research has made great strides, from the discovery and characterization of several NOX/DUOX isoforms in animals, plants, and microorganisms to an improved understanding of structural features and biochemical processes, the development of various genetically modified mouse strains for phenotypic investigation in unchallenged and challenged conditions, and collective efforts in setting up compound screening assays for hit discovery of NOX inhibitors. Considerable progress has been made, even though difficulties associated with a dearth of reliable research tools has hindered swifter progress, impacting at times on reproducibility. To move forward in NOX research, it is important not only to take stock of accomplishments, but also to integrate the promise of forthcoming fundamental discoveries and technologies into the overarching aim of conducting research to ultimately address unmet medical needs. It is time to ask questions concerning unresolved knowledge gaps, the reliability and meaningfulness of current methodologies, and the promise of NOX enzymes as druggable targets. Even if compounds selectively targeting individual NADPH oxidase isoforms can be identified, are NOX/DUOX enzymes promising therapeutic targets, and if so which ones? Do we need to develop and test NADPH oxidase activators as well as inhibitors, and in which disorders will these interventions likely show the most promise? The following reflections are not exhaustive and are

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2_34

575

576

U. G. Knaus et al.

necessarily subjective. They are modeled on recent translational research recommendations for drug target assessment [1] and can serve as guideposts for advancing our knowledge on oxidases to fully fulfill their promise as drug targets for next generation therapeutics.

2

Structure, Expression and Activity of NADPH Oxidases

NADPH oxidase enzymes are multimeric transmembrane complexes with up to six essential components that require numerous co-factors, protein-protein interactions, and posttranslational modifications for catalytic activity. The minimal components indispensable for enzyme activity have been identified for all seven human NOX/DUOX enzymes, but it is likely that not all endogenous activating or inhibiting modifiers are yet known. This might be especially the case for NOX4, an NADPH oxidase mainly considered transcriptionally regulated and constitutively active [2]. Structural insights in the cytosolic components of the NOX2 NADPH oxidase were achieved in early 2000 [3], but only recently has progress been made on the three-dimensional structure of the membrane-bound catalytic core ([4–6], see Part VI Structure). One can predict that further progress will be made on obtaining structural information on the resting and activated forms of several NOX/DUOX enzymes, especially with advances in cryogenic electron microscopy, and these structural data will aid in the de novo design and optimization of pharmaceutical compounds with inhibitory or activating features. Evaluation of NOX/DUOX expression profiles in human cells and tissues in homeostasis, and in patients during the development, progression and resolution of disease is crucial for target-disease linkage. On this front progress has been achieved when cells and tissues were readily accessible, while in other circumstances murine expression patterns served as an imperfect substitute. Ideally, enzyme expression should be assessed at various time points, and for migrating cell populations at different sites (e.g., blood, affected tissue), as the disease condition (e.g., acute, chronic, resolving inflammation) and the changing microenvironment frequently alter expression levels. Such insights will be more easily gained from murine disease models. For certain NADPH oxidases isoform (NOX2, NOX5, DUOX1, DUOX2) specific and sensitive antibodies, which are often not commercially available, confirmed protein expression and provided information on intracellular localization, while for other isoforms this important tool is still lacking or imperfect (NOX1, NOX3, NOX4) [7, 8]. The preparation of improved monoclonal antibodies recognizing human and murine oxidase isoforms, rigorous validation and widespread

distribution to research laboratories is critical, but this will likely require academic-industrial cooperation or funding agency-led initiatives. One needs to bear in mind that oxidase expression levels can be modifier dependent, either in the cellular context (e.g., impaired dimerization, redox regulation, epigenetic silencing) [9–11] or due to the microenvironment (e.g., oxygen concentration, cytokines, microbial sensing) [12–16]. In short supply are antibodies detecting the assembly of oxidase complexes or their activation, or the (induced) localization of oxidases at the cell surface [17–20]. Studies on protein kinases and receptors have shown how enabling such research tools can be. The upregulation of NOX/DUOX expression cannot be used as evidence of catalytic activity and rising superoxide/ hydrogen peroxide levels. Various methods for quantifying reactive oxygen species (ROS) have been developed [21–23] and reagents/kits are commercially available. Highthroughput screening of compounds altering NOX/DUOX activity is technically feasible and can successfully identify lead compounds if several supporting assays and robust counter screens are used [7, 24–26]. Structural information may soon shed light on how NOX4 and DUOX1/2, in contrast to other NOXs, release H2O2 and not superoxide. This ability has likely unique consequences on redox signaling and concomitant cellular functions, yet this area is not well studied. NADPH oxidases reduce molecular oxygen to superoxide via electron transfer from NADPH to FAD and the two hemes incorporated into the third and fifth NOX transmembrane domains but the exact mechanism of electron transfer needs further study which will be facilitated by detailed elucidation of NOX structures (e.g., [27, 28]). The termination of oxidase activity in physiological conditions is unresolved, although understanding this process better may reveal new strategies for designing NOX inhibitors. Currently, cytoskeletal rearrangements, accelerated Rac2 GTP hydrolysis, dephosphorylation, membrane depolarization, cytosolic acidification and cell death have been proposed as termination mechanisms [29–31]. These factors were deduced from the analysis of NOX2 activity in neutrophils or cell lines, commonly stimulated in vitro with phorbol ester, and thereby may not represent physiological mechanisms and are not translatable to other NOX enzymes. In differentiated bronchial epithelial cells the termination of DUOX-induced H2O2 release was independent of a decrease in the intracellular calcium or NADPH concentrations, and oxidase activity could be restimulated by a second, distinct stimulus [32]. Thus, how transient superoxide/H2O2 production by oxidases is achieved still awaits discovery.

34

3

Quo Vadis NADPH Oxidases: Perspectives on Clinical Translation

The Complexity of NOXs as Therapeutic Targets

Single gene disorders can in principle be effectively treated by targeting or replacing the abnormal gene or gene product. In the NOX field, hematopoietic stem/progenitor cell-based lentiviral gene therapy for X-linked chronic granulomatous disease (CGD) is a good example of this approach and appears increasingly promising ([33], see Chap. 33 by G. Santilli and A.J. Thrasher). However, most human diseases—and especially chronic conditions—are complex disorders that involve multiple interacting pathophysiological processes modified by diverse genetic and environmental factors. The selection of optimal therapeutic targets in such conditions is therefore greatly aided by deciphering the distinct role of the putative target within the complex underlying biological processes that drive the disease. It is important to establish whether the activation of a target is a cause or a consequence of such pathological processes and whether activation enhances disease progression or is an adaptive mechanism. Since there is rarely a single key causative molecule or pathway in most complex disorders, it is also important to assess whether enzymes such as NOXs are best targeted directly, at the level of upstream activation or at downstream action. Finally, the relevance of information obtained from rodent or cellular models to in vivo human disease is another important consideration. NOXs have by now been suggested to be involved in a very large number of diseases beyond CGD, including cardiovascular, respiratory, hepatic, neurological, gastrointestinal, cancer and other disorders. However, the above requirements for translation of research discoveries to therapeutic interventions are largely lacking for most of these conditions.

3.1

Intracellular Signaling Related to NOXs

The seminal studies on NOX2, driven by recognition of its pivotal role in CGD, strongly focused on its structure, biochemical function and activity in neutrophils—where rapid, high concentration superoxide production and associated changes in proton fluxes within phagocytic vacuoles are key to its function. It is clear however that the function of NOX2 in other cell types (including other immune cells) is different. NOX2-mediated compartmentalized signaling involving its co-localization with redox-sensitive target proteins appears to be an especially important facet of its role in many non-phagocytic cells [34–38]. Such compartmentalized redox signaling also appears to be central

577

to the roles of NOX1 and NOX4 in many settings [39– 42]. Furthermore, it is abundantly clear that different NOX/DUOX isoforms may mediate distinct signaling events in the same cell type [39, 43, 44]. Yet far too many publications still focus predominantly on non-specific detrimental effects related to “oxidative stress”, without consideration of such compartmentalized subcellular signaling and how it interacts with other pathogenic pathways. An analysis of NOX-mediated cellular signaling from a potential therapeutic perspective also requires consideration of the inter-relationship and interactions with antioxidant pathways [45, 46], which may be important in guiding the choice of the therapeutic target.

4

NADPH Oxidase Animal Models

Global and conditional knock-out, knock-in, and loss-offunction mutant mouse and rat strains for Nox1-5 and Duox1-2 and associated complex components are commercially available (see International Mouse Strain Resource IMSR) or can be sourced from academic researchers. Additional strains are likely in development and resource sharing for robust disease model development and enhanced reproducibility should be encouraged. Global deletion or inactivation of NOX/DUOX in rodents recapitulates for two NADPH oxidase isoforms the key pathophysiology observed in the respective genetic disorders in humans. The innate immune defense deficiencies typically observed in CGD due to inherited loss-of-function variants in NOX2 complex components (CYBB, CYBA, NCF1, NCF2, NCF4) are mirrored in pathogen challenged knock-out or mutant rodent models ([9, 47–52], see Chap. 32 by M-J. Stasia and D. Roos). Similarly, congenital hypothyroidism caused by inactivating DUOX2 and DUOXA2 variants is reproduced in mice harboring a loss-of-function Duox2 mutation and in mice with global Duoxa1/2 deficiency [53, 54]. Inactivation or deletion of the Nox3 complex results in vestibular defects in the inner ear [55], but a link to human disease is not yet firmly established. Environmental changes (e.g., health status, nutrition) or a prolonged observation period (e.g., aging) may reveal additional phenotypes in unchallenged Nox/Duox deficient animals. For various disease challenges, constant or time-dependent global and tissue-specific Nox/Duox deletion strains have been used. To date, genetic modification of animals to underscore the role of a particular oxidase in disease is only available in rodents.

578

4.1

U. G. Knaus et al.

Limitations of Animal Models

Much still needs to be learnt before the knowledge so far acquired on the NOX family enzymes can be translated to effective new therapies for human diseases. Most studies exploring biological functions of NOX/DUOX enzymes have been conducted in in-bred mouse (Mus musculus) strains. Unfortunately, mouse models often fail to recapitulate human pathophysiology or disease. There may be many reasons for this, and it should not be simply ascribed to “species differences”. Indeed, basic cellular and molecular mechanisms are well conserved across mammalian organisms, including Mus musculus. An important caveat, however, are the differences in organ-specific anatomy and physiology that may influence the phenotypic expression of even monogenic disorders [56]. More pertinent reasons why animal models do not replicate human disease or fail to more accurately predict the effectiveness of therapeutic agents include: (a) the complexity of human disease; (b) the dynamic nature of biological processes and contextual action of genes/mediators, and (c) failure to account for the age of the animal in disease models.

4.1.1 Disease Complexity Biological processes in mammalian organisms are complex and adding the variable of “time” (chronicity) to these processes further compounds complexity. Chronic diseases account for a large majority of unmet medical needs of modern societies. These diseases involve multiple genetic (polygenic) and environmental factors, with a dizzying array of gene-gene and gene-environmental interactions. Genome-wide association studies (GWAS) in diseases such as diabetes, atherosclerosis, Alzheimer’s disease, and chronic respiratory disorders typically reveal a number of candidate genes with small effect sizes. Modeling such diseases in model organisms is, therefore, challenging. Gene deletions in inbred strains of mice may produce a phenotype, either in the unstressed state or in stressed conditions. However, these “phenotypes” do not always model human diseases. Rather, they inform gene function within the context of organismal development (in the unstressed state) and/or a specific stressor or challenge in a fully developed, adult animal. Thus, it is not that such models should be abandoned, but rather that their limitations be recognized and acknowledged, and experimental results interpreted with more nuanced understanding of complex biologic processes. 4.1.2 Biological Dynamicity and Contextuality Functional diversity of genes and gene products may be particularly advantageous to organisms subject to rapid changes in their environment [57]. Most genes encode for proteins with multiple functions; for example, an enzyme may possess both catalytic and non-catalytic activities. This

is well recognized for protein kinases that function as scaffolds, allosteric regulators, and mediate protein-DNA binding [58]. Our knowledge of such non-catalytic functions of NOX enzymes are nascent. The catalytic function of NOX enzymes is to generate ROS which can then function in cellular signaling, biosynthetic reactions, or in host defense mechanisms; thus, the reactive nature of ROS itself contributes to functional diversity and the multiplicity of physiological roles for NOX enzymes. An understanding of this functional diversity is particularly relevant when interpreting studies employing animal models of disease. Most animal models require the external application of (or exposure to) a “stressor” designed to produce a pathology akin to the human condition. This “pathology” is often an extension of an adaptive physiological response in the host (usually in mice) subjected to a stressor agent/intervention. However, as with all adaptive responses in mammals, the biology is not only complex, but dynamic and contextual. This dynamicity is facilitated by the multifunctionality of stress-responsive genes. An example of this multifunctionality is the transcription factor nuclear factor kappa B (NF-κB), which is well recognized as a canonical pro-inflammatory mediator; however, inhibition of NF-κB during the resolution phase of inflammation, paradoxically, protracts the inflammatory response by preventing apoptosis of recruited neutrophils [59]. Another example is the recognized role of transforming growth factor-beta1 (TGF-β1) as both a tumor suppressor and tumor promoter, depending on the phase of the disease process [60]. NOX4 is one of the genes that is most highly induced by TGF-β1 and mediates many of the known actions of this multifunctional cytokine [45, 61]. Due to this inherent contextuality and multifunctionality, gene knock-outs in the germline may mis-inform due to the contextual and phase-specific actions of that gene. It is possible that paradoxical effects observed in some knock-out animal models may be attributable to this limitation. Conditional gene knock-out models that allow for gene deletion at specific time points after injury/stress may be more informative, and perhaps more predictive as pre-clinical models. In such cases, care must be taken that pharmacologic agents used for induction do not independently modulate the disease process and confound the resulting phenotype. Furthermore, it must be acknowledged that, unlike genetic targeting of a cell-specific promoter, it is more challenging to deliver pharmacologic agents to specific cell types in a human disease context.

4.1.3 Aging and Antagonistic Pleiotropy In addition to genes and environment, the biology of aging can affect the host response to injury or infection. For example, fibrosis resolution capacity in the lung is markedly diminished with aging [45], and this may explain the

34

Quo Vadis NADPH Oxidases: Perspectives on Clinical Translation

propensity to recalcitrant pulmonary fibrosis in older individuals [62]. Unfortunately, most experimental animal models involve mice that are relatively young; for example, bleomycin lung injury is induced in mice that are typically 6–8 weeks of age, which translates to humans in their teenage years [63]. This may obscure the actions of some genes that mediate beneficial effects at young age, but detrimental effects at post-reproductive age—a concept referred to as antagonistic pleiotropy [64]. It has been argued that NOX enzymes may function as antagonistic pleiotropic genes [65], and this suggests that NOXs may be particularly attractive targets in age-related diseases [66]. Taken together, these limitations indicate that studies in animal models need to progress to a deeper, more mechanistic and more nuanced analysis of the roles of NOXs within complex disease biology.

4.2

Strategies to de-Risk Evaluation of NOX/DUOX as a Drug Target

Given the current limitations of animal models, we would recommend considering the following when designing pre-clinical testing of inhibitors/activators of NOX enzymes: 1. Confirm that the gene-of-interest is expressed in the cellof-interest in tissues derived from the human disease-ofinterest, and that the direction of change in gene/protein expression supports the proposed role of the implicated NOX enzyme(s). 2. Assess whether any genetic evidence from human populations is consistent with the proposed role of the NOX enzyme. 3. Validate drug candidates in human organoid models where available. 4. Consider at least two different, and possibly complementary, animal models when testing the safety and efficacy of pre-clinical drug candidates. 5. Consider non-human primate models where feasible or naturally occurring (“sporadic”) models in diverse mammalian species. 6. Gene and/or pharmacologic targeting strategies must account for the dynamic nature of biological processes with gene deletion and/or drug administration, respectively, appropriately timed to the pathological processof-interest. 7. Consider the age of animals when age has been shown to influence the development and/or progression of the human disease. 8. Incorporate metagenomics (e.g., the microbiome) and environmental factors in the experimental model/design to enhance reproducibility.

579

We believe that establishing the scientific premise by approaches to demonstrate human relevance, such as confirming gene expression of the NOX enzyme-of-interest in human tissues/cells, is critical both as a de-risking strategy and to elevate confidence in successful clinical translation. Fortunately, with the burgeoning publicly available transcriptomic and proteomic databases, including singlecell sequencing of diverse human diseases, this may be more readily achievable today. Such a bedside-to-bench approach may ultimately de-risk translatability from benchto-bedside.

5

NADPH Oxidase-Disease Linkage

For a promising drug target, a causal relationship and confirmed role in the pathophysiology in model organisms and in patients is highly desirable, although targets acting as disease modifiers are also very valuable. In much of the literature on NOXs, it is generally considered that a NOX isoform requires inhibition because of excessive oxidative stress triggered by NOX(s) and/or detrimental NOX-mediated signaling. However, an increasing body of literature also describes beneficial roles for some NOX-modulated pathways, both physiologically and in disease settings. Three therapeutic areas will be discussed here in the context of NADPH oxidases as drug targets for developing inhibitory or activating compounds.

5.1

NADPH Oxidases in Immunity

NADPH oxidases are critical immunoregulators in mucosal, innate, and adaptive immunity, acting as signal transducers in multiple pathways and fulfilling specialized functions in host defense. Epithelial barrier tissues colonized with microbiota constitute the entry point for many pathogenic microorganisms and express multiple oxidases (e.g., NOX1, DUOX1, DUOX2) as a protective antivirulence mechanism, to strengthen and repair the barrier, and to transmit signals for chemokine and cytokine production [67]. Innate immune cells provide most of the acute host defense against invading pathogens, with NOX2 being responsible for the microbicidal oxidative burst of neutrophils and macrophages. Changes in the microenvironment will reprogram macrophages to distinct functional phenotypes which often includes de novo NOX/DUOX isoform expression [68]. NOX2 is also required for dendritic cells to initiate T cell immunity [69]. NADPH oxidase-mediated signaling in activation and function of T lymphocyte subsets remains incompletely defined. The manifestation of life-threatening infections in CGD as well as the survival strategies employed by pathogens to subvert NOX2 complex assembly highlight the importance of oxidant generation in antimicrobial defense. Rodent

580

animal models reflect ROS as an essential host defense mechanism quite well in respiratory and systemic infections, but they are less suitable for oral infection with intestinal bacteria. Host factors and the microbiota often prevent persistent colonization of the rodent intestine by human pathogens, thereby limiting disease progression [70]. Although antibiotic pre-treatment can overcome such limitations, it removes the fundamental host-microbiota-pathogen link in disease onset, progression and resolution. Substitution with murine pathogens is feasible when convergent evolution strains exist. In other circumstances, an essential attachment or entry receptor for a human pathogen is missing in rodents. In this case, humanized mice or other mammals such as ferrets or hamsters will provide a more suitable model. NOX2 constitutes the prime example for the Yin and Yang of ROS in disease. In acute infections or in CGD patients, boosting phagocyte ROS production will be beneficial or even essential. Current preclinical strategies include enhancing neutrophil ROS with biased agonists to the formyl peptide receptor upstream of NOX2 in acute disease, or stimulating mitochondrial ROS to compensate for loss-offunction CGD variants [68]. Advancing therapies that target the hyperinflammation and autoimmune conditions (e.g., granulomatous colitis, rheumatoid arthritis, systemic lupus erythematosus) associated with CGD, currently treated with immunosuppressants, is challenging [71, 72] and may entail gene targeting strategies [33]. On the other hand, activated neutrophils and NOX2-generated superoxide contribute to the oxidative damage in inflammatory diseases (e.g., pancreatitis, pulmonary inflammation, ischemia-reperfusion injury), supporting the development of NOX2 inhibitors [73]. One such inhibitor, an NADPH competitive, NOX2 selective compound, showed activity in an animal model of acute pancreatitis [24]. Therapeutic use of NOX2 inhibitors needs to consider the drug dosage and treatment duration to safeguard from infections and autoimmune complications. Fortifying host protection at mucosal barriers also necessitates rather an oxidant enhancing than inhibitory approach. Various strategies can be considered to augment H2O2 at epithelial barriers [68]. Developing compounds upregulating or activating epithelial oxidases such as DUOX could be a viable tactic as these enzymes provide anti-infective control in the respiratory tract and intestine [15, 74–76]. Inactivating DUOX2 and NOX1 variants are considered inflammatory bowel disease (IBD) risk factors [67, 77, 78], likely due to their role in redox signaling, microbial sensing and antivirulence. This suggests that compensatory ROS induction could be beneficial, at least in certain subsets of patients, as comprehensive understanding of IBD risk loci including NADPH oxidases [79] awaits improved animal models that reflect the complexity of human intestinal inflammation. Moreover, compounds

U. G. Knaus et al.

reversing the epigenetic silencing of DUOX1/2 might be beneficial in certain cancers if these enzymes are indeed tumor suppressors [11, 80, 81], while in other tumors DUOX1 was linked to chromosomal instability and enhanced carcinogenesis ([82]; see Chap. 14 by F. Miot and X. De Deken). The association of DUOX enzymes with tumorigenesis will require further study. In contrast, smallmolecule inhibitors targeting selectively DUOX1 may be beneficial in allergic airway diseases [83]. In summary, the link between the NADPH oxidase NOX2, and likely mucosa associated NOX1/DUOX2, and antimicrobial host defense is strong and at this point supports mainly compounds increasing ROS levels.

5.2

NADPH Oxidases in Organ Fibrosis

Fibrosis involving diverse tissues/organs has received increased attention in recent years as a disease process in which NOX-targeted therapeutics are being actively developed. Fibrosis can be viewed as a stereotypical tissue reaction to injury and is likely evolutionarily conserved to facilitate wound healing and to limit pathogen spread [84, 85]. Most animal models of fibrosis involve an injurious agent or intervention which typically involves an early inflammatory phase followed by a more delayed, late fibrotic phase. Studies using different animal models across distinct organ systems have given mixed, and sometimes contradictory, results with regard to potential anti-fibrotic effects of NOXs [86, 87]. While NOX4 inhibition has consistently revealed anti-fibrotic effects in the lung [61, 88, 89] and liver [90–92], such approaches have produced variable results in the heart [93–95] and kidney [96–98]. The potential explanations for such differences could be attributed to tissue/organ specific functions of NOX4, the disease model employed, the phase of the disease, or even different types of fibrosis that may develop in response to stress/injury. For example, certain types of fibrosis may be predominantly driven by inflammation, independent of TGF-β or NOX4 [99]. A major difference between the currently available animal models and the more recalcitrant fibroses in humans is the reversibility of the disease process in the former. While this might appear as a challenge, one might consider it as an opportunity to study mechanisms of fibrosis resolution in relevant animal models [100]. Indeed, from a therapeutic perspective, it may be more practical to develop strategies that target fibrosis resolution (in contrast to prevention) since most patients with chronic, indolent disease processes typically have established fibrosis at clinical presentation. Based on the premise that pathological fibrosis may be attributable to a deficiency in the capacity for fibrosis resolution, therapeutic interventions at improving or restoring resolution

34

Quo Vadis NADPH Oxidases: Perspectives on Clinical Translation

capacity during the late reparative phases of tissue injury may be worth considering [89, 101–104]. Additionally, studies on how the induction of stress-induced pathways such as those involving NOXs are “switched off” may be useful in promoting stem cell behavior/function and tissue regenerative capacity in chronic, age-related fibrotic disorders that involve diverse organ systems [105, 106].

5.3

NADPH Oxidases in Cardiovascular Disease

A large body of literature implicates NADPH oxidases in cardiovascular physiology and pathology. NOX1 has mainly been linked to vascular disease whereas NOX2 and NOX4 play roles in both cardiac and vascular pathophysiology. Less is known about NOX5, which is not expressed in rodents, but is suggested to be involved in human vascular disease. Conditions where one or more of these NOXs are implicated include hypertension, atherosclerosis, aortic aneurysms, postischemic neovascularization, cardiac hypertrophy, cardiac arrhythmia, and heart failure. As a generalization, the excessive activation of NOX1 or NOX2 promote the development of several of these pathologies, with these two isoforms having distinct roles. In the case of NOX4, however, the situation is more complex. Adaptive (protective) roles for NOX4 have been identified in diverse disease/stress settings including cardiac hypertrophy, atherosclerosis, and neovascularization [41, 42, 44, 107–111], although some studies described detrimental roles [95, 112]. The latter might relate to the use of non-physiologic overexpression approaches or experimental model-dependent uncoupling between NOX4 and downstream adaptive pathways (such as those regulating antioxidant balance). The consideration of NOX1 or NOX2 as a potential therapeutic target for inhibition first requires an assessment of the specific role of the oxidase as distinct from other pathophysiological mechanisms that are the targets of existing therapies. For example, it is well established that increased activation of the renin-angiotensin-aldosterone system (RAAS) and the sympathetic pathway have a central role in hypertension and that beta-blockers and RAAS inhibitors are accordingly very effective drugs. The question then is whether additional NOX inhibition is likely to be beneficial. Moreover, RAAS inhibitors and statins—which are prescribed to many patients with cardiovascular disease—may themselves indirectly reduce NOX1 and NOX2 activity. RAAS inhibitors reduce angiotensin II-mediated activation of NOX1/NOX2, a common activation mechanism in cardiovascular disorders, while statins impact on oxidase activity through the inhibition of Rac prenylation. The potential for additional NOX inhibition may therefore be limited. It will be important to clearly identify the cardiovascular conditions and disease stages

581

(e.g., early versus late) where NOX inhibition may be of value over and above current evidence-based therapies for such conditions. In addressing this question, it will also be important to take into account the interactions with other pathways (e.g., antioxidant and metabolic pathways) and the role of variables such as age and co-morbidity. Another issue is whether NOX-dependent ROS production or a downstream consequence is the best therapeutic target. ROS generated by NOXs can trigger further ROS production from mitochondria or sources such as dysfunctional nitric oxide synthase that collectively contribute to oxidative stress. In a setting such as advanced heart failure, where ROS generation related to mitochondrial dysfunction is central to myocardial pathophysiology, it may be more effective to specifically target mitochondrial ROS (which is achievable using mitochondrial-targeted antioxidants) rather than inhibiting a NOX isoform [113]. However, in early cardiac remodeling in response to hemodynamic overload or myocardial infarction, the inhibition of NOX2 to reduce detrimental intracellular signaling (rather than oxidative stress per se) may be more effective. Current evidence suggests that it may be desirable to enhance rather than inhibit NOX4 in conditions where it has been shown to stimulate adaptive stress pathways, such as in the early stages of heart remodeling in response to hemodynamic overload. In this case, however, the extensive tissue distribution of NOX4 is an important consideration because increasing NOX4 activity may be detrimental in settings such as cancer. We believe that a promising approach to address this problem and achieve functional organ- or tissue-specific targeting may be to choose a therapeutic target downstream of NOX4, which is only activated in the diseased tissue. Despite the large number of studies on NOXs and cardiovascular disease in rodent models to date, it is clear that significantly more work is required before potential translation to new therapeutic targets can be considered.

6

Conclusions

Following the discovery that NADPH oxidases are an evolutionarily conserved family of enzymes with extensive cellular and tissue distribution, a huge amount has been learnt about their basic structure and biochemical function, and their potential involvement in mammalian physiology and disease. However, the prospect of translating this knowledge into new therapeutic approaches for human disease is still a considerable way short of fruition. Here, we have outlined some of the main areas that we believe need to be the focus of research aimed at developing new therapies based on NOX/DUOXs. We also identify common pitfalls in current approaches, many of which may be over-simplistic and insufficiently nuanced, and we highlight the importance of obtaining data

582

that is directly relevant to humans and human disease. With appropriate rigor, focus and attention to complexity, we nevertheless do believe that the field is ripe for translation to the clinic. Acknowledgements The authors gratefully acknowledge the contributions of the members of their research groups. UGK is supported by Science Foundation Ireland (16/IA/4501) and Children’s Health Ireland/National Children’s Research Center (K/17/2). VJT is supported by U.S. National Institutes of Health grants, P01 HL114470, R01 HL139617, R01 HL151702, and R01 HL152246; and by the U.S. Department of Veterans Affairs Merit Award, I01BX003056. AMS is supported by the British Heart Foundation (CH/1999001/11735, RE/18/ 2/34213), the National Institute for Health Research Biomedical Research Centre at Guy’s & St Thomas’ NHS Foundation Trust and King’s College London (IS-BRC-1215-20006), and a Fondation Leducq Transatlantic Network of Excellence award (17CVD04).

References 1. Emmerich CH, Gamboa LM, Hofmann MCJ, Bonin-Andresen M, Arbach O, Schendel P, Gerlach B, Hempel K, Bespalov A, Dirnagl U, Parnham MJ (2021) Improving target assessment in biomedical research: the GOT-IT recommendations. Nat Rev Drug Discov 20(1):64–81. https://doi.org/10.1038/s41573-020-0087-3 2. Martyn KD, Frederick LM, von Loehneysen K, Dinauer MC, Knaus UG (2006) Functional analysis of Nox4 reveals unique characteristics compared to other NADPH oxidases. Cell Signal 18(1):69–82. https://doi.org/10.1016/j.cellsig.2005.03.023 3. Groemping Y, Rittinger K (2005) Activation and assembly of the NADPH oxidase: a structural perspective. Biochem J 386(Pt 3): 401–416. https://doi.org/10.1042/BJ20041835 4. Magnani F, Nenci S, Millana Fananas E, Ceccon M, Romero E, Fraaije MW, Mattevi A (2017) Crystal structures and atomic model of NADPH oxidase. Proc Natl Acad Sci USA 114(26):6764–6769. https://doi.org/10.1073/pnas.1702293114 5. Sun J (2020) Structures of mouse DUOX1-DUOXA1 provide mechanistic insights into enzyme activation and regulation. Nat Struct Mol Biol 27(11):1086–1093. https://doi.org/10.1038/ s41594-020-0501-x 6. Wu JX, Liu R, Song K, Chen L (2021) Structures of human dual oxidase 1 complex in low-calcium and high-calcium states. Nat Commun 12(1):155. https://doi.org/10.1038/s41467-020-20466-9 7. Augsburger F, Filippova A, Rasti D, Seredenina T, Lam M, Maghzal G, Mahiout Z, Jansen-Durr P, Knaus UG, Doroshow J, Stocker R, Krause KH, Jaquet V (2019) Pharmacological characterization of the seven human NOX isoforms and their inhibitors. Redox Biol 26:101272. https://doi.org/10.1016/j.redox.2019. 101272 8. Diebold BA, Wilder SG, De Deken X, Meitzler JL, Doroshow JH, McCoy JW, Zhu Y, Lambeth JD (2019) Guidelines for the detection of NADPH oxidases by immunoblot and RT-qPCR. Methods Mol Biol 1982:191–229. https://doi.org/10.1007/978-1-49399424-3_12 9. Pircalabioru G, Aviello G, Kubica M, Zhdanov A, Paclet MH, Brennan L, Hertzberger R, Papkovsky D, Bourke B, Knaus UG (2016) Defensive mutualism rescues NADPH oxidase inactivation in gut infection. Cell Host Microbe 19(5):651–663. https://doi.org/ 10.1016/j.chom.2016.04.007 10. Yu L, Zhen L, Dinauer MC (1997) Biosynthesis of the phagocyte NADPH oxidase cytochrome b558. Role of heme incorporation

U. G. Knaus et al. and heterodimer formation in maturation and stability of gp91phox and p22phox subunits. J Biol Chem 272(43):27288–27294 11. Luxen S, Belinsky SA, Knaus UG (2008) Silencing of DUOX NADPH oxidases by promoter hypermethylation in lung cancer. Cancer Res 68(4):1037–1045. https://doi.org/10.1158/0008-5472. CAN-07-5782 12. Goyal P, Weissmann N, Grimminger F, Hegel C, Bader L, Rose F, Fink L, Ghofrani HA, Schermuly RT, Schmidt HH, Seeger W, Hanze J (2004) Upregulation of NAD(P)H oxidase 1 in hypoxia activates hypoxia-inducible factor 1 via increase in reactive oxygen species. Free Radic Biol Med 36(10):1279–1288. https://doi.org/ 10.1016/j.freeradbiomed.2004.02.071 13. Yuan G, Khan SA, Luo W, Nanduri J, Semenza GL, Prabhakar NR (2011) Hypoxia-inducible factor 1 mediates increased expression of NADPH oxidase-2 in response to intermittent hypoxia. J Cell Physiol 226(11):2925–2933. https://doi.org/10.1002/jcp.22640 14. Grasberger H, Gao J, Nagao-Kitamoto H, Kitamoto S, Zhang M, Kamada N, Eaton KA, El-Zaatari M, Shreiner AB, Merchant JL, Owyang C, Kao JY (2015) Increased expression of DUOX2 is an epithelial response to mucosal Dysbiosis required for immune homeostasis in mouse intestine. Gastroenterology 149(7): 1849–1859. https://doi.org/10.1053/j.gastro.2015.07.062 15. Strengert M, Jennings R, Davanture S, Hayes P, Gabriel G, Knaus UG (2014) Mucosal reactive oxygen species are required for antiviral response: role of Duox in influenza a virus infection. Antioxid Redox Signal 20(17):2695–2709. https://doi.org/10. 1089/ars.2013.5353 16. Harper RW, Xu C, Eiserich JP, Chen Y, Kao CY, Thai P, Setiadi H, Wu R (2005) Differential regulation of dual NADPH oxidases/peroxidases, Duox1 and Duox2, by Th1 and Th2 cytokines in respiratory tract epithelium. FEBS Lett 579(21): 4911–4917. https://doi.org/10.1016/j.febslet.2005.08.002 17. Makni-Maalej K, Chiandotto M, Hurtado-Nedelec M, Bedouhene S, Gougerot-Pocidalo MA, Dang PM, El-Benna J (2013) Zymosan induces NADPH oxidase activation in human neutrophils by inducing the phosphorylation of p47phox and the activation of Rac2: involvement of protein tyrosine kinases, PI3Kinase, PKC, ERK1/2 and p38MAPkinase. Biochem Pharmacol 85(1):92–100. https://doi.org/10.1016/j.bcp.2012. 10.010 18. Nakamura M, Murakami M, Koga T, Tanaka Y, Minakami S (1987) Monoclonal antibody 7D5 raised to cytochrome b558 of human neutrophils: immunocytochemical detection of the antigen in peripheral phagocytes of normal subjects, patients with chronic granulomatous disease, and their carrier mothers. Blood 69(5): 1404–1408 19. von Lohneysen K, Noack D, Wood MR, Friedman JS, Knaus UG (2010) Structural insights into Nox4 and Nox2: motifs involved in function and cellular localization. Mol Cell Biol 30(4):961–975. https://doi.org/10.1128/MCB.01393-09 20. Meitzler JL, Makhlouf HR, Antony S, Wu Y, Butcher D, Jiang G, Juhasz A, Lu J, Dahan I, Jansen-Durr P, Pircher H, Shah AM, Roy K, Doroshow JH (2017) Decoding NADPH oxidase 4 expression in human tumors. Redox Biol 13:182–195. https://doi.org/10. 1016/j.redox.2017.05.016 21. Zielonka J, Hardy M, Michalski R, Sikora A, Zielonka M, Cheng G, Ouari O, Podsiadly R, Kalyanaraman B (2017) Recent developments in the probes and assays for measurement of the activity of NADPH oxidases. Cell Biochem Biophys 75(3–4): 335–349. https://doi.org/10.1007/s12013-017-0813-6 22. Hardy M, Zielonka J, Karoui H, Sikora A, Michalski R, Podsiadly R, Lopez M, Vasquez-Vivar J, Kalyanaraman B, Ouari O (2018) Detection and characterization of reactive oxygen and nitrogen species in biological systems by monitoring speciesspecific products. Antioxid Redox Signal 28(15):1416–1432. https://doi.org/10.1089/ars.2017.7398

34

Quo Vadis NADPH Oxidases: Perspectives on Clinical Translation

23. Smolyarova DD, Podgorny OV, Bilan DS, Belousov VV (2021) A guide to genetically encoded tools for the study of H2O2. FEBS J 289:5382–5395. https://doi.org/10.1111/febs.16088 24. Hirano K, Chen WS, Chueng AL, Dunne AA, Seredenina T, Filippova A, Ramachandran S, Bridges A, Chaudry L, Pettman G, Allan C, Duncan S, Lee KC, Lim J, Ma MT, Ong AB, Ye NY, Nasir S, Mulyanidewi S, Aw CC, Oon PP, Liao S, Li D, Johns DG, Miller ND, Davies CH, Browne ER, Matsuoka Y, Chen DW, Jaquet V, Rutter AR (2015) Discovery of GSK2795039, a novel small molecule NADPH oxidase 2 inhibitor. Antioxid Redox Signal 23(5):358–374. https://doi.org/10.1089/ars. 2014.6202 25. O'Neill S, Knaus UG (2019) Protein-protein interaction assay to analyze NOX4/p22(phox) Heterodimerization. Methods Mol Biol 1982:447–458. https://doi.org/10.1007/978-1-4939-9424-3_26 26. Zielonka J, Zielonka M, Cheng G, Hardy M, Kalyanaraman B (2019) High-throughput screening of NOX inhibitors. Methods Mol Biol 1982:429–446. https://doi.org/10.1007/978-1-49399424-3_25 27. Cross AR, Jones OT (1991) Enzymic mechanisms of superoxide production. Biochim Biophys Acta 1057(3):281–298 28. Cross AR, Rae J, Curnutte JT (1995) Cytochrome b-245 of the neutrophil superoxide-generating system contains two nonidentical hemes. Potentiometric studies of a mutant form of gp91phox. J Biol Chem 270(29):17075–17077 29. Decoursey TE, Ligeti E (2005) Regulation and termination of NADPH oxidase activity. Cell Mol Life Sci 62(19–20): 2173–2193. https://doi.org/10.1007/s00018-005-5177-1 30. Musset B, Cherny VV, Morgan D, DeCoursey TE (2009) The intimate and mysterious relationship between proton channels and NADPH oxidase. FEBS Lett 583(1):7–12. https://doi.org/10. 1016/j.febslet.2008.12.005 31. Droste A, Chaves G, Stein S, Trzmiel A, Schweizer M, Karl H, Musset B (2021) Zinc accelerates respiratory burst termination in human PMN. Redox Biol 47:102133. https://doi.org/10.1016/j. redox.2021.102133 32. Conner GE (2021) Regulation of dual oxidase hydrogen peroxide synthesis results in an epithelial respiratory burst. Redox Biol 41: 101931. https://doi.org/10.1016/j.redox.2021.101931 33. Kohn DB, Booth C, Kang EM, Pai SY, Shaw KL, Santilli G, Armant M, Buckland KF, Choi U, De Ravin SS, Dorsey MJ, Kuo CY, Leon-Rico D, Rivat C, Izotova N, Gilmour K, Snell K, Dip JX, Darwish J, Morris EC, Terrazas D, Wang LD, Bauser CA, Paprotka T, Kuhns DB, Gregg J, Raymond HE, Everett JK, Honnet G, Biasco L, Newburger PE, Bushman FD, Grez M, Gaspar HB, Williams DA, Malech HL, Galy A, Thrasher AJ, Net CGDc (2020) Lentiviral gene therapy for X-linked chronic granulomatous disease. Nat Med 26(2):200–206. https://doi.org/10. 1038/s41591-019-0735-5 34. Li Q, Harraz MM, Zhou W, Zhang LN, Ding W, Zhang Y, Eggleston T, Yeaman C, Banfi B, Engelhardt JF (2006) Nox2 and Rac1 regulate H2O2-dependent recruitment of TRAF6 to endosomal interleukin-1 receptor complexes. Mol Cell Biol 26(1):140–154. https://doi.org/10.1128/MCB.26.1.140-154.2006 35. Prosser BL, Ward CW, Lederer WJ (2011) X-ROS signaling: rapid mechano-chemo transduction in heart. Science 333(6048): 1440–1445. https://doi.org/10.1126/science.1202768 36. Sullivan MN, Gonzales AL, Pires PW, Bruhl A, Leo MD, Li W, Oulidi A, Boop FA, Feng Y, Jaggar JH, Welsh DG, Earley S (2015) Localized TRPA1 channel Ca2+ signals stimulated by reactive oxygen species promote cerebral artery dilation. Sci Signal 8(358):ra2. https://doi.org/10.1126/scisignal.2005659 37. Emmerson A, Trevelin SC, Mongue-Din H, Becker PD, Ortiz C, Smyth LA, Peng Q, Elgueta R, Sawyer G, Ivetic A, Lechler RI, Lombardi G, Shah AM (2018) Nox2 in regulatory T cells promotes

583 angiotensin II-induced cardiovascular remodeling. J Clin Invest 128(7):3088–3101. https://doi.org/10.1172/JCI97490 38. Hervera A, De Virgiliis F, Palmisano I, Zhou L, Tantardini E, Kong G, Hutson T, Danzi MC, Perry RB, Santos CXC, Kapustin AN, Fleck RA, Del Rio JA, Carroll T, Lemmon V, Bixby JL, Shah AM, Fainzilber M, Di Giovanni S (2018) Reactive oxygen species regulate axonal regeneration through the release of exosomal NADPH oxidase 2 complexes into injured axons. Nat Cell Biol 20(3):307–319. https://doi.org/10.1038/s41556-018-0039-x 39. Hilenski LL, Clempus RE, Quinn MT, Lambeth JD, Griendling KK (2004) Distinct subcellular localizations of Nox1 and Nox4 in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 24(4):677–683. https://doi.org/10.1161/01.ATV.0000112024. 13727.2c 40. Chu X, Filali M, Stanic B, Takapoo M, Sheehan A, Bhalla R, Lamb FS, Miller FJ Jr (2011) A critical role for chloride channel-3 (CIC-3) in smooth muscle cell activation and neointima formation. Arterioscler Thromb Vasc Biol 31(2):345–351. https://doi.org/10. 1161/ATVBAHA.110.217604 41. Santos CX, Hafstad AD, Beretta M, Zhang M, Molenaar C, Kopec J, Fotinou D, Murray TV, Cobb AM, Martin D, Zeh Silva M, Anilkumar N, Schroder K, Shanahan CM, Brewer AC, Brandes RP, Blanc E, Parsons M, Belousov V, Cammack R, Hider RC, Steiner RA, Shah AM (2016) Targeted redox inhibition of protein phosphatase 1 by Nox4 regulates eIF2alpha-mediated stress signaling. EMBO J 35(3):319–334. https://doi.org/10. 15252/embj.201592394 42. Beretta M, Santos CX, Molenaar C, Hafstad AD, Miller CC, Revazian A, Betteridge K, Schroder K, Streckfuss-Bomeke K, Doroshow JH, Fleck RA, Su TP, Belousov VV, Parsons M, Shah AM (2020) Nox4 regulates InsP3 receptor-dependent Ca(2+) release into mitochondria to promote cell survival. EMBO J 39(19):e103530. https://doi.org/10.15252/embj.2019103530 43. Anilkumar N, Weber R, Zhang M, Brewer A, Shah AM (2008) Nox4 and nox2 NADPH oxidases mediate distinct cellular redox signaling responses to agonist stimulation. Arterioscler Thromb Vasc Biol 28(7):1347–1354. https://doi.org/10.1161/ATVBAHA. 108.164277 44. Zhang M, Brewer AC, Schroder K, Santos CX, Grieve DJ, Wang M, Anilkumar N, Yu B, Dong X, Walker SJ, Brandes RP, Shah AM (2010) NADPH oxidase-4 mediates protection against chronic load-induced stress in mouse hearts by enhancing angiogenesis. Proc Natl Acad Sci USA 107(42):18121–18126. https:// doi.org/10.1073/pnas.1009700107 45. Hecker L, Logsdon NJ, Kurundkar D, Kurundkar A, Bernard K, Hock T, Meldrum E, Sanders YY, Thannickal VJ (2014) Reversal of persistent fibrosis in aging by targeting Nox4-Nrf2 redox imbalance. Sci Transl Med 6(231):231ra247. https://doi.org/10.1126/ scitranslmed.3008182 46. Hancock M, Hafstad AD, Nabeebaccus AA, Catibog N, Logan A, Smyrnias I, Hansen SS, Lanner J, Schroder K, Murphy MP, Shah AM, Zhang M (2018) Myocardial NADPH oxidase-4 regulates the physiological response to acute exercise. eLife 7:e41044. https:// doi.org/10.7554/eLife.41044 47. Pollock JD, Williams DA, Gifford MA, Li LL, Du X, Fisherman J, Orkin SH, Doerschuk CM, Dinauer MC (1995) Mouse model of X-linked chronic granulomatous disease, an inherited defect in phagocyte superoxide production. Nat Genet 9(2):202–209. https://doi.org/10.1038/ng0295-202 48. Nakano Y, Longo-Guess CM, Bergstrom DE, Nauseef WM, Jones SM, Banfi B (2008) Mutation of the Cyba gene encoding p22 causes vestibular and immune defects in mice. J Clin Invest 118(3):1176–1185. https://doi.org/10.1172/JCI33835 49. Mori M, Li G, Hashimoto M, Nishio A, Tomozawa H, Suzuki N, Usami S, Higuchi K, Matsumoto K (2009) Pivotal advance: eosinophilia in the MES rat strain is caused by a loss-of-function

584 mutation in the gene for cytochrome b(-245), alpha polypeptide (Cyba). J Leukoc Biol 86(3):473–478. https://doi.org/10.1189/jlb. 1108715 50. Jackson SH, Gallin JI, Holland SM (1995) The p47phox mouse knock-out model of chronic granulomatous disease. J Exp Med 182(3):751–758. https://doi.org/10.1084/jem.182.3.751 51. Jacob CO, Yu N, Yoo DG, Perez-Zapata LJ, Barbu EA, Kaplan MJ, Purmalek M, Pingel JT, Idol RA, Dinauer MC (2017) Haploinsufficiency of NADPH oxidase subunit neutrophil cytosolic factor 2 is sufficient to accelerate full-blown lupus in NZM 2328 mice. Arthritis Rheumatol 69(8):1647–1660. https://doi.org/ 10.1002/art.40141 52. Ellson CD, Davidson K, Ferguson GJ, O’Connor R, Stephens LR, Hawkins PT (2006) Neutrophils from p40phox-/- mice exhibit severe defects in NADPH oxidase regulation and oxidantdependent bacterial killing. J Exp Med 203(8):1927–1937. https://doi.org/10.1084/jem.20052069 53. Johnson KR, Marden CC, Ward-Bailey P, Gagnon LH, Bronson RT, Donahue LR (2007) Congenital hypothyroidism, dwarfism, and hearing impairment caused by a missense mutation in the mouse dual oxidase 2 gene, Duox2. Mol Endocrinol 21(7): 1593–1602. https://doi.org/10.1210/me.2007-0085 54. Grasberger H, De Deken X, Mayo OB, Raad H, Weiss M, Liao XH, Refetoff S (2012) Mice deficient in dual oxidase maturation factors are severely hypothyroid. Mol Endocrinol 26(3):481–492. https://doi.org/10.1210/me.2011-1320 55. Paffenholz R, Bergstrom RA, Pasutto F, Wabnitz P, Munroe RJ, Jagla W, Heinzmann U, Marquardt A, Bareiss A, Laufs J, Russ A, Stumm G, Schimenti JC, Bergstrom DE (2004) Vestibular defects in head-tilt mice result from mutations in Nox3, encoding an NADPH oxidase. Genes Dev 18(5):486–491. https://doi.org/10. 1101/gad.1172504 56. Guilbault C, Saeed Z, Downey GP, Radzioch D (2007) Cystic fibrosis mouse models. Am J Respir Cell Mol Biol 36(1):1–7. https://doi.org/10.1165/rcmb.2006-0184TR 57. Guillaume F, Otto SP (2012) Gene functional trade-offs and the evolution of pleiotropy. Genetics 192(4):1389–1409. https://doi. org/10.1534/genetics.112.143214 58. Kung JE, Jura N (2016) Structural basis for the non-catalytic functions of protein kinases. Structure 24(1):7–24. https://doi.org/ 10.1016/j.str.2015.10.020 59. Lawrence T, Gilroy DW, Colville-Nash PR, Willoughby DA (2001) Possible new role for NF-kappaB in the resolution of inflammation. Nat Med 7(12):1291–1297. https://doi.org/10. 1038/nm1201-1291 60. Akhurst RJ, Derynck R (2001) TGF-beta signaling in cancer – a double-edged sword. Trends Cell Biol 11(11):S44–S51. https://doi. org/10.1016/s0962-8924(01)02130-4 61. Hecker L, Vittal R, Jones T, Jagirdar R, Luckhardt TR, Horowitz JC, Pennathur S, Martinez FJ, Thannickal VJ (2009) NADPH oxidase-4 mediates myofibroblast activation and fibrogenic responses to lung injury. Nat Med 15(9):1077–1081. https://doi. org/10.1038/nm.2005 62. Gulati S, Thannickal VJ (2019) The aging lung and idiopathic pulmonary fibrosis. Am J Med Sci 357(5):384–389. https://doi. org/10.1016/j.amjms.2019.02.008 63. Bernard K, Thannickal VJ (2019) NADPH oxidases and aging models of lung fibrosis. Methods Mol Biol 1982:487–496. https://doi.org/10.1007/978-1-4939-9424-3_29 64. Williams GC (1957) Pleiotropy, natural selection, and the evolution of senescence. Evolution 11:398–411 65. Lambeth JD (2007) Nox enzymes, ROS, and chronic disease: an example of antagonistic pleiotropy. Free Radic Biol Med 43(3): 332–347

U. G. Knaus et al. 66. Thannickal VJ (2010) Aging, antagonistic pleiotropy and fibrotic disease. Int J Biochem Cell Biol 42(9):1398–1400. https://doi.org/ 10.1016/j.biocel.2010.05.010 67. Aviello G, Knaus UG (2018) NADPH oxidases and ROS signaling in the gastrointestinal tract. Mucosal Immunol 11(4):1011–1023. https://doi.org/10.1038/s41385-018-0021-8 68. Dumas A, Knaus UG (2021) Raising the ‘Good’ oxidants for immune protection. Front Immunol 12:698042. https://doi.org/10. 3389/fimmu.2021.698042 69. Segal BH, Grimm MJ, Khan AN, Han W, Blackwell TS (2012) Regulation of innate immunity by NADPH oxidase. Free Radic Biol Med 53(1):72–80. https://doi.org/10.1016/j.freeradbiomed. 2012.04.022 70. Ducarmon QR, Zwittink RD, Hornung BVH, van Schaik W, Young VB, Kuijper EJ (2019) Gut microbiota and colonization resistance against bacterial enteric infection. Microbiol Mol Biol Rev 83:3. https://doi.org/10.1128/MMBR.00007-19 71. Gennery AR (2021) Progress in treating chronic granulomatous disease. Br J Haematol 192(2):251–264. https://doi.org/10.1111/ bjh.16939 72. Henrickson SE, Jongco AM, Thomsen KF, Garabedian EK, Thomsen IP (2018) Noninfectious manifestations and complications of chronic granulomatous disease. J Pediatric Infect Dis Soc 7(suppl_1):S18–S24. https://doi.org/10.1093/jpids/piy014 73. Diebold BA, Smith SM, Li Y, Lambeth JD (2015) NOX2 as a target for drug development: indications, possible complications, and Progress. Antioxid Redox Signal 23(5):375–405. https://doi. org/10.1089/ars.2014.5862 74. Kim HJ, Seo YH, An S, Jo A, Kwon IC, Kim S (2018) Chemiluminescence imaging of Duox2-derived hydrogen peroxide for longitudinal visualization of biological response to viral infection in nasal mucosa. Theranostics 8(7):1798–1807. https://doi.org/10. 7150/thno.22481 75. Sarr D, Gingerich AD, Asthiwi NM, Almutairi F, Sautto GA, Ecker J, Nagy T, Kilgore MB, Chandler JD, Ross TM, Tripp RA, Rada B (2021) Dual oxidase 1 promotes antiviral innate immunity. Proc Natl Acad Sci USA 118(26):e2017130118. https://doi.org/10. 1073/pnas.2017130118 76. Grasberger H, El-Zaatari M, Dang DT, Merchant JL (2013) Dual oxidases control release of hydrogen peroxide by the gastric epithelium to prevent helicobacter felis infection and inflammation in mice. Gastroenterology 145(5):1045–1054. https://doi.org/10. 1053/j.gastro.2013.07.011 77. Hayes P, Dhillon S, O’Neill K, Thoeni C, Hui KY, Elkadri A, Guo CH, Kovacic L, Aviello G, Alvarez LA, Griffiths AM, Snapper SB, Brant SR, Doroshow JH, Silverberg MS, Peter I, McGovern DP, Cho J, Brumell JH, Uhlig HH, Bourke B, Muise AA, Knaus UG (2015) Defects in NADPH oxidase genes NOX1 and DUOX2 in very early onset inflammatory bowel disease. Cell Mol Gastroenterol Hepatol 1(5):489–502. https://doi.org/10.1016/j. jcmgh.2015.06.005 78. Grasberger H, Magis AT, Sheng E, Conomos MP, Zhang M, Garzotto LS, Hou G, Bishu S, Nagao-Kitamoto H, El-Zaatari M, Kitamoto S, Kamada N, Stidham RW, Akiba Y, Kaunitz J, Haberman Y, Kugathasan S, Denson LA, Omenn GS, Kao JY (2021) DUOX2 variants associate with preclinical disturbances in microbiota-immune homeostasis and increased inflammatory bowel disease risk. J Clin Invest 131(9). https://doi.org/10.1172/ JCI141676 79. Dang PM, Rolas L, El-Benna J (2020) The dual role of reactive oxygen species-generating nicotinamide adenine dinucleotide phosphate oxidases in gastrointestinal inflammation and therapeutic perspectives. Antioxid Redox Signal 33(5):354–373. https://doi. org/10.1089/ars.2020.8018 80. Little AC, Sham D, Hristova M, Danyal K, Heppner DE, Bauer RA, Sipsey LM, Habibovic A, van der Vliet A (2016) DUOX1

34

Quo Vadis NADPH Oxidases: Perspectives on Clinical Translation

silencing in lung cancer promotes EMT, cancer stem cell characteristics and invasive properties. Oncogenesis 5(10):e261. https://doi.org/10.1038/oncsis.2016.61 81. Fortunato RS, Gomes LR, Munford V, Pessoa CF, Quinet A, Hecht F, Kajitani GS, Milito CB, Carvalho DP, Menck CFM (2018) DUOX1 silencing in mammary cell alters the response to genotoxic stress. Oxidative Med Cell Longev 2018:3570526. https://doi.org/10.1155/2018/3570526 82. Ameziane El Hassani R, Buffet C, Leboulleux S, Dupuy C (2019) Oxidative stress in thyroid carcinomas: biological and clinical significance. Endocr Relat Cancer 26(3):R131–R143. https://doi. org/10.1530/ERC-18-0476 83. van der Vliet A, Danyal K, Heppner DE (2018) Dual oxidase: a novel therapeutic target in allergic disease. Br J Pharmacol 175(9): 1401–1418. https://doi.org/10.1111/bph.14158 84. Thannickal VJ, Toews GB, White ES, Lynch JP III, Martinez FJ (2004) Mechanisms of pulmonary fibrosis. Annu Rev Med 55: 395–417. https://doi.org/10.1146/annurev.med.55.091902.103810 85. Thannickal VJ, Zhou Y, Gaggar A, Duncan SR (2014) Fibrosis: ultimate and proximate causes. J Clin Invest 124(11):4673–4677. https://doi.org/10.1172/JCI74368 86. Bernard K, Thannickal VJ (2020) NADPH oxidase inhibition in fibrotic pathologies. Antioxid Redox Signal 33(6):455–479. https://doi.org/10.1089/ars.2020.8032 87. Rajaram RD, Dissard R, Jaquet V, de Seigneux S (2019) Potential benefits and harms of NADPH oxidase type 4 in the kidneys and cardiovascular system. Nephrol Dial Transplant 34(4):567–576. https://doi.org/10.1093/ndt/gfy161 88. Crestani B, Besnard V, Boczkowski J (2011) Signalling pathways from NADPH oxidase-4 to idiopathic pulmonary fibrosis. Int J Biochem Cell Biol 43(8):1086–1089. https://doi.org/10.1016/j. biocel.2011.04.003 89. Sanders YY, Lyv X, Zhou QJ, Xiang Z, Stanford D, Bodduluri S, Rowe SM, Thannickal VJ (2020) Brd4-p300 inhibition downregulates Nox4 and accelerates lung fibrosis resolution in aged mice. JCI insight 5(14). https://doi.org/10.1172/jci.insight. 137127 90. Bettaieb A, Jiang JX, Sasaki Y, Chao TI, Kiss Z, Chen X, Tian J, Katsuyama M, Yabe-Nishimura C, Xi Y, Szyndralewiez C, Schroder K, Shah A, Brandes RP, Haj FG, Torok NJ (2015) Hepatocyte nicotinamide adenine dinucleotide phosphate reduced oxidase 4 regulates stress signaling, fibrosis, and insulin sensitivity during development of steatohepatitis in mice. Gastroenterology 149(2):468–480 e410. https://doi.org/10.1053/j.gastro.2015. 04.009 91. Lan T, Kisseleva T, Brenner DA (2015) Deficiency of NOX1 or NOX4 prevents liver inflammation and fibrosis in mice through inhibition of hepatic stellate cell activation. PLoS One 10(7): e0129743. https://doi.org/10.1371/journal.pone.0129743 92. Liang S, Kisseleva T, Brenner DA (2016) The role of NADPH oxidases (NOXs) in liver fibrosis and the activation of Myofibroblasts. Front Physiol 7:17. https://doi.org/10.3389/fphys. 2016.00017 93. Nabeebaccus A, Zhang M, Shah AM (2011) NADPH oxidases and cardiac remodelling. Heart Fail Rev 16(1):5–12. https://doi.org/10. 1007/s10741-010-9186-2 94. Ago T, Kuroda J, Pain J, Fu C, Li H, Sadoshima J (2010) Upregulation of Nox4 by hypertrophic stimuli promotes apoptosis and mitochondrial dysfunction in cardiac myocytes. Circ Res 106(7):1253–1264. https://doi.org/10.1161/CIRCRESAHA.109. 213116 95. Kuroda J, Ago T, Matsushima S, Zhai P, Schneider MD, Sadoshima J (2010) NADPH oxidase 4 (Nox4) is a major source of oxidative stress in the failing heart. Proc Natl Acad Sci USA 107(35):15565–15570. https://doi.org/10.1073/pnas.1002178107

585 96. Babelova A, Avaniadi D, Jung O, Fork C, Beckmann J, Kosowski J, Weissmann N, Anilkumar N, Shah AM, Schaefer L, Schroder K, Brandes RP (2012) Role of Nox4 in murine models of kidney disease. Free Radic Biol Med 53(4):842–853. https://doi. org/10.1016/j.freeradbiomed.2012.06.027 97. Sedeek M, Nasrallah R, Touyz RM, Hebert RL (2013) NADPH oxidases, reactive oxygen species, and the kidney: friend and foe. J Am Soc Nephrol 24(10):1512–1518. https://doi.org/10.1681/ASN. 2012111112 98. Holterman CE, Read NC, Kennedy CR (2015) Nox and renal disease. Clin Sci (Lond) 128(8):465–481. https://doi.org/10.1042/ CS20140361 99. Stenke E, Aviello G, Singh A, Martin S, Winter D, Sweeney B, McDermott M, Bourke B, Hussey S, Knaus UG (2020) NADPH oxidase 4 is protective and not fibrogenic in intestinal inflammation. Redox Biol 37:101752. https://doi.org/10.1016/j.redox.2020. 101752 100. Horowitz JC, Thannickal VJ (2019) Mechanisms for the resolution of organ fibrosis. Physiology 34(1):43–55. https://doi.org/10.1152/ physiol.00033.2018 101. Zhou Y, Huang X, Hecker L, Kurundkar D, Kurundkar A, Liu H, Jin TH, Desai L, Bernard K, Thannickal VJ (2013) Inhibition of mechanosensitive signaling in myofibroblasts ameliorates experimental pulmonary fibrosis. J Clin Invest 123(3):1096–1108. https://doi.org/10.1172/JCI66700 102. Rangarajan S, Bone NB, Zmijewska AA, Jiang S, Park DW, Bernard K, Locy ML, Ravi S, Deshane J, Mannon RB, Abraham E, Darley-Usmar V, Thannickal VJ, Zmijewski JW (2018) Metformin reverses established lung fibrosis in a bleomycin model. Nat Med 24(8):1121–1127. https://doi.org/10.1038/ s41591-018-0087-6 103. Qu J, Yang SZ, Zhu Y, Guo T, Thannickal VJ, Zhou Y (2021) Targeting mechanosensitive MDM4 promotes lung fibrosis resolution in aged mice. J Exp Med 218(5). https://doi.org/10.1084/jem. 20202033 104. Rehan M, Kurundkar D, Kurundkar AR, Logsdon NJ, Smith SR, Chanda D, Bernard K, Sanders YY, Deshane JS, Dsouza KG, Rangarajan S, Zmijewski JW, Thannickal VJ (2021) Restoration of SIRT3 gene expression by airway delivery resolves age-associated persistent lung fibrosis in mice. Nat Aging 1(2): 205–217. https://doi.org/10.1038/s43587-021-00027-5 105. Chanda D, Rehan M, Smith SR, Dsouza KG, Wang Y, Bernard K, Kurundkar D, Memula V, Kojima K, Mobley JA, Benavides GA, Darley-Usmar V, Kim YI, Zmijewski JW, Deshane JS, De Langhe S, Thannickal VJ (2021) Mesenchymal stromal cell aging impairs the self-organizing capacity of lung alveolar epithelial stem cells. eLife 10. https://doi.org/10.7554/eLife.68049 106. Brault J, Vigne B, Meunier M, Beaumel S, Mollin M, Park S, Stasia MJ (2020) NOX4 is the main NADPH oxidase involved in the early stages of hematopoietic differentiation from human induced pluripotent stem cells. Free Radic Biol Med 146:107– 118. https://doi.org/10.1016/j.freeradbiomed.2019.10.005 107. Schroder K, Zhang M, Benkhoff S, Mieth A, Pliquett R, Kosowski J, Kruse C, Luedike P, Michaelis UR, Weissmann N, Dimmeler S, Shah AM, Brandes RP (2012) Nox4 is a protective reactive oxygen species generating vascular NADPH oxidase. Circ Res 110(9):1217–1225. https://doi.org/10.1161/CIRCRESAHA. 112.267054 108. Groeger G, Mackey AM, Pettigrew CA, Bhatt L, Cotter TG (2009) Stress-induced activation of Nox contributes to cell survival signalling via production of hydrogen peroxide. J Neurochem 109(5): 1544–1554. https://doi.org/10.1111/j.1471-4159.2009.06081.x 109. Ray R, Murdoch CE, Wang M, Santos CX, Zhang M, AlomRuiz S, Anilkumar N, Ouattara A, Cave AC, Walker SJ, Grieve DJ, Charles RL, Eaton P, Brewer AC, Shah AM (2011) Endothelial Nox4 NADPH oxidase enhances vasodilatation and reduces blood

586 pressure in vivo. Arterioscler Thromb Vasc Biol 31(6):1368–1376. https://doi.org/10.1161/ATVBAHA.110.219238 110. Schurmann C, Rezende F, Kruse C, Yasar Y, Lowe O, Fork C, van de Sluis B, Bremer R, Weissmann N, Shah AM, Jo H, Brandes RP, Schroder K (2015) The NADPH oxidase Nox4 has antiatherosclerotic functions. Eur Heart J 36(48):3447–3456. https:// doi.org/10.1093/eurheartj/ehv460 111. Langbein H, Brunssen C, Hofmann A, Cimalla P, Brux M, Bornstein SR, Deussen A, Koch E, Morawietz H (2016) NADPH oxidase 4 protects against development of endothelial dysfunction

U. G. Knaus et al. and atherosclerosis in LDL receptor deficient mice. Eur Heart J 37(22):1753–1761. https://doi.org/10.1093/eurheartj/ehv564 112. Vendrov AE, Vendrov KC, Smith A, Yuan J, Sumida A, Robidoux J, Runge MS, Madamanchi NR (2015) NOX4 NADPH oxidase-dependent mitochondrial oxidative stress in agingassociated cardiovascular disease. Antioxid Redox Signal 23(18): 1389–1409. https://doi.org/10.1089/ars.2014.6221 113. Okonko DO, Shah AM (2015) Heart failure: mitochondrial dysfunction and oxidative stress in CHF. Nat Rev Cardiol 12(1):6–8. https://doi.org/10.1038/nrcardio.2014.189

Index

A AlphaFold, 46, 48, 319, 420 Arthropod NADPH oxidases (Noxs) DUOX cuticle formation, 478, 484 immune system, 478–480 tissue regeneration, 481–482 hemocytes, 478 lack of gene for p22phox, 483 Nox4-art, 478, 482–484 Nox5 contraction, 85, 215, 221, 223, 225, 481, 482 immune system, 481, 482 tissue regeneration, 480 Artificial intelligence (AI), 317, 319–320 ATP, 20, 26, 27, 41, 42, 44, 69, 120, 187, 238, 318, 387, 492 Autoinhibition autoinhibitory region, 39

B Babior, B., 5–8, 11, 12, 19, 20, 22, 41, 44, 46, 94, 478, 538 Bacterial genomes Streptococcus pneumoniae Nox (SpNox), 140 B cells, 106, 153, 182, 225, 271, 281, 290, 300, 301, 541, 545, 558, 560, 564, 565 Bokoch, G.M., 29–32, 123–130

C Cathepsin G, 98, 410, 565 Cdc42 Cdc42/RAC interactive binding (CRIB), 33, 34, 126, 127 Cell-free system anionic amphiphiles, 16, 25, 26, 33, 39, 42, 82 arachidonic acid, 16, 74, 117, 120, 264 cell-free assay, 26, 278 lithium dodecyl sulfate (LiDS), 25, 26 myristic acid, 15, 16, 118 oleic acid, 23, 74 semi-recombinant cell-free system, 24, 121 sodium dodecyl sulfate (SDS), 16, 25, 26, 31, 32, 120, 128 Chimeras, 29, 33, 34, 37, 43, 45, 46 Chronic granulomatous disease (CGD) autosomal recessive (AR) CGD, 9, 13, 14, 20, 22, 24, 26–28, 73, 74, 93–97, 251, 280, 281, 541, 547, 549, 563, 568 cell hybrids from different CGD patients, 538 cellular models, 280, 545, 577 clinical aspects, 539–542, 562 diagnosis, 540, 543–545, 548, 549 dysregulated inflammation, 147

genetic diagnosis, 547–549 granuloma, 264 PLB-985 cells, 545 residual levels of ROS formation, 355, 549 X910 CGD, 544–547 X-linked CGD, 82, 314 Cloning cloning of p47phox and p67phox, 27, 28 cloning of X-CGD gene, 105–109 cloning the gene for p22phox, 18 Conformational change in Nox2 affects distances between redox stations, 37–39 electron flow from FAD to heme, 36 hydride transfer from NADPH to FAD, 36–39 in Noxa1, 33, 35, 264, 269 in p47phox circular dichroism, 42, 254 quenching of intrinsic tryptophan fluorescence, 42 in p67phox deletion mutant p67phox(1-526)Δ(259-279), 46 deletion of residues 181-193, 46 point mutations Q115R and K181E, 46 in Rac upon GDP to GTP exchange, 31, 34 Cross-linking tyrosine cross-linking by activation of Udx1, the Duox sea urchin homolog, 237 in Anopheles gambiae, 237 in Rhodnius prolixus, 237 Cryo-EM human and mouse DUOX1-DUOXA1 complexes, 35, 509 structure of human DUOX1 in the presence of high Ca2+, 48, 509 structure of human STEAP1, 525, 528 structure of STEAP4, 524, 525, 527, 528 Cytochrome b558 b cytochrome association of a b cytochrome and a flavoprotein, 9 cytochrome b–245, 3, 9 low potential cytochrome b, 92 Cytochrome c cytochrome c-based O2 – measurements, 324 cytochrome c reductase, 11 Cytosolic components, 22 2’,5’-ADP-agarose, 26 GTP-agarose, 26, 28 p40phox absence of severe bacterial infections in p40phox-deficient patients, 280, 281 deficiency of p40phox, 280 hyperinflammation phenomena, 281

# The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 E. Pick (ed.), NADPH Oxidases Revisited: From Function to Structure, https://doi.org/10.1007/978-3-031-23752-2

587

588 Cytosolic components (cont.) interaction between the PB1 (phox and Bem1) domains in the C-termini of p67phox and p40phox, 276 NCF4 gene, 276 phosphorylation of p40phox, 276, 278 PX domain in p40pho, 43, 265, 268, 276, 278, 295, 296, 436, 542 role in inflammatory reactions, 276, 279 role in NADPH oxidase localization, 277, 278 translocates to the membrane, 277 p47phox interaction with p22phox, 42, 253, 254 NADPH oxidase/Nox2 priming, 252 Neutrophil Cytosolic Factor 1 (NCF1) gene, 251, 344 phosphorylated serines, 252, 254 phosphorylation of p47phox, 42, 251–254, 256, 271, 295, 347, 391 PX domain of p47phox, 40, 254, 296 substrate for protein kinase C, 251 p67phox activation domain, 35, 83, 84, 149, 265–271, 277, 278, 295, 314, 348, 354, 436, 438, 539 β-hairpin-insertion, 266 expanded activation domain, 266, 269 gene NCF2, 169, 264 NoxR, a fungal p67phox-related protein, 264, 269–271 Prdx6 is able to associate with p67phox, 269 Rac-binding site, 266 Rac binding to the TPR domain of p67phox, 266 ternary complex with p47phox and p40phox, 264, 265 tetratricopeptide repeat (TPR) motifs, 149, 264, 265, 267, 295, 298, 406, 436

D Dehydrogenase domain, 235 Deoxycholate, 10, 11, 25, 72, 112 1,2-diacylglycerol (DAG), 41, 42 Diaphorase reducing nitro blue tetrazolium (NBT), 4 Dictyostelium discoideum Noxs Dd-CybA, the amoebal homolog of the p22phox subunit, 405 EF-hands, 404, 408, 409 homologies (identities) with human Nox2, 401 intraphagosomal processing of microbes, 409 NADPH oxidases in Dictyostelium development, 408, 409 NcfA is devoid of the p67phox activation domain, 406 noxA, B and C genes, 401 NoxA, B and C and homologs of the p22phox and p67phox proteins, 399 p67phox-like gene ncfA, 406 Rac1A, -1B and -1C, 406 sentinel cells/S-cells, 409 Diphenylene iodonium (DPI), 94, 192, 233, 350, 352, 357, 416, 446, 452, 454, 458, 459, 471, 472, 480, 481, 492, 493, 499 Dual oxidase (DUOX) arthropods, 216, 478, 480, 482, 483 Ca2+ binding, 139 Caenorhabditis elegans, 153, 231, 232, 239, 467–475 dehydrogenase domain (DH), 346, 364 EF-hand motifs, 139, 231, 263, 264, 401 FAD-binding and NADPH-binding domains, 507 gastrointestinal tract, 239 H2O2 generation, 232, 233 host defense, 153, 154, 159, 238, 239, 364, 429, 507, 579, 580 peroxidase homology domain (PHD), 139–141, 345, 346, 507

Index pleckstrin homology domain, 509 respiratory epithelium, 238 thyroid carcinogenesis, 236, 237 thyroid hormone synthesis, 136, 137, 140, 229, 230, 233, 240, 364, 367, 496 thyroid stimulating hormone (TSH), 229–232, 235, 236 thyroperoxidase (TPO), 136, 140, 153, 230, 231, 234, 240, 385, 473 transmembrane domain, 35, 140, 401, 468, 469, 507, 529 tyrosine crosslinking, 140, 480 Udx1, 140, 237 zebrafish, x, 153, 239, 364, 490, 494, 496, 498 DUOX1 amphipathic helix (preM1), 509 binding sites for NADPH, FAD, heme and oxygen, 511–513 cation binding sites (CBS1 and CBS2), 509 distance between NADPH and FAD, 37, 38 distance between two hemes, 515 DUOX1 and DUOX2 genes, 231 EF module, 507–509, 511, 515, 516 electron transfer pathway, 38, 513–515 glycosylation, 231, 233 2:2 heterotetrameric assembly, 509 NBD adopts a "Rossmann-like fold", 509 DUOX1/DUOXA1 pair complex architecture, 510–513 cryo-EM, 235, 365 glycosylation, 233 H2O2 production, 238, 507 transition from the endoplasmic reticulum to the Golgi, 233 DUOX2 DUOX1 and DUOX2 genes, 231 DUOX2 mutations, 154, 235, 236, 240, 577 glycosylation, 231, 233 hypothyroidism, 137, 140, 153, 154, 232, 233, 236, 240, 316, 510, 513, 515 inflammatory bowel diseases (IBD) inactivating DUOX2 variants in, 577 DUOX2/DUOXA2 pair transition from the endoplasmic reticulum to the Golgi, 233 Duoxa-/- mice congenital hypothyroidism, 233 maturation defect of Duox proteins, 233 DUOXA1 DUOXA genes, 232, 235 extracellular domain of, 509, 510 M2-M3 and M4-M5 loops, 510 N-terminal peptide (NTP), 510 similarities to the claudin superfamily, 510 DUOXA2 DUOXA2 genes, 240 hypothyroidism, 137, 232, 236, 510, 577

E Electron paramagnetic resonance (EPR), 9, 10, 16, 94, 114–116, 328–330 Electron transfer one electron transfer, 148 two electron transfer, 116 Essential for reactive oxygen species protein (EROS) cytochrome b-245 chaperone 1 (CYBC1), 386, 387 Cybc1-/-, 386 cytochrome b-245 chaperone 1, 386 purinergic P2X7 receptor ATP-activated cation channel, 387

Index F Fatty acids critical micellar concentration (CMC), 23, 25 Krafft points, 25 long chain, 22, 23, 25 saturated, 22, 23, 25, 112 unsaturated, 22, 23, 25, 112 Ferredoxin-NADP+ reductase (FNR), viii, 21, 22, 35, 36, 38, 39, 94–96, 140, 263, 447, 513, 529, 538 Ferric reductase iron reductase, 19 Flavin adenine dinucleotide (FAD) FAD-binding domain isoalloxazine binding site, 21 ribitol binding site, 21 FADH2, 37, 114, 128, 148, 527 FAD semiquinone, 37 flavoproteins, 94 riboflavin, 383 Flavin mononucleotide (FMN), 5, 7, 23, 95, 114, 383, 538 Formyl-methyl-leucyl-phenylalanine (fMLF), 125, 153, 251, 252, 264, 276, 278, 279, 391, 392, 537, 539, 543, 544 Formyl-Met-Leu-Phe (fMLP), 23, 31, 43, 251 Fridovich, I., 6, 7, 67–69, 416 Fungal Noxs (multicellular) absence of p47phox and p40phox homologues, 436 ascospore germination, 269, 431 crippled growth, 431 filamentous fungi, 140, 429, 431, 438 fruiting body, 269, 416 NoxA, NoxB and NoxC, 429 NoxC Ca2+ signaling, 482 EF-hand, 264, 270, 409 NoxR homologue of p67phox, 264, 269–271, 440 Phox and Bem1 (PB1) domain, 264, 439 RacA, 270, 271, 436

G GDP nonhydrolysable GDP analogue, 15, 28 Gel filtration, 10, 12, 20, 45, 97 GTP binding, 20, 26, 30–32, 34, 83, 124–126, 277, 289, 292, 293, 297, 387, 539 hydrolysis, 30, 31, 289, 297, 388, 576 nonhydrolysable GTP analogue, 15, 28, 30 GTPase GTPase activating protein, 28, 30, 138, 289, 446 GTPase binding domain, 33 RhoGTPases, 388 Guanine nucleotide exchange factor (GEF) Dbl family, 31, 291, 298 Dbl homology (DH) domain, 31, 291 pleckstrin homology (PH) domain, 31 P-Rex, 290, 296, 388 Tiam, 30, 44, 290, 301, 388 Trio, 30, 44, 388 Vav, 278, 290, 296, 301, 388

H Hearing loss age-related, 172 cisplatin-induced, 171, 172

589 noise-induced, 172 Heme bis-heme motif, 18, 48 carbon monoxide (CO), 16 high-spin, 16, 115, 117 inner heme, 18, 512, 513, 515, 529, 530 low-spin, 16, 115–117 midpoint potentials, 18, 115, 538 outer heme, 17, 18, 47, 48, 165, 263, 512, 513, 529–531 outer sphere, 16, 18, 527, 529 reduced minus oxidized spectra, 545 Hexose monophosphate shunt (HMPS), 4–7, 112, 538 High-performance liquid chromatography (HPLC), 95, 328, 332, 470 Histidine His101, 18, 19 His115, 18, 19 His209, 18, 19 His222, 18, 19, 180 Human Genome Organization (HUGO), 3, 84, 138 Hydride transfer, 36–39, 83, 84, 527, 531 Hydrogen peroxide (H2O2), 4, 67, 69, 70, 72, 98, 222, 229, 240, 249, 263, 323, 330–334, 344, 352, 387, 399, 416, 440, 445, 458, 459, 468–474, 478, 491, 496, 497, 507, 513, 515, 517, 538, 543, 576 Hydrophobic binding, 25, 29–31, 40, 43, 45, 96, 383, 387, 405, 469, 515 interactions, 29, 40, 43, 45, 47, 292, 383, 387, 510, 517 residues, 29, 40, 43, 45, 47, 148, 292, 383, 387, 405, 469, 483, 510, 515

I Inflammatory bowel disease (IBD) Crohn’s disease, 239 Interferon gamma, 256, 300, 391 Interleukin interleukin 1 (IL-1), 225, 540, 566 interleukin 3 (IL-3), 563 interleukin 4 (IL-4), 151, 238 interleukin 6 (IL-6), 154, 279, 566 interleukin 10 (IL-10), 151, 154, 256 interleukin 12 (IL-12), 566 interleukin 13 (IL-13), 151, 237, 238 interleukin 17 (IL-17), 151, 155, 256 interleukin 23 (IL-23), 566 interleukin 27 (IL-127), 566 Iron sulphur protein, 94

K Kakinuma, K., 9, 15, 16, 22, 36, 94, 111–118 Karnovsky, M., 4–6, 13, 22, 67, 68, 70, 538 Klebanoff, S.J., vii, 7, 68, 70, 538

L Leukocyte polymorphonuclear, 23, 125 Lipopolysaccharide (LPS), 152, 182, 184, 185, 190, 238, 252, 385, 391, 407, 410, 478, 481 Lysosomes, 6, 230, 540

M Machine learning, 319 Macrophage cytosols, 27, 28, 126

590 Macrophage (cont.) guinea pigs, 22, 126 membranes, 27, 30, 44, 98, 126 Membrane liposomes, 10, 26, 31, 43, 292 solubilized, 10, 19, 26, 38, 43, 517 Metchnikoff, I., 68, 77, 78 Methods to measure reactive oxygen species (ROS) detection of H2O2 oxidation of amplex red, 332 oxidation of luminol and analogs, 331 oxidation of phenolic compounds, 332 oxidation of reduced fluorophores, 331 detection of Nox-derived superoxide (O2 –) and H2O2 in cell-free and cellular systems, 323–330, 335 detection of superoxide (O2 –) electron paramagnetic resonance (EPR) spin trapping, 328, 330 lucigenin chemiluminescence, 326, 327 oxidation of coelenterazine, 327 oxidation of ethidium-based probes, 328 oxidation of luminol and analogs, 327 reduction of acetylated ferricytochrome c, 324 reduction of ferricytochrome c, 324 reduction of tetrazolium salts, 324–326 Microbiomes, 152, 154–156, 158, 159, 239, 479, 480, 579 ML-171, 129

N NADH, 4–6, 13, 20, 23, 70–75, 92, 112–114, 186, 332, 416, 478, 513, 538 NADPH binding domain adenine, 527 nicotinamide, 21 pyrophosphate, 21 ribose, 513 NADPH oxidase (Nox) activators, 16, 19, 22, 30, 149, 167, 169, 251, 254, 264, 315, 384 antagonistic pleiotropy, 579 BioGRID, 380 complex assembly, 25, 316, 349, 380, 542, 545, 579 dehydrogenase region (DHR), 9, 435 interactome, 380 isoforms, 3, 86, 87, 180, 181, 190, 191, 215, 216, 221, 225, 323, 324, 327, 334, 344, 347, 348, 353–359, 362, 365, 399, 407, 434, 478, 482, 489, 490, 493–496, 499, 500, 529, 579, 581 Nox-associated proteins, 380, 381, 389, 391–392 negative regulator of reactive oxygen species, 391, 392 protein disulfide isomerase, 383, 384 proteins cross-talking with Nox complexes, 380–392 synaptotagmin like-1 NADPH oxidase (Nox) inhibitors, 386 NADPH oxidase (Nox) inhibitors cell-free systems, 9, 23–25, 39, 251, 252, 323, 539 DUOX1 and DUOX2, 3, 251, 263, 330, 364 heterologous reconstituted cell lines, 149, 347 need for isoform-specific Nox inhibitors, 365, 367 Nox targeting strategies, 347–349 Nox1, 84, 128, 138, 215, 250, 263, 287, 313, 344, 379, 401, 415, 429, 447, 481, 489, 507, 531, 576 Nox2, 84, 91, 105, 126, 138, 215, 251, 263, 287, 313, 326, 344, 379, 399, 415, 429, 446, 471, 477, 489, 513, 531, 538, 576 Nox4, 84, 129, 138, 215, 313, 323, 346, 379, 401, 416, 449, 478, 489, 517, 576

Index Nox5, 86, 139, 215, 251, 263, 313, 346, 379, 401, 418, 429, 447, 478, 489, 529 pan-Nox inhibitors, ix, 347, 349–351, 353, 365 peptidic inhibitors, 347, 349, 354, 360 Nematode Noxs biological roles aging, 471 cuticle, 470 development, 470 immunity, 470–471 stress, 472 vulva, 470 bli-3 structure, 471–473 blistered, 467, 471, 473 Caenorhabditis elegans, 467, 470–474 peroxidase domain, 467, 468, 470, 471, 473, 474 regulators DUOXA, 472 exogenous peroxidases, 473 small G proteins, 472 tetraspanin, 472 Sydney Brenner’s bli mutants, 467 Neutrophils azurophil granules, 542, 547 extracellular nets, 386 specific granules, 9, 92, 94, 268 Nitric oxide synthase (NOS) endothelial nitric oxide synthase NOS (eNOS), 355 inducible nitric oxide synthase NOS (iNOS), 152, 279 Nitro blue tetrazolium (NBT) formazan, 92, 544 Nox1 binding to the SH3 domain of Noxa1, 149 colon adenocarcinoma, 129, 157 dependence of Nox1 activity on Rac1, 149 dextran sulfate sodium (DSS)-induced colitis, 152 loss-of-function (LoF) genetic variants associated with human inflammatory bowel disease, 151, 153 Mox1, 84, 138, 148 oncogenic KRas-Val12 in the original NIH-3T3 cell clones, 148 overexpression, 157 Nox2 (gp91phox) amphiphile-independent Nox2 activation, 19, 33, 42, 43 CGD, viii, 19, 147, 149, 153, 355, 401, 455, 498, 513, 539, 545, 577, 579, 580 C-terminal F570, 39 CYBB, 105–109, 250, 386, 455, 492, 493, 495, 549, 577 FAD binding domain, 348 glycosylation, 47 insertion sequence, 21, 36 location of hemes, 18 loop D, 36, 47 NADPH binding domain, viii, 12, 347 polybasic region, 36 purification, 20, 47, 347 redox stations, 37–39 transmembrane region, 36 X-linked CGD, 404 Nox3 cisplatin ototoxicity, 173 drug target, 173 expression and subunit composition, 166–169 inner ear expression, 173 otoconia, 100, 166, 169–171, 173, 315

Index otoconial agenesis, 173 pathological significance, 165–173 pattern of Nox3 expression, 166 physiological functions, 165–173 Nox4 (renal oxidase, renox) cancer, 184, 194, 195, 416, 499, 581 downregulation, 183, 184, 189 fibrosis, 181, 182, 185, 192, 193, 352, 360, 361, 580 H2O2 production, 180–182, 189, 236, 483 inhibitors, 193, 196, 198, 351, 352, 360–362 Nox4-/- mice, 188–193, 196 pathophysiology, 188 polymerase delta interacting protein-2 (Poldip2), 181, 318, 359 splice variants, 180, 181, 189 subcellular localization, 181, 449, 459 Nox5 Ca2+ binding, 139, 217, 219, 349 calmodulin, 36, 215, 218, 318, 345, 363 cancer, 216, 220, 221, 225, 363, 499 crystal structure Cylindrospermum stagnale Nox5 (csNox5), 17, 418, 529 EF hands, 139, 140, 216–218, 318, 346, 404, 437 Hsp70, 215, 219, 391 Hsp90, 215, 219, 345, 362, 364, 387, 391 isoforms, 220 molecular chaperones, 215, 218, 391 nitric oxide, 219 sperm function, 221, 225 structure, 21, 37, 216, 217, 365, 418, 530 Noxa1 activates Nox1, 268 enhances Nox3-catalyzed superoxide production, 268 peroxiredoxin 6, 269 poor activator of Nox2, 268 Rac binding, 268 Noxo1 constitutive association with p22phox, 255 discovery, 254 expression of Noxo1, 256 partner of Nox1, 254 phosphorylation of Noxo1, 151, 255, 256 PX domain of Noxo1, 151, 254 splicing, 255 structure of Noxo1, 254, 255 subcellular localization, 255

O Octyl glucoside, 14 Oxygen anaerobic conditions, 4, 8, 9, 13, 14, 67 consumption, 4–6, 23, 68, 70, 72, 98, 117, 133, 249, 323, 537, 538, 543 dioxygen, 418, 513 uptake, 4–7, 11, 13, 23, 68, 112, 113, 224, 249, 537

P p21-activated kinase (PAK), 33, 34, 126, 127, 129, 292–298, 301, 381, 382, 388, 424–426 p21-binding domain (PBD), 33–35, 382 p22phox CYBA, 153, 167, 169, 250, 344, 386, 402, 405, 406, 408, 545 Cyba -/- mice, 155 Cyba variant in mice, 154

591 Cyba-/- mice, 172 P156Q mutation, 318 Poldip2-p22phox, 389 unique binding surface for Nox4, 180 Paradigm shifts, 3–50 p-chloromercuribenzoate (PCMB), 11, 12, 14, 23 Peptide walking, 40, 45, 298, 354 Peroxidase hypochlorous acid, 471 myeloperoxidases, 87, 134–136, 350, 468, 471 thyroperoxidase, 136, 140, 229, 473 Phagocyte lysosomes, 279 membrane liposomes, 30, 42 phagolysosomes, 409 phagosomal pH regulation, 542 Phagocytosis, 3–6, 13, 23, 44, 67, 69, 70, 77, 93, 98, 111, 112, 133, 136, 150, 249, 264, 268, 271, 276, 278, 300, 400, 401, 409, 455, 478, 490, 498, 499, 537, 539, 542, 543, 547 Phorbol esters phorbol myristate acetate (PMA), 8–14, 19, 20, 22–24, 26, 27, 31, 41, 43, 45, 73, 96, 97, 151, 156, 233, 251, 252, 254, 255, 264, 268, 275, 276, 278, 279, 314, 315, 358, 359, 364, 405, 410, 472, 474, 537, 539, 543–545 Phospholipase phospholipase A2, 22, 151, 269, 385 phospholipase C, 25, 294 Phospholipids anionic membrane phospholipids, 43, 44 phosphatidic acid (PA), 26, 31, 43, 215, 252, 254, 277, 296, 381, 539 phosphatidylcholine (PC), 26, 43 phosphatidylethanolamine (PE), 26 phosphatidylglycerol (PG), 31, 43 phosphatidylinositol (PI), 31, 43, 254 phosphatidylinositol 3-phosphate (PtdIns(3)P), 43, 265, 268, 271, 296, 385, 386 phosphatidylinositol 3,4-bisphosphate (PtdIns(3,4)P2), 33, 40, 254, 381, 539 phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P3), 30, 31, 290, 296, 388 phosphatidylserine (PS), 26, 31, 277, 296 Phosphorylation of serines, 28, 40, 41 of tyrosines, 137, 349 Plant NADPH oxidases (Noxs) Arabidopsis, 140, 446, 452, 458 Arabidopsis genome (AtRBOHA-J), 404, 446 biological function plant immunity, 446, 455–456 pollen fertility, 453–454 seed germination, 454–455 stomata closing, 456–458 stress responses, 140 systemic responses, 458–459 Ca2+-biding EF-hand motifs, 447 cellular distribution, 179, 180 discovery of plant Noxs, 446–447 gene expression, viii, 158, 167, 182, 193, 197, 223, 225, 361, 362, 458, 579 G-protein activation, 446 phosphorylation, 15, 39, 41, 42 regulation of activation, 447–449 respiratory burst oxidase homologues (RBOHs) cloned first from rice, 446 RBOHA, 140, 446

592 Plant NADPH oxidases (Noxs) (cont.) structure of RBOH proteins, 447 PLB-985 cells, 36, 37, 545 Positional cloning, 105–109, 545 Potassium channels, 99, 100 Proline polyproline type II helix, 40 proline-rich region (PRR), 39, 139, 251, 255, 318, 436, 448 Protein kinase C (PKC), 22–24, 26, 39, 41, 42, 73, 82, 120, 137, 150, 151, 168, 215, 219, 221, 222, 225, 233, 251–253, 255, 264, 268, 271, 276, 294, 295, 313, 314, 317, 344, 349, 362, 381, 382, 472, 474, 475, 479, 543 Protein–protein interaction BioID assay, 319 PX domains, 40, 42–44, 86, 128, 149, 151, 254, 255, 265–268, 271, 276–278, 295, 296, 347, 380, 382, 390, 539, 542

Q Quastel, J.H., 4–7, 20, 69, 71, 538

R Rac β hairpin, 34, 35, 46 binding to p67phox, 34, 298, 406 conformational change after GDP/GTP exchange, 34 insert region, 83 polybasic region, 31, 289, 349 prenylation geranylgeranyl, 29 switch I, 30, 34, 35, 267, 298 switch II, 298 Rac1 macrophages, 28, 29, 126, 250, 276, 296, 300, 384, 387 Rac2 hematopoietic system, 289 mutation D57N, 30, 301 Rac3 in the brain and testis, 289 Rac-GDP, 149 Rac-GTP, 31–33, 35, 41–43, 46, 264–267, 277, 295, 298 Rap1A association of Rap1A with cytochrome b558, 15 Reactive oxygen species (ROS), 7, 25, 36, 40, 47, 70, 76, 84–86, 129, 133–135, 137, 138, 141, 147, 148, 151–156, 158, 159, 165, 166, 171, 173, 179–182, 184–187, 190–194, 196, 197, 215–217, 219–222, 224, 225, 231, 233, 237–240, 249–252, 255, 256, 263, 264, 279–281, 287, 291, 292, 294, 300, 313, 315, 316, 318–319, 323–335, 343, 344, 346, 348–362, 364, 382–386, 388–392, 399, 407–410, 414, 416–418, 421, 422, 429–435, 440, 445–450, 452–459, 470–473, 477–484, 489–500, 507, 538–545, 549, 550, 557, 564, 565, 568, 576, 578, 580, 581 Redox station, 21 Respiratory burst, 11, 15, 67–74, 76, 78, 81, 82, 92, 93, 98, 101, 112, 113, 119, 124, 126, 133–135, 237, 249, 278, 281, 383, 384, 386, 391, 429, 448, 450, 455, 459, 480, 490, 491, 498, 537, 538 Reverse genetics, 17, 95, 107, 490 Rho GTPases CAAX motif, 289, 292 effectors, 287, 289, 292–298, 301, 302 human diseases, 287, 300, 302 molecular switch, 289, 296 regulators, 287, 289, 291, 294, 296–298, 300, 301, 440

Index Rho GDP dissociation inhibitor (RhoGDI) microsyntenic clusters with genes for protein disulfide isomerase (PDI) family, 388 RhoGDI1, 30, 292 RhoGDI2, 292, 388 RhoGDI3, 292, 388 Rossi, F., 4–6, 10, 15, 71, 72, 111, 112

S SDS-polyacrylamide gel electrophoresis (SDS-PAGE), 10–12, 14, 15, 27, 251, 254 SH3 domain SH3 tandem, 41 Six-transmembrane epithelial antigen of the prostate enzymes (STEAPs) cofactor swap, 529 electron transport across the membrane from FAD to heme, 527 ferric reductase (FRD) superfamily, 529 hydride transfer between NADPH and FAD, 527 monoclonal antibody, 528 outer sphere reduction mechanism, 527, 529, 531 single heme binding, 524 STEAP1, 521–523, 525, 527–529, 531 STEAP2, 521–523, 525, 528 STEAP3, 521, 522, 524, 527, 528 STEAP4, 521–531 Small angle X-ray scattering (SAXS), 45, 295 Small interfering RNA (siRNA), 184, 191, 192, 194, 195, 221, 239, 313, 318, 390 Src homology region 3 domain, 28, 167, 251, 292, 314, 380, 435, 545 Superoxide (O2 -) dismutation, 7, 166, 263, 324, 330, 331, 344, 517, 538 production, 6–8, 11, 12, 14–19, 22–24, 28, 30, 32, 33, 39, 44, 45, 48, 446, 453 Superoxide dismutase (SOD), 7, 67–70, 72, 81, 92, 120, 137, 141, 166, 222, 249, 313, 316, 319, 324–327, 332, 335, 408, 416, 418, 445, 452, 454, 470, 480, 538, 543, 544

T Tks4, 128, 129, 345, 359, 382, 383 Tks5, 128, 129, 382, 383 Toll-like receptor (TLR) TLR2, 152 TLR4, 152, 158, 182, 238, 290, 318 TLR5, 152, 189, 238 TLR7, 281, 541 Translocation of cytosolic components, 8, 14, 26, 28, 32, 33, 36, 41, 42, 48, 49, 83 to the membrane, 22, 27, 31, 42, 151, 296, 406, 436 of p47phox, 32, 39, 296, 359, 406, 436 of p67phox, 34, 39, 265, 296, 349, 359, 406 of RAC, 31, 296 Treatments for chronic granulomatous disease (CGD) gene therapy adeno-associated viruses, 549, 559, 568 CRISPR/Cas9 platform, 561, 566 ex vivo gene therapy, 560, 566 gene editing, 561, 562, 566, 568 hematopoietic stem and progenitor cells (HSPCs), 561, 566 in vivo gene therapy, 559 lentiviral vectors, 549, 566, 568 retroviral vectors, 549, 560, 564 hematopoietic stem cell transplantation (HSCT) bone marrow transplantation, 563 conditioning regimen, 562, 568

Index graft-versus-host disease, 541, 549, 558 human leukocytes antigen (HLA), 558 matched donors, 559 Trimera, 43, 44, 47 Triton-X 100, 72 Tumor necrosis factor alpha (TNF-α), 182, 252, 313, 315

U Ubiquinone, 10, 94, 95

V Vignais, P.V., 11, 15, 27, 30, 119–121

W Wiskott-Aldrich syndrome protein (WASP), 34, 294, 297, 298, 301, 410, 421, 422, 425, 426

X X-ray crystallography, 30, 34, 35, 37, 38, 347, 358, 365, 524 X-ray structure, 40, 253

593 Y Yeast NADPH oxidase Yno1 actin cytoskeleton, 414, 416, 420–422 Candida albicans (C. albicans), 418, 435 Cdc24, 268, 270, 271, 421, 436, 438, 440 Cdc42, 392 FRE, 415 shmoos, 414, 423 superoxide dismutase (Sod1), 92, 416

Z Zebrafish NADPH Oxidases biological functions antimicrobial defense, 494 circadian rhythm, 499 development, 494 regeneration, 496–498 wound healing, 496–498 duoxA gene, 493 nox1, nox2, nox5 and duox isoforms, 493, 494, 496 p22phox, p40phox, p47phox, p67phox, Noxo1 and Noxa1 and several rac genes, 493 tumorigenesis, 498–499 Zymosan, 8, 13, 24, 72, 94, 251, 278, 279, 539, 544