Ribosome-inactivating Proteins: Ricin and Related Proteins 1118125657, 9781118125656

Citation preview

Ribosomeinactivating Proteins Ricin and Related Proteins

• • • • •

Edited by:

Fiorenzo Stirpe, Douglas A. Lappi

Ribosome-inactivating Proteins

Ribosome-inactivating Proteins Ricin and Related Proteins

Edited by FIORENZO STIRPE Dipartimento di Medicina Specialistica, Diagnostica e Sperimentale Alma Mater Studiorum Università di Bologna Bologna, Italy

DOUGLAS A. LAPPI Advanced Targeting Systems, Inc. San Diego, California, USA

This edition first published 2014 © 2014 by John Wiley & Sons, Inc. Editorial Offices 1606 Golden Aspen Drive, Suites 103 and 104, Ames, Iowa 50010, USA The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 9600 Garsington Road, Oxford, OX4 2DQ, UK For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Blackwell Publishing, provided that the base fee is paid directly to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license by CCC, a separate system of payments has been arranged. The fee codes for users of the Transactional Reporting Service are ISBN-13: 978-0-8138-2083-5/2014. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. Limit of Liability/Disclaimer of Warranty: While the publisher and author(s) have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. It is sold on the understanding that the publisher is not engaged in rendering professional services and neither the publisher nor the author shall be liable for damages arising herefrom. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Manuscript no. 14-068-B from the Kansas Agricultural Experiment Station, Manhattan. Library of Congress Cataloging-in-Publication Data Ribosome-inactivating proteins: biology and applications / edited by Fiorenzo Stirpe, Douglas A. Lappi. pages cm Includes bibliographical references and index. ISBN 978-1-118-12565-6 (hardback) 1. Proteins–Synthesis. 2. Ribosomes–Structure. I. Stirpe, Fiorenzo, editor of compilation. II. Lappi, Douglas A., editor of compilation. QP551.R522 2014 571.6′58–dc23 2014002677 A catalogue record for this book is available from the British Library. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Cover image: Upper RH illustration: Ricinus communis L: Koehler’s Medizinal-Pflanzen (1887) Lower RH illustration: Saponaria officinalis L: English Botany, or Coloured Figures of British Plants, 3th ed. [J.E. Sowerby et al], vol. 3 (1864) Upper LH and lower RH illustrations: designed by Dr. Valeria Severino (Seconda Università degli Studi di Napoli) Cover design by Soephian Zainal Set in 10.5/12pt Times by SPi Publisher Services, Pondicherry, India

1

2014

Contents

Contributors Preface 1

Introduction and History Fiorenzo Stirpe

2

Occurrence and Taxonomical Distribution of Ribosome-inactivating Proteins Belonging to the Ricin/Shiga Toxin Superfamily Chenjing Shang, Willy J. Peumans, and Els J. M. Van Damme

3

Ribosome-inactivating Proteins from Phytolaccaceae Augusto Parente, Angela Chambery, Antimo Di Maro, Rosita Russo, and Valeria Severino

4

Ribosome-inactivating Proteins in Caryophyllaceae, Cucurbitaceae, and Euphorbiaceae Tzi Bun Ng and Jack Ho Wong

vii xi 1

11 28

44

5

Non-toxic Type 2 Ribosome-inactivating Proteins Pilar Jiménez, Manuel José Gayoso, and Tomás Girbés

67

6

The Intracellular Journey of Type 2 Ribosome-inactivating Proteins Robert A. Spooner and J. Michael Lord

83

7

Shiga Toxins: The Ribosome-inactivating Proteins from Pathogenic Bacteria Maurizio Brigotti

97

8

The Structure and Action of Ribosome-inactivating Proteins Jon D. Robertus and Arthur F. Monzingo

111

9

Updated Model of the Molecular Evolution of RIP Genes Willy J Peumans, Chenjing Shang, and Els J. M. Van Damme

134

10

Enzymology of the Ribosome-inactivating Proteins Yaeta Endo

151

11

A Long Journey to the Cytosol: What do We Know about Entry of Type 1 RIPs Inside a Mammalian Cell? Rodolfo Ippoliti and Maria Serena Fabbrini

161

v

vi

12

CONTENTS

Ribosome-inactivating Proteins: Pathology from Cells to Organs Gareth D. Griffiths

178

13

Antiviral and Antifungal Properties of RIPs Gabriela Krivdova, Kira C. M. Neller, Bijal A. Parikh, and Katalin A. Hudak

198

14

Insecticidal and Antifungal Activities of Ribosome-inactivating Proteins Lúcia Rosane Bertholdo Vargas and Célia Regina Carlini

212

15

Immunology of RIPs and their Immunotoxins Giulio Fracasso and Marco Colombatti

223

16

Ribosome-inactivating Proteins in Cancer Treatment Douglas A. Lappi and Fiorenzo Stirpe

244

17

Nervous System Research with RIP Conjugates: From Determination of Function to Therapy Douglas A. Lappi, Jack Feldman, Dale Sengelaub, and Jill McGaughy

253

Embryotoxic and Abortifacient Activities of Ribosome-inactivating Proteins Wood Yee Chan, Jack Ho Wong, and Tzi Bun Ng

270

18

19

The Potential for Misuse of Ribosome-inactivating Proteins Gareth D. Griffiths

Index Color plates appear between pages 116 and 117

281

287

Contributors

Maurizio Brigotti

Dipartimento di Medicina Specialistica, Diagnostica e Sperimentale Alma Mater Studiorum Università di Bologna Bologna, Italy

Célia Regina Carlini

Centro de Biotecnologia–UFRGS Universidade Federal do Rio Grande do Sul Porto Alegre, Brazil

Angela Chambery

Department of Environmental, Biological and Pharmaceutical Sciences and Technologies Second University of Naples Caserta, Italy

Wood Yee Chan

School of Biomedical Sciences Faculty of Medicine The Chinese University of Hong Kong Hong Kong, China

Marco Colombatti

Department of Pathology and Diagnostics University of Verona Verona, Italy

Antimo Di Maro

Department of Environmental, Biological and Pharmaceutical Sciences and Technologies Second University of Naples Caserta, Italy

Yaeta Endo

Cell-Free Science and Technology Research Center Ehime University Ehime, Japan; Center for Molecular Biology of RNA University of California, Santa Cruz Santa Cruz, California, USA

Maria Serena Fabbrini

Ministry of Instruction, University and Research (MIUR) Monza, Italy

vii

viii

CONTRIBUTORS

Jack Feldman

Systems Neurobiology Laboratory University of California, Los Angeles Los Angeles, California, USA

Giulio Fracasso

Department of Pathology and Diagnostics University of Verona Verona, Italy

Manuel José Gayoso

Departamento de Farmacología, Biologíca Celular e Histología Facultad de Medicina Universidad de Valladolid Valladolid, Spain

Tomás Girbés

Nutrición y Bromatologia Facultad de Medicina and Centro de Investigación en Nutrición, Alimentacióny Dietética CINAD-Parque Científico Universidad de Valladolid Valladolid, Spain

Gareth D. Griffiths

Cellular Toxicity Team Biology and Biomedical Sciences Defence Science & Technology Laboratory (DSTL) Porton Down Salisbury, UK

Katalin A. Hudak

Department of Biology York University Toronto, Ontario, Canada

Rodolfo Ippoliti

Department of Life, Health and Environmental Sciences University of L’Aquila L’Aquila, Italy

Pilar Jiménez

Nutrición y Bromatologia Facultad de Medicina and Centro de Investigación en Nutrición, Alimentacióny Dietética CINAD-Parque Científico Universidad de Valladolid Valladolid, Spain

Gabriela Krivdova

Department of Biology York University Toronto, Ontario, Canada

Douglas A. Lappi

Advanced Targeting Systems, Inc. San Diego, California, USA

J. Michael Lord

School of Life Sciences University of Warwick Coventry, UK

CONTRIBUTORS

Jill McGaughy

Department of Psychology University of New Hampshire Durham, New Hampshire, USA

Arthur F. Monzingo

Institute for Cellular and Molecular Biology University of Texas at Austin Austin, Texas, USA

Kira C. M. Neller

Department of Biology York University Toronto, Ontario, Canada

Tzi Bun Ng

School of Biomedical Sciences Faculty of Medicine The Chinese University of Hong Kong Hong Kong, China

Augusto Parente

Department of Environmental, Biological and Pharmaceutical Sciences and Technologies Second University of Naples Caserta, Italy

Bijal A. Parikh

Department of Pathology and Immunology Washington University School of Medicine St. Louis, Missouri, USA

Willy J. Peumans

Aalst. Belgium

Jon D. Robertus

Department of Molecular Biosciences Institute for Cellular and Molecular Biology University of Texas at Austin Austin, Texas, USA

Rosita Russo

Department of Environmental, Biological and Pharmaceutical Sciences and Technologies Second University of Naples Caserta, Italy

Dale Sengelaub

Department of Psychological and Brain Sciences, and Program in Neuroscience Indiana University Bloomington, Indiana, USA

Valeria Severino

Department of Environmental, Biological and Pharmaceutical Sciences and Technologies Second University of Naples Caserta, Italy

Chenjing Shang

Laboratory of Biochemistry and Glycobiology Department of Molecular Biotechnology Ghent University Ghent, Belgium

ix

x

CONTRIBUTORS

Robert A. Spooner

School of Life Sciences University of Warwick Coventry, UK

Fiorenzo Stirpe

Dipartimento di Medicina Specialistica, Diagnostica e Sperimentale Alma Mater Studiorum Università di Bologna Bologna, Italy

Els J. M. Van Damme

Laboratory of Biochemistry and Glycobiology Department of Molecular Biotechnology Ghent University Ghent, Belgium

Lúcia Rosane Bertholdo Vargas

Instituto de Biotecnologia Universidade de Caxias do Sul Caxias do Sul, Brazil

Jack Ho Wong

School of Biomedical Sciences Faculty of Medicine The Chinese University of Hong Kong Hong Kong, China

Preface

Ribosome-inactivating proteins (RIPs) are a class of proteins that range from a few proteins known for more than a century, to a large number identified in the last few years. Some of them are potent toxins. An abundant literature has appeared on the subject, with thousands of articles, reviews, and books. Research on RIPs has been stimulated not only for the sake of knowledge, but also for their potential applications, at first in medicine and subsequently in agriculture, some of which now seem to be close to use. In spite of this significant amount of research, a lot remains to be learned. Many questions remain unanswered, and new ones are posed by results obtained. To mention just one example, the role of RIPs in nature is still unclear. These, in part unexpected, developments led us to plan a book on these proteins. We were fortunate enough to obtain the collaboration of some of the best experts on the various aspects of RIPs, who have well described the research on these proteins. They had absolute freedom, not only in reviewing the literature, but also in expressing their views and making new proposals, even when these were different from the opinions of other authors and, at times, even the editors. This, we hope, makes the book not only informative, but also a stimulus for further research. The book is organized in 19 chapters, each assigned to relevant experts. It starts with an introduction summarizing the research that led from ricin to a new class of proteins and their possible practical applications, followed by a description of the occurrence and distribution of RIPs in nature, based on a modern original search on the genome of these proteins and their presence in the available genomes of plants and animals. Almost all the type 1 RIPs are described, divided by plant families of origin – namely Phytolaccaceae, Caryophyllaceae, Cucurbitaceae, and Euphorbiaceae – followed by the non-toxic type 2 RIPs and the Shiga and Shiga-like toxins. The properties of RIPs are extensively described, beginning with their structures, which are compared and related to their enzymatic activity and to the action of inhibitors. This is followed by the evolution of RIP genes. The true toxins enter cells by the clever and insidious use of a cell-binding protein; their traverse into the cell and to their target is a tale of incredible evolutionary prowess. Even RIPs with no cell-binding chain apparently enter cells for antiviral activity. The enzymatic action, the entry into cells, and the intracellular destination of RIPs are fully described. The pathological damage caused by toxic RIPs is also well discussed. The antiviral, antifungal, embryotoxic, and abortifacient properties of RIPs lead to possible applications of these proteins in agriculture and in medicine, as they are or linked to antibodies or other carriers, with limitations due to their immunological properties which are well described and discussed.

xi

xii

PREFACE

RIPs conjugated to targeting proteins, such as antibodies that recognize cancer cell surface markers, have held promise over the years as miraculous anticancer agents if they could just be targeted specifically to the cancer cells and nothing else. The adventures and difficulties in developing a proper drug in which the marvels of modern science can be used is the subject of Chapter 16. This topic has been worked on for 40 years now and can, and should, be the subject of a completely separate book. The use of this same idea has been transferred to neuroscience research, described in Chapter 17, as scientists have begun the work on the Brain Activity Map project. Finally, fears over the possible uses of toxic RIPs for criminal purposes and as biological weapons for warfare and terroristic attacks are summarized in Chapter 19 from a realistic viewpoint. We are grateful to all our authors for having accepted our proposal and our comments on their work. These authors are leaders in their fields and have contributed their results, in several cases, for many years in the best peer-reviewed scientific publications. We, the editors, are proud that they have joined us in describing the many fascinating facets of ribosome-inactivating proteins. We also thank Wiley for their help in all aspects of this effort, including a most important aspect: publishing. We especially thank Denise Higgins for organizing and formatting all the chapters so that they all made sense. This was no small task and, without her help, this book would never have made it to the publishers. All of us, authors and editors, are tremendously grateful for her dedication.

1

Introduction and History Fiorenzo Stirpe Dipartimento di Medicina Specialistica, Università di Bologna, Italy

Introduction

The history of RIPs, especially the toxic ones, has been well reviewed recently.1 This present chapter will summarize the research steps that in the last 40 years have led to significant advancements in the knowledge of these proteins, of their mechanism of action, and of their possible practical applications in medicine and in agriculture. Ribosome-inactivating proteins (RIPs), initially discovered in higher plants, have been the subject of numerous studies (reviews by Van Damme,2 Nielsen,3 Hartley,4 Girbés,5 Stirpe,6, 7 Ng,8 and Puri9). More than 50 RIPs have been identified and purified, but it has become clear that they can, in some circumstances, be expressed in many plants and other organisms in which they have not been detected because of assay sensitivity or other reasons. Thus, they must have an important function to justify their persistence throughout the evolution of proteins, which are an expensive material to make. Furthermore, it is becoming more and more apparent that important uses of RIPs can be envisaged. Identification and Distribution in Nature

The studies on the proteins that eventually were denominated ribosome-inactivating proteins (RIPs) began in Dorpat at the end of the nineteenth century when ricin, a potent toxin from the seeds of Ricinus communis (castor bean plant), was identified and isolated by Stillmark who described it in his thesis as a “ferment” (remarkably for the time!).10 Shortly afterwards, abrin, a toxin similar to ricin, was isolated from the seeds of Abrus precatorius.11 Research on these toxins was then rather scarce for a long period, until a revival of the studies on ricin and abrin was prompted by the report that these proteins were more toxic to malignant, than to normal, cells.12 The same authors found that the toxins inhibited protein synthesis by cells, a first step toward the discovery of their mechanism of action.13 Unfortunately, subsequent interest in the toxins, especially ricin, was stimulated also by the fear of a possible use for warfare or terrorist actions, as in the case of a Bulgarian journalist murdered with a micro-bullet probably containing ricin.14 Ribosome-inactivating Proteins: Ricin and Related Proteins, First Edition. Edited by Fiorenzo Stirpe and Douglas A. Lappi. © 2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.

1

2

RIBOSOME-INACTIVATING PROTEINS

Significant, pioneering progress in knowledge of the toxins was achieved by S. Olsnes, A. Pihl, and collaborators in the Institute of Biochemistry of the Norwegian Radium Hospital. They elucidated the structure of both ricin and abrin, establishing that they constituted two unequal polypeptide chains linked by a disulfide bond, an A chain which inhibited protein synthesis, and a B chain with the properties of a lectin specific for sugars with the structure of galactose. They found also that the latter chain binds to galactose residues on the surface of most cells, allowing the entry of the A chain, which exerts its toxic action.15 Furthermore, they confirmed that the toxins inhibit protein synthesis, not only in cells, but also in a cell-free system.16 The knowledge that ricin and abrin were produced by taxonomically unrelated plants, and still had a very similar structure and function, led our group to research whether other, similar, toxins were present in nature. In our laboratory this research was conducted by examining seed extracts from plants known to be toxic. On the basis of Olsnes’ observations, the extracts were tested for inhibition of protein synthesis in a cell-free system, a rabbit reticulocyte lysate, a much simpler and more rapid test to perform than to evaluate the effects on inhibition of protein synthesis by, or on viability of, cells. A number of seed extracts were screened in this way, starting from those from plants that in the old literature were reported as containing toxins similar to ricin, such as crotin and curcin.17 Much to our surprise (and for a while, disappointment!), we found that indeed in the seed extracts from many plants there were proteins that inhibited cell-free protein synthesis, but these were hardly toxic to cells.18 The meaning of these observations was elucidated when an antiviral protein (pokeweed antiviral protein, PAP) was isolated from the leaves of Phytolacca americana, that inhibited protein synthesis resembling the A chain of ricin.19, 20 Thus, it soon became clear that the proteins present in our and other plant extracts also had antiviral activity against both plant and animal viruses (review by Kaur et al.21) and were similar to the A chain of ricin. This led to classifying the ribosome-inactivating proteins into two types: type 2, consisting of an A and a B chain; and type 1, consisting of an A-like single polypeptide chain with enzymatic activity.22 Type 3 RIPs, consisting of an active moiety linked to a peptidic chain with unknown function, have also been described (reviewed in Peumans et al.23). A new classification of RIPs is proposed in Chapter 2. Another form of PAP was identified in pokeweed summer leaves,24 and another in pokeweed seeds.22 On the whole, these results indicated (i) that RIPs could be present in the same plant as isoforms, which subsequent research showed could be several, and (ii) that in some plants RIPs could be in different organs, and not only in seeds as ricin and abrin. It was found also that trichosanthin, a protein from Trichosanthes kirilowii and other plant proteins used in China to induce abortion, were RIPs25 (reviews by Ng et al.26 and Chapter 4). The structure of ricin, and subsequently of other RIPs of both types, was elucidated by crystallographic studies (review by Robertus and Monzingo27 and Chapter 8), which led to the identification of active and sugar-binding sites. The search for other toxins similar to ricin continued, and modeccin, a highly toxic protein described as a “toxalbumin”28 and already at that time suspected to be similar to ricin, was purified and characterized almost at the same time in Oslo29 and in Bologna.30 A galactose-specific haemagglutinating lectin purified from mistletoe31 turned out to be a type 2 RIP.32 Other type 2 RIPs were found in several other plants,5 including some belonging to the genus Adenia.33 From Adenia stenodactyla stenodactylin was isolated, probably the most potent toxin of plant origin, with an LD50 30 [Au]; 3 [AC]; 1 [AP]

Grasses

Zea mays*: 7 [Au]; 2 [AB]; 1 [AC]; 1 [AD] Sorghum bicolor*: 14 [Au]; 1 [AB] Setaria italica*: >5 [Au]; 2 [AC] Amborella trichopoda: No RIP gene Selaginella moellendorffii: No RIP gene Physcomitrella patens: No RIP gene Chlamydomonas reinhardii: No RIP gene Volvox carteri: No RIP gene

Figure 2.1 Schematic overview of the presence/absence of RIP genes in the currently completed plant genomes. The dendrogram reflects only the overall phylogeny of the species listed. The presence or absence of RIP genes is indicated. *denotes preliminary results.

16

RIBOSOME-INACTIVATING PROTEINS

Parent type: type [AB]

Classical type 2 RIP

[AB] lineage Offspring type: type [AΔB]

Vacuolar form (common)

Classical type 1 RIP

Cytoplasmic form (rare)

Malus domestica EP

Parent type: type [AX]

Cannabis sativa EP

[AX] lineage Offspring type: type [AΔx]

Cytoplasmic form (common)

Fagus sylvatica EP

Vacuolar form (rare)

Muscari armeniacum type 1 RIP

Parent type: type [AC] [AC] lineage

?

Hordeum vulgare JIP 60

Offspring type: type [AΔc]?

Possibly some poaceae type [Au] RIP

Parent type: type [AD] [AD] lineage

?

Zea mays EP

Offspring type: type [AΔD]?

Possibly some poaceae type [Au] RIP

[APM41] lineage

Oryza sativa EP

[APc19] lineage

Triticum aestivum EP

Signal peptide

EP=expressed protein Figure 2.2

RIP domain

X domain

C domain

D domain

Peptidase C19 domain

Schematic overview of the domain architectures identified in plant RIP genes.

Peptidase M41 domain

Magnoliids Angiosperms

Commelinids

Monocots

Eudicots

Fabids

Rosids Core eudicots

Asterids

Lamiids Campanulids

Malvids

Amborellales Nymphaeales Austrobaileyales Piperales Canellales Magnoliales [AB] Chloranthales Commelinales Zingiberales Poales [Au], [AB], [AC], [AD], [AP] Arecales [AΔX], [AB] Dasypogonaceae Asparagales [AΔB], [AΔX] Liliales Pandanales Dioscoreales Petrosaviales Alismatales Acorales Ceratophyllales Ranunculales [AΔB], [AB] Sabiaceae Proteales Buxales Trochodendrales Gunnerales Cucurbitales [AΔB], [AB], [ΔAB] Fagales [AΔX] Rosales [AΔB], [AΔX], [AB] Fabales [AB], [AX] Celastrales Oxalidales Malpighiales [AΔB], [AB], [AX] Zygophyllales Malvales [AΔX], [AB], [AX] Brassicales Huerteales Sapindales [AB] Picramniales Crossosomatales Myrtales Geraniales Vitales Saxifragales Dilleniaceae Berberidopsidales Santalales [AB] Caryophyllales [AΔB], [ΔX] Cornales [AB] Ericales [AΔB], [AB] Garryales Gentianales Lamiales [AΔB] Solanales Boraginaceae Aquifoliales Escalloniales Asterales [AB] Dipsacales [AB], [ΔAB] Paracryphiales Apiales [AΔX], [AB], [AX] Bruniales

Figure 2.3 Schematic overview of the documented occurrence of (expressed) RIP genes within the major taxa of Angiosperms. The dendrogram (based on APG III) reflects only the overall phylogeny of the taxa listed.

18

RIBOSOME-INACTIVATING PROTEINS

Eudicots (+)(3) Monocots (+)(3) Magnoliids (+)(2) Ceratophyllales (–) Chloranthales (–) Austrobaileyales (–) Nymphaeales (–) Amborellales (–) Coniferophyta (–) Gnetophyta (–) Ginkgophyta (+)(1) Cycadophyta (–) Polypodiopsida (–) Equisetopsida (–) Marratiopsida (–) Ophioglossopsida (–) Psilotopsida (–) Lycopodiophyta (–) Bryophyta (–) Marchantiophyta (–) Anthocerophyta (–) Chlorophyta (–) Figure 2.4 Schematic overview of the documented occurrence of (expressed) RIP genes within the major taxa of Viridiplantae. The dendrogram reflects only the overall phylogeny of the taxa listed (based on Palmer et al.).15 (1) Refers to Gnetum gnemon, containing two [AB] RIPs. (2) refers to Cinnamomum camphora, containing three [AB] RIPs. (3) For details on the occurrence of RIP genes in Eudicots and Monocots, we refer to Figure 2.3.

the latter taxon, a summary of the results of the transcriptome screenings is superimposed on the phylogenetic tree of the Angiosperm Phylogeny Group III system (APG III, 2009), which is the most recent version of a modern, primarily molecular-based, system of plant taxonomy. Further details about each RIP-positive group are given in Tables 2.1 and 2.2, listing approximately 100 non-grass and over 20 grass species. Though illustrative, the results of the present overview have to be interpreted with care because they are highly biased by the availability of sequence information. No conclusion can be drawn, indeed, neither with respect to the possible absence of RIP genes from a particular species, family, or higher order taxon nor for what concerns the apparent frequent occurrence in some taxonomic groups (e.g., Caryophyllaceae, Cucurbitaceae). What can be concluded with certainty is that RIP genes (i) occur over a wide taxonomic range of flowering plants and (ii) are particularly prominent (or possibly even ubiquitous) in Poaceae. Moreover, it appears that the genome of most grasses possesses a fairly complex family of RIP genes, some of which might be confined to the Poaceae.15

OCCURRENCE AND TAXONOMICAL DISTRIBUTION OF RIBOSOME-INACTIVATING PROTEINS

19

Table 2.1 Detailed overview of the documented taxonomic distribution of the different types of RIPs within the Magnoliids and Monocot divisions of the Angiosperms (flowering plants). Taxa are ordered according to the dendrogram shown in Figure 2.3. MAGNOLIIDS Laurales

Lauraceae

Cinnamomum camphora

3 [AB]

MONOCOTS Commelinids Arecales

Arecaceae

Elaeis guineensis* Phoenix dactylifera

[AB] [AΔX], [ΔAX]

Commelinids Poales

Bromeliaceae Poaceae

[Au] [Au], [AB] >30 [Au], 3 [AC], [AP] [Au] [Au], [AC] 15 [Au], [AC], [AP] [Au] [Au] [Au] >20 [Au], [AC] [Au] Multiple [Au] Multiple [Au] >20 [Au], [AB], [AP] [Au], [AB] 14 [Au], [AB] 7 [Au], 2 [AB], [AC], AD] [AB], [Au], [ΔAC] >5 [Au], 2 [AC]

Asparagales

Agavaceae Asparagaceae Hyacinthaceae

Ananas comosus B# Phyllostachys edulis E Oryza sativa* P Agrostis capillaries Avena barbata Brachypodium distachyon* Festuca arundinacea Festuca pratensis Aegilops speltoides Hordeum vulgare Leymus cinereus Pseudoroegneria spicata Secale cereale Triticum aestivum P# Saccharum officinarum A Sorghum bicolor* C Zea mays* C Panicum virgatum A Setaria italica* D Yucca filamentosa Asparagus officinalis Charybdis maritime Drimiopsis kirkii Hyacinthus orientalis Muscari armeniacum Iris brevicaulis Iris fulva Iris hollandica Ophiogon japonicus Polygonatum multiflorum

Iridaceae

Ruscaceae

2 [AΔX] 8 [AΔX] [AΔX] [AΔX] [AΔX] [AΔX] 3 [AΔB] 2 [AΔB] 3 [AΔB], 2 [AB] [AΔX] 2 [AB]

*Completed genomes #BEP clade (Bambusoideae, Ehrhartoideae and Pooideae); PACCAD clade (Panicoideae, Aristidoideae, Centothecoideae, Chloridoideae, Arundinoideae, Danthonioideae)

Bacterial RIPs – Shiga Toxin Group Genuine Shiga and Shiga-Like Toxins

The best known and, until recently the only identified, group of bacterial RIPs are the so-called Shiga and Shiga-like toxins. The Shiga toxin itself is produced and secreted by Shigella dysenteriae. Nearly identical toxins were also found in some Escherichia coli strains (more specifically in the Shiga toxigenic group of E. coli (STEC)). All members of the Shiga toxin family have the same

20

RIBOSOME-INACTIVATING PROTEINS

Table 2.2 Detailed overview of the documented taxonomic distribution of the different types of RIPs within the Eudicotyledons of the Angiosperms (flowering plants). Taxa are ordered according to the dendrogram shown in Figure 2.3. EUDICOTYLEDONS RANUNCULALES

C O R

E E U D I C O T Y L E D O N S

R O S I D S

F A B I D S

Ranunculaceae

Menispermaceae Cucurbitales Cucurbitaceae

Fagales Rosales

Fagaceae Cannabaceae Rosaceae

M A L V I D S

Fabales

Fabaceae

Malpigiales

Euphorbiaceae

Malvales

Passifloraceae Saliceae Malvaceae

Sapindales

Sapindaceae

Actaea racemosa Adonis aestivalis Aquilegia coerulea* Eranthis hyemalis Cissampelos mucronata Bryonia dioica Citrullus lanatus* Cucumis figarei Cucumis melo* Cucumis sativus* Gynostemma pentaphyllum Luffa cylindrica Momordica balsemina Momordica charantia Siraitia grosvenorii Trichosanthes cordata Trichosanthes cucumerina Trichosanthes dioica Trichosanthes kirilowii

3 [AΔB] [AB] [AB] [AB] [AB] [AΔB], [AB] 2 [AΔB], 2 [AB]pseudo [AΔB] 3 [AB], 3 [ΔAB] 2 [AB], 3 [ΔAB] 4 [AΔB] 2 [AΔB] [AΔB] 2 [AΔB] [AB] ≥3 [AΔB], [ΔAB] or [AB]? 2 [AB] [AΔB] [AB] 4 [AΔB], [AB]

Fagus sylvatica Cannabis sativa* Humulus lupulus Malus domestica* Prunus armeniaca Prunus mume Prunus persica* Abrus precatorius Abrus pulchellus Acacia mangium Euphorbia esula Euphorbia characias Euphorbia serrata Gelonium multiflorum Jatropha curcas* Manihot esculenta* Ricinus communis* Vernicia fordi Adenia volkensii Populus trichocarpa* Gossypium hirsutum Gossypium raimondii* Theobroma cacao*

8 [AΔX] 7 [AΔB], 2 [AΔX], [AX] [AΔX] 3 [AΔB], [ΔAB] ≥1 [AΔB] ≥ 2[AΔB] 4 [AΔB] 5 [AB] 4 [AB] [AX] 9 [AΔB], 5 [ΔAB] [ΔAB] 2 [AΔB] [AΔB] 7 [AΔB] 4 [AΔB] 7 [AΔB], 8 [AB] [AΔB] [AB] [AΔX], 3 [ΔAB] [AB] [AB] 2 [AΔX] (or truncated [AX]), 3 [AX], [ΔAX] [AB]

Paullinia cupana

OCCURRENCE AND TAXONOMICAL DISTRIBUTION OF RIBOSOME-INACTIVATING PROTEINS

Table 2.2

(Continued )

A S T E R I D S

Santalales

Caryophyllales

Olaceae

Ximenia americana

[AB]

Santalaceae

Viscum album Viscum articulatum Mesembryanthemum crystallinum Amaranthus tricolor Amaranthus viridis Atriplex patens Beta vulgaris Celosia cristata Chenopodium album Chenopodium quinoa Spinacia oleracea Dianthus caryophyllus Dianthis chinensis Saponaria officinalis Silene latifolia Stellaria media Bougainvillea x buttiana Bougainvillea spectabilis Mirabilis jalapa Mirabilis expansa Phytolacca americana Phytolacca acinosa Phytolacca heterotepala Phytolacca insularis Phytolacca octandra Hydrangea macrophylla Actinidia deliciosa Camellia sinensis Diospyros kaki Ipomopsis aggregata Clerodendron inerme Clerodendron aculeatum

10 [AB] [AB] 5 [AΔB]

Artemisia annua Centaurea maculosa Centaurea solstitialis Chrysanthemum x morifolium Helianthus tuberosus Parthenium argentatum Sambucus nigra Sambucus ebulus Sambucus sieboldiana Panax ginseng Bupleurum chinense

[AB] [AB] [AB] [AB] [AB] [AB] Multiple [AB] and [ΔAB] Multiple [AB] and [ΔAB] Multiple [AB] and [ΔAB] [AB], [AX], [ΔAX] Most probably [AΔB]

Aizoaceae Amaranthaceae

Caryophyllaceae

Nyctaginaceae

Phytolaccaceae

L A M I I D S C A M P A N U L I D S *Completed genomes

Cornales Ericales

Hydrangaceae Actinidiaceae Theaceae Ebenaceae Polemoniaceae Lamiaceae

Asterales

Asteraceae

Dipsacales

Adoxaceae

Apiales

Araliaceae Apiaceae

2 [AΔB] 4 [AΔB] [AΔB] 6 [AΔB], [AΔX] [AΔB] [AΔB] [AΔB] 2 [AΔB] [AΔB] 3 [AΔB] 4 [AΔB] 5 [AΔB] [AΔB] 3 [AΔB] 3 [AΔB] [AΔB] [AΔB] 6 [AΔB] [AΔB] [AΔB] 2 [AΔB] [AΔB] [AB] [AΔB], [AB] [AΔB], 2 or 3 [AB], 3 [ΔAX] [AΔB] or [AB]? [AB] [AΔB] [AΔB]

21

22

RIBOSOME-INACTIVATING PROTEINS

Genes with a Shiga-toxin A domain Genuine Shiga-toxin genes (discistronic) Enterobacter cloacae Citrobacter freundii Escherichia coli Shigella sonnei Shigella dysenteriae

RIP [As]-Bs Shiga and shiga-like toxins Genes with a Shiga-toxin A domain only RIP [As]

Rickettsiella grylli Brukholderia sp. CCGE1002

Protein homolog of Stx A subunit

Genes with a domain distantly to the Shiga-toxin A domain

RIP [As]-U: dicistronic Non-secreted

Burkholderia ambifaria

RIP [As]-P-W: tricistronic Non-secreted

Flavobacterium columnare

Genes with an A domain resembling plant type A RIP Streptomyces secreted type A RIP

Streptomyces coelicolor Streptomyces lividans TK24 Streptomyces scabiei Streptomyces somaliensis

Streptomyces non-secreted type A RIP

Streptomyces lysosuperificus Streptomyces sp. Mg1

RIP [A]

Micromonospora sp. ATCC 39149 Chimeric protein

RIP [AM] Monocistronic

Signal peptide

Shiga toxin A or related domain Domain resembling plant RIP Unknown domains

Part of Pasteurella ToxA domain Figure 2.5

Schematic overview of the domain architectures identified in bacterial RIP genes.

canonical AB5 molecular structure. Basically, an A subunit (equivalent to a RIP domain) with catalytic activity forms through non-covalent interactions a complex with a so-called B subunit, which itself is a pentamer of five identical smaller polypeptides (Figure  2.5) that interact with specific glycolipids like globotriaosylceramide (Gb3).2 Though the term AB5, as well as the underlying overall structure of the Shiga toxins, might be reminiscent of the canonical AB structure of ricin and other type 2 RIPs from plants, there are three fundamental differences. First, there is no evolutionary relationship between the B chain of ricin and the polypeptide constituting the B subunit of the Shiga toxin. Second, the AB5 structure of the Shiga toxin results from non-covalent interactions between the A and B subunits, whereas in ricin the A and B chains are covalently linked

OCCURRENCE AND TAXONOMICAL DISTRIBUTION OF RIBOSOME-INACTIVATING PROTEINS

Table 2.3

23

Overview of documented domain architectures of RIP genes.*

PLANTS [AΔB]** [AΔx] [Au] [AB] [AC] [AD] [AX] [APM41] [APC19]

Type [A] derived from type [AB] RIP (most classical type 1 RIPs) Type [A] derived from type [AX] RIP (some classical type 1 RIPs; e.g. muscarin) Type [A] from undefined origin (e.g. maize RIP b-32; described as type 3 RIP) Chimer with C-terminal ricin-B domain (type 2 RIPs) Chimer with C-terminal (unknown) C domain (JIP 60; type 3 RIP) Chimer with C-terminal (unknown) D domain Chimer with C-terminal (unknown) X domain Chimer with C-terminal peptidase M41domain Chimer with C-terminal peptidase C19 domain

FUNGI [AΔxf] [AXf]

Fungal homolog of plant [AΔx] RIP Fungal homolog of chimeric plant [AX] RIP

INSECTS [AI]

Chimer with C-terminal (unknown) I domain

BACTERIA [A] [As] [As]-[Bs]° [As]-[U]° [As]-[P]-[W]° [AM]

Bacterial homolog of plant type [A] RIP (Streptomyces coelicolor RIPsc) Type [A] equivalent to Shiga toxin A-subunit Shiga and Shiga-like toxins Chimer with unknown C-terminal domain Chimer with PMT-C3 and unknown C-terminal domain Chimer with C-terminal (unknown) M domain

*For a schematic representation see Figures 2.2, 2.5, 2.6, and 2.7 **Domain (s) between square brackets are synthesized on a single mRNA °Polycistronic gene

through an interchain disulfide bond. Third, the A and B chains of ricin are synthesized on a single precursor polypeptide that undergoes a complex post-translational processing, whereas the A and B chains of the Shiga toxins are synthesized as two separate polypeptides on a single discistronic mRNA. Genuine Shiga and Shiga-like toxins and/or their genes have been identified in five different bacterial species, all of which are classified in the family Enterobacteriaceae (Gammaproteobacteria; Table  2.3). The genes encoding these toxins are presumably part of the genome of lambdoid prophages. Protein Homologs of Stx A Subunit

A (monocistronic) gene was identified in the genome of Rickettsiella grylli that resembles the genuine Stx genes but lacks the second cistron encoding the B polypeptide (Figure 2.5). Since no other gene occurs in the genome with a sequence encoding a B-polypeptide homolog, one can reasonably expect that Rickettsiella grylli does not produce a Shiga-like toxin. In principle Rickettsiella grylli can express a protein consisting only of an A chain. Alternatively, an as yet unidentified (toxic?) protein composed of an A chain homolog and another non-covalently bound subunit might be synthesized. A similar gene occurs in the Betaproteobacterium Burkholderia sp. CCGE1002, but in this species the RIP gene is located on plasmid pBC201. In this case also the nature and structure of the presumed RIP remain unclear.

24

RIBOSOME-INACTIVATING PROTEINS

Distantly Related Non-secreted Protein Homolog of Stx A Subunit

Proteins/genes distantly related to the Stx A subunit have been identified in two bacterial species. A Burkholderia ambifaria gene encodes a protein that clearly corresponds to a RIP domain and is most probably synthesized on a dicistronic mRNA. The latter fact might indicate that Burkholderia ambifaria synthesizes a novel type of toxin consisting of an A domain and an unknown noncovalently bound subunit(s) (Figure 2.5). The genome of Flavobacterium columnare (Bacteroidetes; Flavobacteriales) also possesses a gene with a RIP domain (Figure 2.5). This RIP domain is synthesized most probably on a tri-cistronic mRNA. Interestingly, the protein translated from the second cistron corresponds to the C-terminal domain of Pasteurella multocida ToxA (a bacterial protein toxin that modulates G proteins). This domain belongs to a family of transglutaminase-like proteins, with active site Cys–His–Asp catalytic triads.16 Unlike the proteins belonging to the Stx group described above, the Burkholderia ambifaria and Flavobacterium columnare RIPs are synthesized without a signal peptide and accordingly might be retained within the cytoplasm of the bacterial cells.

RIPs in Actinobacteria Streptomyces Coelicolor type [A] RIP

The production and biochemical characterization of the gene product SCO7092 (RIPsc) from the Gram-positive soil bacterium Streptomyces coelicolor expressed in E. coli demonstrated for the first time the occurrence of a bacterial RIP other than the classical Stx. Moreover, sequence comparisons revealed that RIPsc was more closely related evolutionarily to some plant type [A] RIPs than to the Stx A subunit (Figure 2.5).17 Besides S. coelicolor a nearly identical protein occurs in S. lividans, whereas similar proteins (46% and 41% sequence similarity, respectively) are found in S. scabei and S. somaliensis. All these RIPs are synthesized with a signal peptide and hence are likely to be secreted. Other Streptomyces RIPs

RIP sequences were retrieved in two additional Streptomyces species. The genome of Streptomyces lysosuperificus possesses a single gene encoding a protein with a RIP domain. Two similar genes located on different loci occur in the genome of Streptomyces sp. Mg1. All three (putative) RIPs are synthesized without a signal peptide and hence might be retained in the cytoplasm (Figure 2.5). Within the Actinobacteria a single RIP gene was identified outside the genus Streptomyces. Micromonospora sp. ATCC 39149 (Actinomycetales; Micromonosporaceae) possesses a gene encoding a chimeric RIP comprising an N-terminal A domain fused to a C-terminal domain of approximately 250 amino acid residues that hitherto has not been found in any known protein (Figure 2.5). This novel chimeric protein will further be referred to as a type [AM] RIP (A fused to unknown C-terminal domain).

Fungal RIPs Previously Described Fungal Ribosome-Inactivating Proteins: No Evidence for the Presence of an N-glycosidase Domain

Several reports claim the isolation and characterization of RIPs from fungi.9–11 However, until now no evidence has been presented that any of these presumed RIPs comprises a domain that is structurally and evolutionarily related to the A chains of either bacterial Stx or plant RIPs.

OCCURRENCE AND TAXONOMICAL DISTRIBUTION OF RIBOSOME-INACTIVATING PROTEINS

25

One group of fungal proteins/genes that is often referred to as RIPs belongs to the family of fungal ribotoxins. These ribotoxins are a family of extracellular ribonucleases that inactivate ribosomes by specifically cleaving a single phosphodiester bond located at the universally conserved sarcin/ricin loop of the large rRNA.12 Though ribotoxins inactivate ribosomes, both their catalytic activity and structure fundamentally differ from those of the RIPs with an N-glycosidase domain, and hence fall beyond the scope of this review. Besides ribotoxin homologs, a number of partly characterized proteins were described as fungal RIPs. Most of these proteins were isolated from mushrooms (e.g., Calvatia caelata, Flammulina velutipes, Hypsizigus marmoreus, Lyophyllum shimeji, and Pleurotus tuber-regium). Though the authors ascribe a protein synthesis inhibitory activity to all these proteins, they did not provide sufficient sequence information for formal identification. Evidently, this obvious lack of sequence information – in an era where scores of fungal genomes have been completely sequenced – can no longer be justified, and accordingly none of these “novel “ or “new” fungal RIPs can for the time being be considered members of the N-glycosidase family, and hence will not be further discussed here. In Silico Analyses Provide Evidence for the Presence of Genes with an N-glycosidase Domain in a Few Fungi

Screening of genome and transcriptome databases using some recently identified plant RIP genes as a query revealed the occurrence of two different types of genes with an N-glycosidase domain. A first type referred to as [AΔXf], that occurs in the genome of the nematode trapping fungus Arthrobotrys oligospora (Ascomycota) encodes a protein that consists of a single RIP domain and most probably can be considered a homolog of plant type [AΔX] RIP (Figure 2.6). No homologs/ orthologs of the A. oligospora RIP gene could be retrieved from any other sequenced fungal genome or transcriptome. The second type of fungal RIP gene encodes a chimeric protein that, based on its sequence, can be considered a fungal homolog of the plant [AX] chimers and accordingly will be referred to as [AXf]. Type [AXf] genes were identified in eight Ascomycota species, all of which belong to the Sordariomycetes (Table 2.4). Pseudogenes lacking an ORF due to the occurrence of frame shifts and/or stop codons were retrieved in two additional species of the same group. Two important conclusions can be drawn from Table 2.4. First, RIP genes are confined to a few fungal taxonomic groups. Second, the absence of a Magnaporthe poae [AXf] homolog from the genomes of M. grisea and M. oryzeae illustrates that even within the very same genus the [AXf] gene is not retained.

RIP [Axf]

Arthrobotrys oligospora

Cytoplasmic RIP [AXf] Vacuolar

[Axf] domain Figure 2.6

Epichloe sp. Neotyphodium gansuense Magnaporthe oryzeae Cordyceps sp.

[Xf] domain

Schematic overview of the domain architectures identified in fungal RIP genes.

26

Table 2.4

RIBOSOME-INACTIVATING PROTEINS

Summary of the occurrence and domain architecture of RIP genes in Bacteria.

P R O T E O B A C T E R A A C T I N O B A C T E R I A Bactero-idetes

Gammaproteobacteria

Enterobacteriales; Enterobacteriaceae

Betaproteobacteria

Legionellales; Coxiellaceae Burkholderiales; Burkholderiaceae

Actinobacteridae; Actinomycetales

Flavobacteriia; Flavobacteriales

Enterobacter cloacae Citrobacter freundii Escherichia coli Shigella sonnei Shigella dysenteriae Rickettsiella grylli

[As][Bs]*° [As][Bs] 2 [As][Bs] [As][Bs] [As][Bs] [As]

Burkholderia ambifaria MC40-6 Burkholderia sp CCGE1002

[As][U] [As]

Micromonosporineae; Micromonosporaceae Streptomycineae; Streptomycetaceae

Micromonospora sp. ATCC 39149

[AM]

Streptomyces coelicolor A3 Streptomyces lividans TK24 Streptomyces lysosuperificus Streptomyces scabiei Streptomyces somaliensis Streptomyces sp. Mg1

[A] [A] [A] [A] [A] [A]

Flavobacteriaceae

Flavobacterium columnare

[As][P] [W]

*Domain(s) between square brackets correspond to a single cistron. A sequence of two or three single domains between square brackets means that they correspond to a single polycistronic gene. °[As] and [Bs]: A and B domains identical or related to the A and B domains, respectively, of the Shiga toxins [U]; [W]: undefined domains

RIP [AI] Figure 2.7

Culex quinquefasciatus aedes aegypti

Schematic overview of the domain architectures identified in insect RIP genes.

Insect RIPs

Hitherto no RIP has been isolated from any animal species. However, in silico analyses of genome and transcriptome databases revealed that two mosquito species express genes encoding chimeric proteins comprising an N-terminal RIP domain fused to an unidentified C-terminal domain of approximately 200 amino acid residues. To distinguish this novel type of chimer from previously described RIPs it will further be referred to as the [AI] type, built up of an A domain and an “Insect specific” C-terminal domain (Figure 2.7). A single type [AI] RIP gene occurs in the genome of Culex quinquefasciatus (southern house mosquito), whereas three paralogs are found in Aedes aegypti (yellow fever mosquito). The presence of perfectly matching (partial) EST sequences indicates that all four insect RIP genes are expressed.

OCCURRENCE AND TAXONOMICAL DISTRIBUTION OF RIBOSOME-INACTIVATING PROTEINS

27

Conclusions

A renewed screening of the databases allowed the updating of the overall taxonomic distribution of the RIP domain and the domain architecture of the corresponding genes. In silico analyses revealed that the RIP domain is more widespread than was previously inferred from the classical biochemical and molecular approach, and has to be extended to fungi and insects. In addition, the identification of several novel chimeric forms implies that the heterogeneity of RIP genes in terms of domain architecture is no longer covered by the classical Shiga toxins and type 1, type 2, and type 3 plant RIPs, and argues for a novel classification system. Therefore one has to take into account that, as can be expected from the preliminary analyses of the RIP gene families in grass species, the present picture is still incomplete. Acknowledgments

This research was supported by the Research Council of Ghent University and the Fund for Scientific Research (FWO-Vlaanderen, Brussels, Belgium). Chenjing Shang acknowledges the receipt of a CSC Grant from the Chinese Government.

References 1. 2. 3. 4. 5. 6.

7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Stirpe F, Battelli MG. Ribosome-inactivating proteins: progress and problems. Cell Mol Life Sci. 2006;63:1850–1866. Johannes L, Römer W. Shiga toxins–from cell biology to biomedical applications. Nat Rev Microbiol. 2010;8:105–116. Barbieri L, Battelli MG, Stirpe F. Ribosome-inactivating proteins from plants. Biochim Biophys Acta. 1993;1154:237–282. Nielsen K, Boston RS. Ribosome-inactivating proteins: A plant perspective. Ann Rev Plant Physiolg. 2001;52:785–816. Van Damme EJM, Hao Q, Chen Y, et al. Ribosome-inactivating proteins: A family of plant proteins that do more than inactivate ribosomes. Crit Rev Plant Sci. 2001;20:395–465. Walsh TA, Morgan AE, Hey TD. Characterization and molecular cloning of a proenzyme form of a ribosome-inactivating protein from maize. Novel mechanism of proenzyme activation by proteolytic removal of a 2.8-kilodalton internal peptide segment. J Biol Chem. 1991;266:23422–23427. Mehta AD, Boston RS. Ribosome-inactivating proteins. In: Bailey-Serres J, Gallie DR, (eds.) A Look Beyond Transcription: Mechanisms Determining mRNA Stability and Translation in Plants. Rockville, MD: American Society of Plant Physiologists, 1998:145–152. Chaudhry B, Müller-Uri F, Cameron-Mills V, et al. The barley 60 kDa jasmonate-induced protein (JIP60) is a novel ribosome-inactivating protein. Plant J. 1994;6:815–824. Girbés T, Ferreras JM, Arias FJ, Stirpe F. Description, distribution, activity and phylogenetic relationship of ribosomeinactivating proteins in plants, fungi and bacteria. Mini Reviews in Medicinal Chemistry 2004;4:461–476. Reyes AG, Anné J, Meija A. Ribosome-inactivating proteins with an emphasis on bacterial RIPs and their potential medical applications. Future Microbiol. 2012;7:705–717. Stirpe F. Ribosome-inactivating proteins. Toxicon. 2004;44:371–383. Lacadena J, Alvarez-García E, Carreras-Sangrà N, et al. Fungal ribotoxins: molecular dissection of a family of natural killers. FEMS Microbiol Rev. 2007;31:212–237. Peumans WJ, Van Damme EJM. Evolution of plant ribosome-inactivating proteins. In: Lord JM, Hartley MR, (eds.) Toxic Plant Proteins Plant Cell Monographs Vol. 18. Berlin: Springer, 2010:1–26. Jiang SY, Ramamoorthy R, Bhalla R, et al. Genome-wide survey of the RIP domain family in Oryza sativa and their expression profiles under various abiotic and biotic stresses. Plant Mol Biol. 2008;67:603–614. Palmer JD, Soltis DE, Chase MW. The plant tree of life: an overview and some points of view. Am J Bot. 2004;91:1437–1445. Wilson BA, Ho M. Recent insights into Pasteurella multocida toxin and other G-protein-modulating bacterial toxins. Future Microbiol. 2010;5:1185–1201. Reyes AG, Geukens N, Gutschoven P, et al. Streptomyces coelicolor encodes a type I ribosome-inactivating protein. Microbiol. 2010;156:3021–3030.

3

Ribosome-inactivating Proteins from Phytolaccaceae Augusto Parente, Angela Chambery, Antimo Di Maro, Rosita Russo, and Valeria Severino Department of Environmental, Biological and Pharmaceutical Sciences and Technologies, Second University of Naples, Italy

Introduction

Ribosome-inactivating proteins (RIPs; rRNA N-β-glycosidases; EC 3.2.2.22) have been isolated from several species belonging to the Phytolacca genus (Fam. Phytolaccaceae), comprising many perennial plants. The interest in RIPs from Phytolaccaceae is undoubtedly related to their historical relevance and dates back to 1925 when Duggar et al. reported that extracts of Phytolacca americana L. (common name: American pokeweed; syn. P. decandra L.) were endowed with antiviral properties against tobacco mosaic virus (TMV).1 Kassanis and Kleczkowski (1948) independently isolated a glycoprotein responsible for the antiviral effect of soluble extracts of P. esculenta van Houtte (common name: Indian Poke; syn. P.  acinosa Roxb.).2 Subsequently, a protein responsible for the antiviral activity was purified from P. americana3 and characterized as the first type 1 RIP. It was named with the acronym PAP for “Pokeweed Antiviral Protein” (previously “Phytolacca americana peptide” or “Phytolacca americana protein”).4, 5 It was demonstrated that PAP was able to inhibit protein synthesis by affecting the interaction of the elongation factors EF-1 and EF-2 with the large ribosomal subunit.5 Following these first observations, many other type 1 RIPs were isolated from the Phytolacca genus, including several PAP isoforms (see below). In P. americana, very similar and sometimes identical PAP forms were found in different plant tissues (roots, leaves, and seeds), comprehensively reviewed by Irvin.6 In addition, a differential seasonal expression was evidenced for at least three different PAP isoforms (i.e., PAP-I from spring leaves, PAP-II from early summer leaves, and PAP-III from late summer leaves).7, 8 Furthermore, P. dioica (common name: Ombù tree) is a rich source of several highly conserved RIPs isolated both from leaves and seeds.9–13 Type 1 RIPs from P. heterotepala H. Walter (common name: Mexican pokeweed)14 and from the Korean pokeweed P. insularis Nakai were also characterized.15 Additional RIPs have been isolated from P. octandra L. (common name: Red Inkplant; syn. P. icosandra) and for P. dodecandra L’Herit (syn. P. abyssinica Hoffm).16 However, only partial structural and/or functional information are available for these proteins.

Ribosome-inactivating Proteins: Ricin and Related Proteins, First Edition. Edited by Fiorenzo Stirpe and Douglas A. Lappi. © 2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.

28

RIBOSOME-INACTIVATING PROTEINS FROM PHYTOLACCACEAE

29

Physicochemical Properties and Structure Physicochemical Properties and Amino Acid Composition

The physicochemical properties of RIPs from Phytolaccaceae are well within the canonical parameters of other type 1 RIPs (Table 3.1). In particular, basic theoretical pI values ranging from 7.7 (Dioicin 2) and 9.3 (PAP-alpha) and an average molecular mass of 29.3 kDa for non-glycosylated forms have been calculated. In addition, very similar theoretical E0.1%1mg/mL at 280 nm and aliphatic indices have been computed from their amino acid composition (Table 3.1). All RIPs from Phytolaccaceae have an expected prevalence of basic amino acid residues (ratio Lys + Arg/Asp + Glu higher than 1), two tryptophanyl and four cysteinyl residues. The amino acid residues present standard deviation values below 2, with the notable exceptions of Arg (3; min 8 for PAP-II, Dioicin 2, PAP icos-II, max 21 for PIP), Asn (4; min 13 for Dioicin 2, max 26 for PD-L3-4), Asp (3; min 11 for PAP-I and PAP-alpha, max 19 for Dioicin 2 and PAP icos-II), Lys (5; min 10 for PIP, max 34 for PAP icos-II), Ser (4; min 13 for PAP-S1aci, max 28 for PD-S2), Thr (3; min 16 for PIP, max 26 for PAP icos-II), Val (3; min 11 for PD-L1-2 and Dioicin 2, max 20 for PAP-I, PAP-aci and PH-L4). Primary Structure: Sequence Similarity of RIPs from Phytolaccaceae

A comprehensive list of ribosome-inactivating proteins from Phytolaccaceae for which complete primary structures are available is reported in Table 3.2. Complete sequences of RIPs from Phytolaccaceae have been determined by direct sequencing of purified proteins and/or deduced from sequencing of genomic/cDNA clones. A multiple alignment of RIPs from Phytolaccaceae yielding a 280 amino acid residue consensus sequence is shown in Figure 3.1. The sequence comparison revealed the presence of 62 identical residues among all sequences with a high identity/similarity region at positions 173–230 (i.e., 22 identical residues out of 57). The amino acid residues involved in the formation of the catalytic site (i.e., Glu189, Arg192, Trp223, and Ser227 numbering according to the consensus sequence) and Tyr78 and Tyr132 of the binding site are fully conserved (Figure 3.2). The invariant four cysteinyl residues (Cys36, Cys91, Cys113, and Cys277) involved in the formation of intramolecular disulfide bonds (Cys36-277 and Cys91-113) are also conserved together with the two Trp residues at positions 223 and 255. RIPs from Phytolaccaceae exhibit high percentages of identity (above 35%) attesting the high conservativity at both structural and functional level (Figure 3.3a). The unrooted phylogenetic tree of RIPs from Phytolaccaceae shows the presence of two major groups (Figure 3.3b). The first group includes Dioicin 2, PAP-II, and PAP icos-II, located on a separate branch with respect to a second group containing most of the other known RIPs from Phytolaccaceae. As previously reported, a closer phylogenetic relationship has been evidenced for PD-L1-2 and PD-L3-4 from P. dioica leaves and RIPs from P. americana seeds with respect to PD-S2 isolated from seeds of the same plant.13 Glycosylation of RIPs from Phytolaccaceae

Similarly to other type 1 RIPs, some RIPs from Phytolaccaceae are glycoproteins. Consensus N-glycosylation sequences are present at the N-terminal at conserved position 12 (numbering according to the consensus sequence) in PAP-S1aci, PAP-S, PAP-S2, PAP-II, PAP-alpha, PD-L1-2, PD-L3-4, Dioicin 2, PIP2, PAP icos-II) and/or at the C-terminal ends (i.e., at positions 274) in PAP-S1aci, PAP-S, PAP-S2, PAP-alpha, and PD-L1-2. Additional N-glycosylation sites may be present within internal sequence regions.

PIP2

PIP

PH-L4

PAP-S1aci

PAP-aci

Dioicin 2

PD-L3-4

PD-L1-2

PD-S2

PAP-H

PAP-alpha

PAP-II

PAP-I

PAP-S2

PAP-S

15 17 14 16 16 16 17 14 15 15 13 15 12 12 15 12 15 Ala Arg 9 11 11 8 9 11 11 10 9 8 11 9 11 21 10 11 8 Asn 25 23 25 15 25 16 18 25 26 13 24 24 23 25 23 22 14 Asp 13 12 11 18 11 13 12 14 13 19 13 13 14 12 14 15 19 Cys 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Gln 7 8 9 8 8 10 8 10 9 9 8 8 8 7 8 8 8 Glu 14 13 14 15 11 16 14 13 13 18 12 14 14 13 13 14 15 Gly 16 12 14 19 15 18 13 14 12 15 15 16 13 16 13 13 19 His 2 2 2 2 3 3 3 2 2 2 2 2 2 3 2 2 1 Ile 20 18 18 13 19 20 17 20 17 18 18 20 19 22 18 18 14 Leu 25 24 23 20 24 22 24 24 25 23 23 25 23 23 25 22 20 Lys 23 22 20 32 22 21 23 19 20 30 20 23 20 10 22 20 34 Met 5 6 5 7 6 5 5 6 6 7 4 4 3 5 7 4 7 Phe 8 11 9 8 8 11 9 8 8 8 8 9 8 8 7 7 8 Pro 12 11 12 9 11 10 11 9 9 10 14 13 12 13 12 14 11 Ser 17 20 19 16 16 23 28 23 23 19 20 13 23 19 15 20 14 Thr 18 18 20 23 22 17 17 22 21 24 20 21 19 16 23 20 26 Trp 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Tyr 12 11 10 12 13 9 12 11 11 11 11 12 11 11 13 11 12 Val 14 17 20 15 16 17 17 11 16 11 20 14 20 19 17 19 13 Residues 261 262 262 262 261 264 265 261 261 266 262 261 261 261 263 258 264 Aliphatic index 88.5 87.8 88.5 71.8 88.2 86.8 85.4 83.3 86.3 77.7 88.1 88.5 89.6 93.0 88.2 86.5 70.2 pI 8.9 9.2 9.0 9.0 9.3 8.6 9.2 8.3 8.5 7.7 9.0 8.9 8.5 9.2 8.9 8.3 9.1 Molecular mass (kDa) 29.2 29.4 29.3 29.2 29.2 29.3 29.5 29.2 29.1 29.9 29.1 29.2 29.2 29.4 29.6 28.9 29.5 E2800.1%1cm 1.00 0.94 0.89 1.00 1.05 0.84 0.99 0.95 0.95 0.92 0.95 1.00 0.95 0.94 1.03 0.96 0.99

PAP icos-I

Table 3.1 Amino acid composition and physicochemical properties of RIPs from Phytolaccaceae. All computations have been performed by using sequences retrieved from the UniProtKB database (see Table 3.2) by means of the ProtParam tool available online at www.expasy.org.73 E2800.1%1cm values were calculated assuming all pairs of Cys residues form cystines. PAP-R sequence was not included sharing 100% identity with PAP-I. PAP icos-II

PAP-acia PAP-S1aci PAP-Sb PAP-S2 PAP-S1 PAP-Ic PAP-Rd PAP-II PAP-alpha PAP-H PD-S2 PD-L1-2 PD-L3-4 Dioicin 2 PH-L4 PIP PIP2 PAP icos-Ia PAP icos-IIa

Q941G8 D0VWY1 P23339 Q8S946 Q8S947 P10297 Q53YN2 Q40772 Q03464 Q8RYA4 P34967 P84853 P84854 P85208 Q6EH50 – Q9XFF8 Q4A525 Q4A524 P. acinosa P. acinosa P. americana P. americana P. americana P. americana P. americana P. americana P. americana P. americana P. dioica P. dioica P. dioica P. dioica P. heterotepala P. insularis P. insularis P. octandra P. octandra

Plant

Sequencing source cDNA Unknown Protein Genomic DNA Genomic DNA cDNA Genomic DNA cDNA Genomic DNA cDNA Protein Protein Protein Protein Protein/cDNA cDNA Genomic DNA cDNA cDNA

Plant tissue – Seeds Seeds Seeds Seeds Leaves Roots Leaves – Roots Seeds Leaves Leaves Leavesb Leaves – – – –

PAP – – paps2 paps1 PAP1 papra PAP2 papalphaa papha – – – – RIP1 pipa PIP2 pap papII

Gene name nd 2Q8W 1GIK nd nd 1PAF nd 1LLN 1APA nd nd 3H5K 2Z4U nd nd nd nd nd nd

PDB code

Submitted to EMBL DB Submitted to PDB DB – Submitted to EMBL DB Submitted to EMBL DB Spring leaves – Summer leaves – Hairy roots – – – – Gene synonym: PhRIPI Not submitted in DB – Submitted to EMBL DB Submitted to EMBL DB

Note

b

a

When specific names were not assigned to the protein/gene submitted in database, RIPs were named by the authors. An additional PAP-S sequence deduced from cDNA has been submitted by Poyet et al.81 to UniProtKB under the accession number P93444. c PAP-C is reported as synonym of PAP-I referring to protein isolated from cultured pokeweed cells.82 d 100% sequence identity with PAP-I. e Unpublished data.

Protein name

Accession number

Peng et al.e Hogg et al.e 74 Honjo et al.e Honjo et al.e 21,75 76 77 78,79 63 80 12 12 11 14,57,65 56 15 Lin et al.e Lin et al.e

Refs.

Table 3.2 List of RIPs from Phytolaccaceae for which complete sequence data are available. The UniProtKB has been used as reference database. nd, not determined; DB, database; PDB, Protein Data Bank.

Figure 3.1 Multiple alignment of RIPs from Phytolaccaceae reported in Table  3.2. Sequences were aligned by using the ClustalW algorithm. Identical residues (*), conserved substitutions (:) and semi-conserved substitutions (.) are reported. The four cysteinyl (in bold) and the two tryptophanyl (in white) residues are shaded. The amino acid residues involved in the formation of the catalytic site (i.e., Glu189, Arg192, Trp223, and Ser227 numbering according to the consensus sequence) and Tyr78 and Tyr132 of the binding site are indicated by arrows. PAP-R and PAP-S1 sequences were not included in the alignment sharing a percentage of identity above 98% with PAP-I and PAP-S.

RIBOSOME-INACTIVATING PROTEINS FROM PHYTOLACCACEAE

32 Tyr 1

33

G

lu

18 9 -x

8

-x -A

Ty r7

rg

19 2

H2N N

H

OH O P O

H

O P

O

O 3′ O

N

2

O

O H

O

2

N

5′ C

O

H

N

Trp233-x-x-x-Ser227

N

N-β-glycosidic bond

COOH Figure 3.2 Representation of the active site of RIPs from Phytolaccaceae. The cleft containing the fingerprint of the highly conserved catalytic amino acid residues (numbering according the consensus sequence of Figure 3.1) involved in the adenine cleavage is schematically reported. (This figure also appears in the color plate section.)

Although it has been reported that RIPs N-glycosylation seems not to play a pivotal role in their enzymatic activity,17 some evidences obtained by using P. dioica RIPs as experimental system, indicate a correlation between glycan chains and the enzymatic activity.18 In particular, P. dioica contains four RIPs isoforms, named PD-L1-2 and PD-L3-4, two by two sharing the same primary structure, differing only in their glycosylation degree.12 Specifically, PD-L1 presents three N-glycosylation sites (Asn10, Asn43, and Asn255) occupied by the well-known paucidomannosidic pattern (i.e., Man3-GlcNAc2-Fuc1-Xyl), while PD-L2 lacks the glycan moiety on Asn255. Similarly, PD-L3 presents the glycan chain only on Asn10, while PD-L4 is not glycosylated. This standard plant glycan structure has been also reported for PD-S1 and PD-S2.19 Recently, Severino et al., by using a deglycosylated recombinant form of PD-L1, demonstrated that the removal of three glycan chains increased the deadenylation activity, without affecting their capability of inhibiting protein synthesis. It has been hypothesized that the absence of multiple glycosylation sites surrounding the active site could be responsible of the increased accessibility to substrates and, therefore, of the higher enzymatic efficiency measured as adenine release.18 Three-Dimensional Structure

The first X-ray diffraction studies on RIPs were performed on PAP, allowing the determination of its complete three-dimensional (3D) structure.20, 21 Later on, other RIPs from Phytolaccaceae were

Figure 3.3 (a) Identity/similarity matrix of sequences reported in Table  3.2 obtained by using the BoxShade tool available on-line at http://www.ch.embnet.org/.

Figure 3.3 (Continued) (b) Phylogenetic tree of RIPs from Phytolaccaceae. The neighbor-joining clustering method was used with Poisson corrected distances. PAP-R and PAP-S1 sequences were not included in the alignment sharing a percentage of identity above 98% with PAP-I and PAP-S.

36

RIBOSOME-INACTIVATING PROTEINS

solved at high resolution (Table 3.2), including additional PAP isoforms22, 23 and RIPs isolated from P. dioica and P. acinosa.24–27 The analysis of crystal structures revealed that all RIPs from Phytolaccaceae share an almost identical three-dimensional organization, according to their high percentages of sequence identity as discussed above. The superposition of α-carbon backbone of the seven structures available in PDB database clearly discloses the presence of identical secondary structure elements (Figure 3.4a). In particular, similarly to other RIP structures (e.g., ricin A-chain,28 trichosanthin,29 and saporin SO630), those isolated from Phytolaccaceae contain two major domains known as “RIP fold.” The N-terminal domain consists of β-strands and α-helices, while the C-terminal one contains predominantly α-helices. As an example, these domains are evidenced in yellow and blue, respectively, on the ribbon structure of PAP-I (Figure 3.4b). An identical spatial position of the highly conserved catalytic amino acid residues (i.e., Tyr72, Tyr123, Glu176, Arg179, Trp208, and Ser212, numbering of the PAP-I sequence) is also visible on the 3D structures overlay (Figure 3.4a). However, despite the high conservativity of RIPs 3D structures, some variations are located in loop regions.27 Interestingly, it has been reported that these differences may be related to the differential activity of RIPs from Phytolaccaceae. Indeed, although PD-L1 and PD-L4 share a high percentage of sequence identity, only the first is able to cleave the supercoiled double-stranded pBR322 DNA (see next section).31 As reported by Ruggiero et al., a  conformational change of the loop lining the catalytic site and containing the Asp91 residue, (a)

(b)

Ser212

Trp208 Arg179 Tyr72 Glu176 Tyr123

Figure 3.4 Three-dimensional structures of RIPs from Phytolaccaceae available in Protein Data Bank (PDB). (a) Superimposition of the following structures: PD-L1-2 (PDB code, 3H5K), PD-L3-4 (PDB code, 2Z4U), PAP-S1aci (PDB code, 2Q8W), PAP-S (PDB code 1GIK), PAP-I (PDB code, 1PAF), PAP-II (PDB code, 1LLN), and PAP-alpha (PDB code, 1APA). (b) Representative ribbon structure of PAP-I. N-terminal and C-terminal domains are colored in yellow and blue, respectively. (This figure also appears in the color plate section.)

37

RIBOSOME-INACTIVATING PROTEINS FROM PHYTOLACCACEAE

seems to open the active site cleft, thus allowing the accommodation of supercoiled DNA.27 Moreover, the same loop conformation is observed in the PAP 3D structure, which also displays a similar ability to induce DNA cleavage.32 These findings suggest that sequence variations and specific structural determinants can be related to the different RIP’s properties. Enzymatic Activities of RIPs from Phytolaccaceae

RIPs isolated from Phytolaccaceae, as generally recognized for other RIPs, are endowed with multiple enzymatic activities. The classical rRNA N-β-glycosidase activity toward ribosomes has been reported for RIPs isolated from P. americana, P. dioica, and P. heterotepala.6 In particular, it has been reported that several PAP isoforms display a broad substrate specificity, being able to depurinate eukaryotic ribosomes from plants, yeast, and animals.6 The N-β-glycosidase activity assayed on yeast and animal ribosomes has been also reported for RIPs isolated from P. dioica leaves10 and seeds,9 respectively. Yeast ribosomes were also depurinated in vitro by PH-L4 from P. heterotepala.14 In addition, a specificity towards prokaryotic ribosomes was evidenced for PAP-I and PAP-S on the large subunit of the E. coli 23S rRNA as substrate.33 As a consequence of ribosome depurination, these RIPs are able to inhibit the protein synthesis by blocking the function of the elongation factors EF-1 and EF-2.6 When assayed for the inhibition of protein synthesis by using a lysate of rabbit reticulocytes as cell-free system, RIPs from Phytolaccaceae gave IC50 values ranging from 0.05 nM (PAP-R) to 0.66 nM (Dioicin 1) as summarized in Table 3.3. It has been reported that the assay experimental conditions, including buffer ionic strength and other additional factors (e.g., ATP addition), may significantly affect the IC50 determination, thus preventing an accurate comparison of values obtained from different laboratories.6 However, the data available so far indicate that RIPs from Phytolaccaceae are endowed with similar inhibitory properties that are also common to other type 1 RIPs.17

Table 3.3

Effects of RIPs from Phytolaccaceae on protein synthesis assayed on a cell-free system.

Species

RIP

IC50 (nM)

Refs.

P. americana

PAP-I PAP-II PAP-R PAP-S

0.24 0.25 0.05 0.04

5 83 84 68

P. dioica

PD-S1 PD-S2 PD-S3 PD-L1 PD-L2 PD-L3 PD-L4 Dioicin 1 Dioicin 2

0.12 0.06 0.08 0.10 0.11 0.23 0.13 0.66 0.23

9 9 9 10 10 10 10 11 11

P. dodecandra

Dodecandrin

0.04

16

P. insularis

PIP2

0.04

15

P. heterotepala

PH-L4

0.08

14

38

RIBOSOME-INACTIVATING PROTEINS

In addition, it has been reported that some RIPs have a polynucleotide adenine glycosidase activity, being able to release adenines from polynucleotide substrates other than rRNA and ribosomes, including DNA and poly(A).34, 35 Therefore, the denomination of adenine polynucleotide glycosylase (APG) has been proposed for RIPs.36 In particular, APG activity on herring sperm DNA (hsDNA) has been reported for several PAP isoforms (i.e., PAP, PAP-II, PAP-R, PAP-S).34 In addition, PAP and PAP-II caused a concentration-dependent depurination of genomic human immunodeficiency virus type 1 (HIV-1) RNA37 and of viral capped mRNAs38 (see next section: Biological Activity). PD-Ls also release several adenines from poly(A) and hsDNA. This activity was weaker in PD-L1 and PD-L3 with respect to PD-L2 and PD-L4.10 APG activity on hsDNA was also evidenced for Dioicin 1, Dioicin 2, and PH-L4.11, 14 Noteworthy is the fact that the APG activity of RIPs from P. dioica appears to be related to amino acid residue(s) other than those classically included in the active site. It has been hypothesized that a conserved seryl residue located in the proximity of the catalytic tryptophan plays a role in this activity.39 Several activities on nucleic acids other than depurination have been also attributed to RIPs.32, 35, 40 In particular, for some RIPs, including those isolated from Phytolaccaceae (i.e., PAP, PD-L1-2, Dioicin 1, and Dioicin 2), a DNase activity on circular plasmids, resulting in their linearization and topological changes, has been revealed.11, 31, 32 However, the putative DNase activity has been a matter of debate mainly supported by concerns about the presence of contaminating nucleases within RIP preparations.41, 42 Although a chromatographic procedure based on the use of a Procion Red derivatized Sepharose chromatography has been proposed to obtain nuclease-free RIPs,42 PD-L1-2, Dioicin 1, and Dioicin 2 purified on Red Sepharose, still retained the ability to cleave supercoiled pBR322 dsDNA, generating relaxed and linear molecules while PD-L3-4, purified by applying the same procedure, did not show the same behavior.11, 31 In addition, the nicking activity on supercoiled pBR322, resulting in the production of circular and linear forms, was also evidenced for recombinant PD-L1.27 On the basis of the comparison of PD-L1 and PD-L4 crystal structures, potential determinants responsible for differential DNase activity have been proposed (see paragraph  3.2.4).27 Although DNase activity seems to be somehow inefficient with respect to N-β-glycosidase activity occurring at high RIP concentrations, it is now not doubted that some RIPs possess the ability to act as nucleases. Future investigations will definitively clarify the structural determinants of this activity.

Biological Activity Antiviral Activity

Several RIPs belonging to the Phytolacca genus are endowed with a broad-spectrum antiviral activity against plant and animal viruses.6 From an historical point of view, PAP was initially discovered due to its ability to inhibit the transmission of TMV in plants.1 Similarly, soluble extracts of P. esculenta were found to be able to inhibit the infectivity of TMV and other plant viruses (e.g., tomato bushy stunt virus, potato virus X, tobacco necrosis virus and cucumber virus).2 PAP was later classified as a RIP due to its ability to enzymatically inactivate ribosomes, thus representing the first type 1 RIP to be isolated from plants.5 Since then, several works have been devoted to investigate the antiviral activity of type 1 RIPs. In particular, RIPs from Phytolaccaceae, especially those from P. americana,43 have been found to be effective in protecting host plants, against different plant viruses (Table 3.4). Indeed, the resistance to infection against mosaic viruses, worldwide distributed phytopathogens infecting more than 150 types of plants, following RIP treatment has been widely investigated. The antiviral properties of PAP against plant viruses have been extensively reviewed.44 Furthermore,

39

RIBOSOME-INACTIVATING PROTEINS FROM PHYTOLACCACEAE

Table 3.4

RIPs from Phytolaccaceae inhibiting infection by plant and animal viruses. RIP

Virus

Experimental system

Plant viruses

PAP PAP-II PAP PAP PAP PAP PAP PAP PAP PAP PAP PAP PAPa PAPa Dioicin 2 Dioicin 2 PIPa PIPa PIPa PIP2a PhRIPI

TMV TMV SBMV AIMV CMV PVY PMV TMV PVX ACMV CaMV TMV BMV TMV TMV TNV PVX PVY PLV TMV PVX

Phaseolus vulgaris Phaseolus vulgaris Phaseolus vulgaris Chenopodium amaranticolor Chenopodium amaranticolor Chenopodium amaranticolor Chenopodium amaranticolor Nicotiana tabacum protoplasts Gomphrena globosa Nicotiana benthamiana Brassica campestris Nicotiana tabacum protoplasts Hordeum vulgare L. protoplasts Nicotiana tabacum Nicotiana tabacum, cv. Samsun NN Phaseolus vulgaris Trasgenic potato plants Trasgenic potato plants Trasgenic potato plants Nicotiana tabacum, cv. Samsun NN Nicotiana tabacum, cv. Samsun NN

83 83 3 85 85 85 85 86 85 85 85 87 88 89 90 90 56 56 56 15 57

Animal viruses

PAP PAP PAP PAP-S PAP PAP PAP PAP-S PAP-S

HIV-1 HTLV-I HBV HBV LCMV Poliovirus HSV HSV-1 Poliovirus

293 T PAP-expressing cells 293 T cells HepG2 PAP-expressing cells HepG2 2.2.15 cells Mice HeLa cells Vero cells HEp-2 cells HEp-2 cells

54 91 50 50 51 47 49 92 92

a

Refs.

Recombinant protein

antiviral activity towards mammalian viruses including influenza virus,45 poliovirus,46 herpes simplex virus,47, 48 hepatitis B virus,49 lymphocytic choriomeningitis virus,50 and HIV-151, 52 has been reported. Mansouri et al. recently demonstrated that the infectivity of HIV-1 particles released by PAP-expressing 293 T cells is increased through the activation of the ERK1/2 Mitogen Activated Protein Kinase pathway.53 Other RIPs with antiviral activities belong to the Phytolacca genus. It was reported that the contemporaneous inoculation in the same leaf of Dioicin 2 with TMV and tobacco necrosis virus (TNV) protected Nicotiana tabacum and Phaseolus vulgaris from infection, respectively.54 Furthermore, when recombinant PIP and PIP2 are inoculated with pathogens or expressed in transgenic plants, they are able to confer resistance to host plants (e.g., potato plants) against TMV, potato virus X, potato virus Y, and potato leafroll virus.15, 55 Similarly, PhRIPI inducible expression in Nicotiana tabacum, resulted in the enhanced resistance of tobacco leaves against PVX infection.56 The mechanism of action for the antiviral activity of RIPs has not yet been clarified. On the basis of their enzymatic activity and their localization within vacuoles, protein bodies, or cell walls, it has been proposed that RIPs may act as “suicidal agents” released from subcellular compartments

40

RIBOSOME-INACTIVATING PROTEINS

following plant infection by a virus through ribosomal inactivation and consequent prevention of viral replication.57 However, recent evidence suggests that RIP antiviral activity does not solely depend on ribosome depurination. An unexpected activity of PAP has been related to the specific capped RNAs recognition and depurination.38 Although the cap structure is not removed, the depurination of viral capped mRNAs leads to the inhibition of their translation.58 Furthermore, mRNA depurination also negatively affects the correct packaging of brome mosaic virus (BMV) viral genome into particles.59, 60 Further investigations will clarify the molecular mechanism involving viral mRNAs depurination in antiviral activity. Antifungal Activity

RIPs isolated from wheat, barley, and maize have been shown to be endowed with antifungal properties.6 The capability to inactivate fungal ribosomes is further increased by the synergic action of glucanase and chitinase enzymes.61 Although the lack of antifungal activity was initially reported for PAP,43 it has been later shown that PAP-H, purified from Agrobacterium rhizogenes-transformed hairy roots of P. americana and constitutively secreted in the root exudates, was able to depurinate fungal ribosomes in vitro and in vivo.62 However, its antifungal activity against soil-borne fungi was revealed only when assayed in root exudates containing additional enzymatic activities (i.e., chitinase, glucanase, and proteases). In contrast, purified PAP-H did not inhibit the fungal growth, thus supporting the hypothesis of the occurrence of a synergic mechanism enabling PAP-H to penetrate fungal cells.62 Moreover, the expression of some non-toxic PAP mutants was reported to be effective in protecting creeping bentgrass against fungal diseases.63 In addition, PhRIPI was able to enhance the resistance against Alternaria alternata and Botrytis cinerea fungal pathogens upon inducible expression in tobacco.64 These findings suggest that Phytolacca species may constitute a still unexplored source of potential antifungal proteins for future biotechnological applications. Other Biological Activities

Some type 1 RIPs were found to have insecticidal activity. In particular, entomotoxic properties have been reported for PAP-S and other type 1 RIPs against two different lepidopteran (i.e., the soybean caterpillar Anticarsia gemmatalis and fall armyworm Spodoptera frugiperda).65 It has been found that the ingestion of low doses of RIPs affected the development and survival rate of both insects, thus supporting their potential utilization to enhance plant resistance against pests.65 Some works also investigated the toxic effects of RIPs from Phytolaccaceae on cultured cells and in vivo model systems. Differently from type 2 RIPs, the cytotoxicity of type 1 RIP both in vitro and in vivo is very low, mainly due to the lack of the lectin B chain that allows the toxin internalization within cells.66 However, a cytotoxic action on Hela cells has been evidenced for PAP-S and PAP.66, 67 The toxicity was found to be strictly dependent from incubation times and protein concentration, being revealed only on prolonged incubation times and at micromolar concentrations.66 It has been proposed that type 1 RIP cytotoxicity may involve the toxin internalization through pinocytosis.68 According to this hypothesis, the cellular uptake of native glycosylated PD-Ls as well as of recombinant PD-L1, was investigated in NIH3T3 cells.18 Interestingly, a peculiar distribution in grains, indicative of a vesicular compartmentalization and apparently related to the glycosylation degree, was evidenced. This behavior is independent from the recognition of the mannosylated glycoproteins, since NIH3T3 cells lack mannose receptors.69 A recent study also demonstrated that in 293 T cells the PAP expression induced the activation of c-Jun NH2-terminal kinase (JNK) depending on rRNA depurination.70 However, JNK activation did not lead to apoptosis nor cause cell death. Accordingly, previous works reported that the intraperitoneal injection of PAP-S into mice is lethal only at high doses with an LC50 of 1 mg/kg.67 Also, PAP intravenous doses of

RIBOSOME-INACTIVATING PROTEINS FROM PHYTOLACCACEAE

41

about  5 mg/kg did not cause death, evidencing a lower toxicity of PAP with respect to PAP-S.67 The capability of PAP to attenuate liver fibrogenesis through the down-regulation of the Wnt/βcatenin pathway, has been also reported, suggesting its potential use for therapeutic applications.71

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23. 24. 25.

Duggar BM, Armstrong JK. The effect of treating the virus of tobacco mosaic with the juice of various plants. Annals of the Missouri Botanical Garden. 1925;12:359–366. Kassanis B, Kleczkowski A. The isolation and some properties of a virus-inhibiting protein from Phytolacca esculenta. J Gen Microbiol. 1948;2:143–153. Wyatt SD, Shepherd RJ. Isolation and characterization of a virus inhibitor from Phytolacca americana. Phytopathology. 1969;59:1787–1794. Irvin JD. Purification and partial characterization of the antiviral protein from phytolacca americana which inhibits eukaryotic protein synthesis. Arch Biochem Biophys. 1975;169:522–528. Obrig TG, Irvin JD, Hardesty B. The effect of an antiviral peptide on the ribosomal reactions of the peptide elongation enzymes, EF-I and EF-II. Arch Biochem Biophys. 1973;155:278–289. Chessin M, De Borde D, Zipf A. Antiviral Proteins in Higher Plants. Boca Raton, FL: CRC Press, 1995:180. Rajamohan F, Venkatachalam TK, Irvin JD, Uckun FM. Pokeweed antiviral protein isoforms PAP-I, PAP-II, and PAP-III depurinate RNA of human immunodeficiency virus (HIV)-1. Biochem Biophys Res Commun. 1999;260:453–458. Houston LL, Ramakrishnan S, Hermodson MA. Seasonal variations in different forms of pokeweed antiviral protein, a potent inactivator of ribosomes. J Biol Chem. 1983;258:9601–9604. Parente A, De Luca P, Bolognesi A, et al. Purification and partial characterization of single-chain ribosome-inactivating proteins from the seeds of Phytolacca dioica L. Biochim Biophys Acta. 1993;1216:43–49. Di Maro A, Valbonesi P, Bolognesi A, et al. Isolation and characterization of four type-1 ribosome-inactivating proteins, with polynucleotide:adenosine glycosidase activity, from leaves of Phytolacca dioica L. Planta. 1999;208:125–131. Parente A, Conforto B, Di Maro A, et al. Type 1 ribosome-inactivating proteins from Phytolacca dioica L. leaves: differential seasonal and age expression, and cellular localization. Planta. 2008;228:963–975. Di Maro A, Chambery A, Carafa V, et al. Structural characterization and comparative modeling of PD-Ls 1–3, type 1 ribosome-inactivating proteins from summer leaves of Phytolacca dioica L. Biochimie. 2009;91:352–363. Parente A, Berisio R, Chambery A, Di Maro A. Type 1 Ribosome-Inactivating Proteins from the Ombú Tree (Phytolacca dioica L.). In: Lord JM, Hartley MR, (eds.) Toxic Plant Proteins Plant Cell Monographs Vol. 18. Berlin: Springer 2010:79–106. Di Maro A, Chambery A, Daniele A, et al. Isolation and characterization of heterotepalins, type 1 ribosome-inactivating proteins from Phytolacca heterotepala leaves. Phytochemistry. 2007;68:767–776. Song SK, Choi Y, Moon YH, et al. Systemic induction of a Phytolacca insularis antiviral protein gene by mechanical wounding, jasmonic acid, and abscisic acid. Plant Mol Biol. 2000;43:439–450. Ready MP, Adams RP, Robertus JD. Dodecandrin, a new ribosome-inhibiting protein from Phytolacca dodecandra. Biochim Biophys Acta. 1984;791:314–319. Barbieri L, Battelli MG, Stirpe F. Ribosome-inactivating proteins from plants. Biochim Biophys Acta. 1993;1154:237–282. Severino V, Chambery A, Di Maro A, et al. The role of the glycan moiety on the structure-function relationships of PD-L1, type 1 ribosome-inactivating protein from P. dioica leaves. Mol Biosyst. 2010;6:570–579. Chambery A, Di Maro A, Parente A. Primary structure and glycan moiety characterization of PD-Ss, type 1 ribosome-inactivating proteins from Phytolacca dioica L. seeds, by precursor ion discovery on a Q-TOF mass spectrometer. Phytochemistry. 2008;69:1973–1982. Irvin JD, Robertus JD, Monzingo AF. Preliminary x-ray diffraction studies on an anti-viral protein. Biochem Biophys Res Commun. 1977;74:775–779. Monzingo AF, Collins EJ, Ernst SR, et al. The 2.5 A structure of pokeweed antiviral protein. J Mol Biol. 1993;233:705–715. Zeng ZH, He XL, Li HM, et al. Crystal structure of pokeweed antiviral protein with well-defined sugars from seeds at 1.8A resolution. J Struct Biol. 2003;141:171–178. Kurinov IV, Uckun FM. High resolution X-ray structure of potent anti-HIV pokeweed antiviral protein-III. Biochem Pharmacol. 2003;65:1709–1717. Ruggiero A, Chambery A, Di Maro A, et al. Crystallization and preliminary X-ray diffraction analysis of PD-L1, a highly glycosylated ribosome inactivating protein with DNase activity. Protein Pept Lett. 2007;14:407–409. Ruggiero A, Chambery A, Di Maro A, et al. Atomic resolution (1.1 A) structure of the ribosome-inactivating protein PD-L4 from Phytolacca dioica L. leaves. Proteins. 2008;71:8–15.

42 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

RIBOSOME-INACTIVATING PROTEINS

Ruggiero A, Chambery A, Di Maro A, et al. Crystallization and preliminary X-ray diffraction analysis of PD-L4, a ribosome inactivating protein from Phytolacca dioica L. leaves. Protein Pept Lett. 2007;14:97–100. Ruggiero A, Di Maro A, Severino V, et al. Crystal structure of PD-L1, a ribosome inactivating protein from Phytolacca dioica L. leaves with the property to induce DNA cleavage. Biopolymers. 2009;91:1135–1142. Montfort W, Villafranca JE, Monzingo AF, et al. The three-dimensional structure of ricin at 2.8 A. J Biol Chem. 1987;262:5398–5403. Zhou K, Fu Z, Chen M, et al. Structure of trichosanthin at 1.88 A resolution. Proteins. 1994;19:4–13. Savino C, Federici L, Ippoliti R, et al. The crystal structure of saporin SO6 from Saponaria officinalis and its interaction with the ribosome. FEBS Lett. 2000;470:239–243. Aceto S, Di Maro A, Conforto B, et al. Nicking activity on pBR322 DNA of ribosome inactivating proteins from Phytolacca dioica L. leaves. Biol Chem. 2005;386:307–317. Wang P, Tumer N. Pokeweed antiviral protein cleaves double-stranded supercoiled DNA using the same active site required to depurinate rRNA. Nucleic Acids Res. 1999;27:1900–1905. Hartley MR, Legname G, Osborn R, et al. Single-chain ribosome inactivating proteins from plants depurinate Escherichia coli 23S ribosomal RNA. FEBS Lett. 1991;290:65–68. Barbieri L, Valbonesi P, Bonora E, et al. Polynucleotide:adenosine glycosidase activity of ribosome-inactivating proteins: effect on DNA, RNA and poly(A). Nucleic Acids Res. 1997;25:518–522. Nicolas E, Beggs JM, Haltiwanger BM, Taraschi TF. A new class of DNA glycosylase/apurinic/apyrimidinic lyases that act on specific adenines in single-stranded DNA. J Biol Chem. 1998;273:17216–17220. Barbieri L, Valbonesi P, Bondioli M, et al. Adenine glycosylase activity in mammalian tissues: an equivalent of ribosomeinactivating proteins. FEBS Lett. 2001;505:196–197. Rajamohan F, Kurinov IV, Venkatachalam TK, Uckun FM. Deguanylation of human immunodeficiency virus (HIV-1) RNA by recombinant pokeweed antiviral protein. Biochem Biophys Res Commun. 1999;263:419–424. Hudak KA, Wang P, Tumer NE. A novel mechanism for inhibition of translation by pokeweed antiviral protein: depurination of the capped RNA template. RNA. 2000;6:369–380. Chambery A, Pisante M, Di Maro A, et al. Invariant Ser211 is involved in the catalysis of PD-L4, type I RIP from Phytolacca dioica leaves. Proteins. 2007;67:209–218. Ling J, Liu WY, Wang TP. Cleavage of supercoiled double-stranded DNA by several ribosome-inactivating proteins in vitro. FEBS Lett. 1994;345:143–146. Day PJ, Lord JM, Roberts LM. The deoxyribonuclease activity attributed to ribosome-inactivating proteins is due to contamination. Eur J Biochem. 1998;258:540–545. Barbieri L, Valbonesi P, Righi F, et al. Polynucleotide:Adenosine glycosidase is the sole activity of ribosome-inactivating proteins on DNA. J Biochem. 2000;128:883–889. Chen ZC, White RF, Antoniw JF, Lin Q. Effect of pokeweed antiviral protein (PAP) on the infection of plant viruses. Plant Pathology. 1991;40:612–620. Irvin JD. Pokeweed antiviral protein. Pharmacology and Therapeutics. 1983;21:371–387. Tomlinson JA, Walker VM, Flewett TH, Barclay GR. The inhibition of infection by cucumber mosaic virus and influenza virus by extracts from phytolacca americana. J Gen Virol. 1974;22:225–232. Ussery MA, Irvin JD, Hardesty B. Inhibition of poliovirus replication by a plant antiviral peptide. Ann N Y Acad Sci. 1977;284:431–440. Aron GM, Irvin JD. Inhibition of herpes simplex virus multiplication by the pokeweed antiviral protein. Antimicrob Agents Chemother. 1980;17:1032–1033. Teltow GJ, Irvin JD, Aron GM. Inhibition of herpes simplex virus DNA synthesis by pokeweed antiviral protein. Antimicrob Agents Chemother. 1983;23:390–396. He YW, Guo CX, Pan YF, et al. Inhibition of hepatitis B virus replication by pokeweed antiviral protein in vitro. World J Gastroenterol. 2008;14:1592–1597. Uckun FM, Rustamova L, Vassilev AO, et al. CNS activity of Pokeweed anti-viral protein (PAP) in mice infected with lymphocytic choriomeningitis virus (LCMV). BMC Infect Dis. 2005;5:9. Erice A, Lieler CL, Meyers DE, et al. Inhibition of zidovudine (AZT)-sensitive strains of human immuno deficiency virus type 1 by pokeweed antiviral protein targeted to CD4+ cells. Antimicrobial Agents and Chemotherapy. 1993;37:835–838. Zarling JM, Moran PA, Haffar O, et al. Inhibition of HIV replication by pokeweed antiviral protein targeted to CD4+ cells by monoclonal antibodies. Nature. 1990;347:92–95. Mansouri S, Kutky M, Hudak KA. Pokeweed antiviral protein increases HIV-1 particle infectivity by activating the cellular mitogen activated protein kinase pathway. PLoS One. 2012;7:e36369. Faoro F, Conforto B, Di Maro A, et al. Activation of plant defence response contributes to the antiviral activity of Diocin 2 from Phytolacca dioica. IOBC/wprs Bull. 2009;44:53–57.

RIBOSOME-INACTIVATING PROTEINS FROM PHYTOLACCACEAE

55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.

66. 67. 68. 69. 70. 71.

43

Moon YH, Song SK, Choi KW, Lee JS. Expression of a cDNA encoding Phytolacca insularis antiviral protein confers virus resistance on transgenic potato plants. Mol Cells. 1997;7:807–815. Corrado G, Scarpetta M, Alioto D, et al. Inducible antiviral activity and rapid production of the Ribosome-Inactivating Protein I from Phytolacca heterotepala in tobacco. Plant Science. 2008;174:467–474. Bonness MS, Ready MP, Irvin JD, Mabry TJ. Pokeweed antiviral protein inactivates pokeweed ribosomes; implications for the antiviral mechanism. Plant J. 1994;5:173–183. Hudak KA, Bauman JD, Tumer NE. Pokeweed antiviral protein binds to the cap structure of eukaryotic mRNA and depurinates the mRNA downstream of the cap. RNA. 2002;8:1148–1159. Karran RA, Hudak KA. Depurination of Brome mosaic virus RNA3 inhibits its packaging into virus particles. Nucleic Acids Res. 2011;39:7209–7222. Karran RA, Hudak KA. Depurination within the intergenic region of Brome mosaic virus RNA3 inhibits viral replication in vitro and in vivo. Nucleic Acids Res. 2008;36:7230–7239. Leah R, Tommerup H, Svendsen I, Mundy J. Biochemical and molecular characterization of three barley seed proteins with antifungal properties. J Biol Chem. 1991;266:1564–1573. Park SW, Lawrence CB, Linden JC, Vivanco JM. Isolation and characterization of a novel ribosome-inactivating protein from root cultures of pokeweed and its mechanism of secretion from roots. Plant Physiol. 2002;130:164–178. Dai WD, Bonos S, Guo Z, et al. Expression of pokeweed antiviral proteins in creeping bentgrass. Plant Cell Rep. 2003;21:497–502. Corrado G, Bovi PD, Ciliento R, et al. Inducible Expression of a Phytolacca heterotepala Ribosome-Inactivating Protein Leads to Enhanced Resistance Against Major Fungal Pathogens in Tobacco. Phytopathology. 2005;95:206–215. Vargas RB, Martins JN, Bordin D, et al. Type 1 ribosome-inactivating proteins—Entomotoxic, oxidative and genotoxic action on Anticarsia gemmatalis (Hu¨ bner) and Spodoptera frugiperda (J.E. Smith) (Lepidoptera: Noctuidae). Journal of Insect Physiology. 2009;55:51–58. Aron GM, Irvin JD. Cytotoxicity of pokeweed antiviral protein. Cytobios. 1988;55:105–111. Barbieri L, Aron GM, Irvin JD, Stirpe F. Purification and partial characterization of another form of the antiviral protein from the seeds of phytolacca americana L. (pokeweed). Biochem J. 1982;203:55–59. Goldmacher VS, Tinnel NL, Nelson BC. Evidence that pinocytosis in lymphoid cells has a low capacity. J Cell Biol. 1986;102:1312–1319. Hebert E, Monsigny M. Galectin-3 mRNA level depends on transformation phenotype in ras-transformed NIH 3 T3 cells. Biol Cell. 1994;81:73–76. Chan Tung KW, Mansouri S, Hudak KA. Expression of pokeweed antiviral protein in mammalian cells activates c-Jun NH2-terminal kinase without causing apoptosis. Int J Biochem Cell Biol. 2008;40:2452–2461. Wenting L, Chuanlong Z, Xiliu C, et al. Pokeweed antiviral protein down-regulates Wnt/β-catenin signalling to attenuate liver fibrogenesis in vitro and in vivo. Digestive and Liver Disease. 2011;43:559–566.

4

Ribosome-inactivating Proteins in Caryophyllaceae, Cucurbitaceae, and Euphorbiaceae Tzi Bun Ng and Jack Ho Wong School of Biomedical Sciences, The Chinese University of Hong Kong, China

Introduction

Ribosome-inactivating proteins (RIPs) are distributed among different plant families and divided into two types: type 1 and type 2. Type 1 RIPs are single-chain proteins whereas type 2 RIPs are composed of an RIP chain and a lectin chain. In the last decade there have been a number of excellent reviews on various aspects of this important class of proteins. 1–14 The intent of this chapter is to discuss type 1 RIPs with an emphasis on those from the Caryophyllaceae, Cucurbitaceae, and Euphorbiaceae families. The small type 1 RIPs and type 1 RIPs from other families excluding Phytolaccaceae are discussed briefly. Type 1 RIPs from Phytolaccaceae and type 2 RIPs will be dealt with in other chapters. The Caryophyllaceae RIPs include agrostin, dianthin, gypsophilin, petroglaucin 1, petroglaucin 2, petrograndin, saporin, and Vaccaria pyramidata. The Cucurbitaceae RIPs comprise lagenin, luffangulin, luffins, luffacylin, luffinS, luffin P1, balsamin, Momordica antiviral protein, alpha- and beta-momorcharins, charantin (small RIP), cochinin B, momorgrosvin, momorcochin, Marah oreganos RIPs, moschatin, Cucurbita moschata RIP, pepocin, sechiumin, trichokirin, trichokirin-S1, trichomaglin, trichoanguin, trichosanthin, karasurin-A, karasurin-B, karasurin–C, and trichosanthrip. The Euphorbiaceae RIPs include curcin, gelonin, lychnin, and mapalmin.

Caryophyllaceae RIPs Agrostin

Agrostin from Agrostemma githago seeds inhibited 3[H]-thymidine incorporation and induced apoptosis in human leukemic HL-60 cells by down-regulation of the intracellular bcl-2 protein level.15

Ribosome-inactivating Proteins: Ricin and Related Proteins, First Edition. Edited by Fiorenzo Stirpe and Douglas A. Lappi. © 2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.

44

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

45

Dianthin

Dianthin 30 and dianthin 32 from Dianthus caryophyllus (carnation) leaves are 29.5-kDa and 31.7-kDa mannose-containing glycoproteins that suppressed protein synthesis in a rabbit reticulocyte lysate with an IC50 (concentration giving 50% inhibition) of 0.31 nM and 0.11 nM, respectively. They markedly reduced the number of local lesions produced by tobacco-mosaic virus on Nicotiana glutinosa leaves.16

Gypsophilin

Gypsophilin from Gypsophila elegans is a 28-kDa protein with a pI close to 10 and an N-terminal amino acid closely resembling those of other RIPs. The enzyme exhibits RNA N-glycosidase activity. Its IC50 values for ribosomes from rat liver, wheat germ, and E. coli are 39.8 pM, 0.24 nM, and 0.82 uM, respectively. It is distributed within vacuoles in the cytoplasm and found in abundance in the intercellular spaces.17

Petroglaucin 1 and Petrograndin

Petroglaucin 1 and petrograndin isolated from Petrocoptis glaucifolia and P. grandiflora, respectively, did not display similarity in N-terminal amino acid sequence with petroglaucin 2 from P. glaucifolia. Petroglaucin 1 resembled Cucurbitaceae RIPs whereas petroglaucin 2 and petrograndin resembled the Caryophyllaceae RIPs saporins and dianthin 30 in sequence. Although they repressed protein synthesis at subnanomolar concentrations in rabbit reticulocyte lysates and other eukaryotic cell-free systems, there was no activity on bacterial ribosomes.18

Saporin

Saponaria officinalis RIP, SO-6, displays pronounced sequence homology to pokeweed antiviral protein II (Family Phytolaccaceae). Its antiserum demonstrate crossreactivity with other proteins (SO-5, SO-8, and SO-9) from Saponaria officinalis seeds, but not with dianthin 32, gelonin, and Momordica charantia inhibitor.19 The genotoxic effects of saporin from Saponaria officinalis root cultures have been examined. Saporin induced formation of micronuclei in cultured human lymphocytes producing a reduction in cell viability and a rise in apoptosis as evidenced by the emergence of cytosolic oligonucleosomes. However, sister-chromatid exchange was indiscernible.20 Specific neural lesions can be brought about by targeting cytotoxins such as RIPs to specific neuronal populations in accordance with their common expression of a particular surface molecule. General molecular neurosurgery protocols are adopted. The first depends on the use of targeted toxins to produce anatomically-restricted lesions based on retrograde axonal transport of the toxin. The second involves the use of anti-neuronal immunotoxins to produce type-selective and anatomically-restricted lesions.21 Cys255saporin-3 has been produced by mutation of saporin-3 DNA by introducing a cysteine residue, followed by protein expression and purification. Cys255saporin-3, saporin-3 isomer, and commercially available saporin were equipotent in inhibiting protein synthesis. The reactivity of

46

RIBOSOME-INACTIVATING PROTEINS

Cys255saporin-3 was established by labeling with a thio-specific fluorescent probe and conjugation with a nonspecific mouse IgG. Thus, a single cysteine within the saporin molecule furnishes one way for antibody conjugation that guarantees a uniform and reproducible modification of a saporin variant with full bioactivity.22

RIPs from Saponaria Ocymoides and Vaccaria Pyramidata

RIPs from Saponaria ocymoides and Vaccaria pyramidata seeds had a pI higher than 9.5 and a molecular mass of 30.2 kDa and 28 kDa, respectively. They resembled saporin-S6 and dianthin 30, also from Caryophyllaceae, in N-terminal amino acid sequence, and displayed partial crossreaction with sera against these proteins. They inhibited protein synthesis in a rabbit- reticulocyte lysate (with IC50 below 1 nM) and various cell lines (with IC50 ranging from 4 nM to > 3000 nM) and demonstrated rRNA N-glycosidase activity.23

Cucurbitaceae RIPs Bryodin

Bryodin is a strongly basic 30-kDa glycoprotein from Bryonia dioica (white bryony) roots that inhibited protein synthesis by a rabbit reticulocyte lysate with an IC50 of 0.12 nM and was considerably less potent in inhibiting whole-cell protein synthesis, with IC50 values in the range of 46 nM to 2.27 uM. Bryodin minimized lesions caused by tobacco mosaic virus in tobacco foliage.24

Colocin 1 and Colocin 2

RIPs from Citrullus colocynthis seeds (colocin 1 and 2) are 30-kDa glycoproteins with an alkaline isoelectric point. They suppress protein synthesis by a rabbit reticulocyte lysate and phenylalanine polymerization by isolated ribosomes and demonstrated rRNA N-glycosidase activity.25

Cucumis Melo

Partial characterization of the translational inhibitor present in seeds of Cucumis melo has been reported.26

Cucurmosin

Tyr70, Tyr109, Glu158, and Arg161 form the active site residues of cucurmosin from pumpkin (Cucurbita moschata) sarcocarp as an RNA N-glycosidase. Cucurmosin was potently cytotoxic to three human and murine cancer cell lines, but toxicity to normal cells was low. The crystal structure of cucurmosin has been determined at 1.04 A. There is a large N-terminal domain comprised of seven alpha-helices and eight beta-strands, and a smaller C-terminal domain with three alphahelices and two beta-strands. Asn225 is a glycosylation site.27

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

47

Hispin

The 21-kDa RIP designated as hispin from hairy melon (Benincasa hispida var. chieh-qua) seeds inhibited translation in the cell-free rabbit reticulocyte lysate system with an IC50 of 165 pM and exhibited N-glycosidase activity. Antifungal activity was also observed.28 Lagenin

Lagenin from Lagenaria siceraria seeds inhibits cell-free translation in a rabbit reticulocyte system with an IC50 of 0.21 nM. It had a low molecular weight of 20 kDa, and an N-terminal sequence with a lower extent of similarity to those of other Cucurbitaceae RIPs.29 Luffangulin

A 5.6-kDa RIP designated as luffangulin, with an N-terminal sequence resembling that of the 6.5-kDa-arginine/glutamate-rich polypeptide from sponge gourd (Luffa cylindrica) seeds, was isolated from Luffa acutangula seeds. The peptide inhibits cell-free translation with an IC50 of 3.5 nM but was devoid of HIV-1 reverse transcriptase inhibitory activity. On CM-cellulose, luffangulin and the 28-kDa RIP luffaculin appear as two adjacent peaks.30 Luffa Cylindrica Protein Synthesis Inhibitors

Three proteins which inhibited protein synthesis in rabbit reticulocyte lysates, designated as 19 K-PSI, 15 K-PSI, and 9 K-PSI, with a molecular mass of 19 kDa, 15 kDa, and 9 kDa, respectively, were isolated from Luffa cylindrica seeds. Although devoid of protein synthesis inhibitory activity in HeLa cells, the cell-free protein synthesis inhibitory activity of 19 K-PSI was 340- and 83-fold more potent than those of ricin A-chain and luffin-a, respectively. The inhibitory activities of 15 Kand 9 K-PSIs are attenuated. 19 K-PSI is a glycoprotein characterized by the possession of three intramolecular disulfide bonds and a blocked N-terminal residue, whereas 15 K-PSI is glycine-rich and 9 K-PSI has an abundance of arginine and glutamic acid and/or glutamine residues.31 The positive charge on lysine residues in luffin plays a crucial role in its immunological and protein synthesis inhibitory activities.32 Agrostin, luffin, and saporin are approximately equipotent in suppressing the mitogenic response of murine splenocytes to concanavalin A, with maximal inhibition achieved at about 80 nM and approximately 50% inhibition attained at 830 pM.33 Luffin

Two immunologically distinct glycoproteins from Luffa cylindrica seeds, 28-kDa luffin-a and 28.5-kDa luffin-b, vary in the content of aspartic acid, threonine, proline, and alanine but are similar in the content of other amino acids. They inhibit protein synthesis in a cell-free rabbit reticulocyte system and thymidine uptake by human choriocarcinoma cells.34 Luffin-a is composed of 248 amino acids and sugar chains attached to Asn residues at positions 28, 33, 77, 84, 206, and 227. It manifests 33% sequence identity to ricin A-chain.35

48

RIBOSOME-INACTIVATING PROTEINS

His140, Tyr165, and Lys231 exposed on the surface of momordin-a and luffin-a are identified as residues critical to protein synthesis inhibitory activity. Chemical modification of histidine, tyrosine, and lysine residues of momordin-a and luffin with diethylpyrocarbonate, KI3 and trinitrobenzensulfonic acid lead to a drastic decrement of protein synthesis inhibitory activity.36

Luffacylin (Small RIP)

The 7.8-kDa RIP from sponge gourd seeds, designated as luffacylin, exhibited an N-terminal sequence with remarkable homology to that of the 6.5-kDa arginine-glutamate rich polypeptide from the same source. Luffacylin inhibits translation in a rabbit reticulocyte lysate system with an IC50 of 140 pM and reacts positively in the N-glycosidase assay. Luffacylin suppresses mycelial growth in Fusarium oxysporum and Mycosphaerella arachidicola.37

LuffinS (Small RIP)

LuffinS(1), LuffinS(2), and LuffinS(3) are characterized by a molecular weight of circa 8 kDa, and A, P, and T as the N-terminal amino acid residues, respectively. LuffinS(2) has the N-terminal amino acid sequence PRRGKEAFD. LuffinSs are more potent than trichosanthin in inhibiting cellfree protein synthesis, with IC50 values of 130 pM, 1 nM, and 630 pM, respectively.38 LuffinS from Luffa cylindrica seeds is about 10 kDa in molecular weight. Its cell-free protein synthesis inhibitory activity is akin to trichosanthin.39

Luffin P1 (Small RIP)

A 5226.5-kDa peptide from Luffa cylindrica seeds, luffin P1, which inhibits protein synthesis in the cell-free rabbit reticulocyte lysate translation system with an IC50 of 0.6 nM, is identical in N-terminal sequence to that (from G3 to R13) of 6.5 K Arg/Glu rich peptide from the same seeds. Luffin P1 closely resembles trypsin inhibitor C2 peptide from pumpkin seeds and demonstrates a trypsin inhibitory activity with an IC50 of 22 uM.40 Luffin P1 inhibits cell-free protein synthesis with an IC50 of 0.88 nM. The recombinant hIL2-Luffin P1 immunotoxin suppresses T-cell proliferation in mixed lymphocyte reaction and ConA response with an IC50 of 1.8–10 nM, and lengthens the survival of major MHC-mismatched skin and kidney allografts in animal models.41

Balsamin

Balsamin from Momordica balsamina seeds suppresses protein synthesis in a cell-free rabbit reticulocyte lysate with an IC50 of 90.6 ng/ml. It displays RNA N-glycosidase activity. Its N-terminal sequence demonstrates over 80% similarity to α-momorcharin and lesser similarity to β-momorcharin, bryodin I, and luffin a. Its secondary structure is composed of alpha-helix (23.5%), beta-strand (24.6%), turn (20%), and random coil (31.9%).42 The amino acid sequence of momordin H from Momordica balsamina, after likely leader processing, is homologous to trichosanthin (57%) and momordin I (51%).43

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

49

Bitter Gourd Protein (MRK29)

Thai bitter gourd protein (MRK29), isolated from Momordica charantia fruits and seeds, is a 28.6-kDa protein with an isoelectric point of or exceeding 9, and the N-terminal amino acid sequence DVNFRLSGADPRXYGMFIED which is more similar to alpha-momorcharin than beta-momorcharin. MRK29 inhibited HIV-1 reverse transcriptase, and at the concentration of 6 nM brought about 82% reduction of viral core protein p24 expression in HIV-infected cells.44

MAP30

MAP30 acts against various stages of the HIV life cycle, on acute infection and replication in chronically-infected cells. MAP30 exhibited antitumor, viral DNA inactivating, viral integrase inhibitory, and cell-free ribosome-inactivating activities. The cloning and expression of the gene encoding biologically active re-MAP30 furnishes material for conducting clinical studies and structure–function research.45

Alpha- and Beta-Momorcharins

A purification scheme, entailing ion exchange chromatography on DEAE-cellulose and Mono-S FPLC, was developed for the purification of alpha- and beta-momorcharins from Momordica charantia seeds. The N-terminal amino acid sequence of beta-momorcharin (DVNFDLSTATAK TYTKFIED) differs from that of alpha-momorcharin (DVSFRLSGADPRSYGMFIKD) in 10 out of the 20 positions. Their N-glycosidase activity is optimal at pH7, is not inhibited by K + ions, and is not appreciably affected by NH4+ ions. The activity of alpha-momorcharin is not drastically altered by Mn2+ ions. The activity of beta-momorcharin is inhibited by about 40% in the presence of (1–10 mM) Mn2+ ions.46 The structures of two main oligosaccharides were established. These two oligosaccharides are the first examples with xylose (or fucose) but no alpha-mannosyl linkage among the N-linked oligosaccharides of glycoproteins from both animal and plant origins.47 The alpha- and beta-momorcharins suppress concanavalin A-, phytohaemagglutinin-, and lipopolysaccharide-induced mitogenic response in murine splenocytes and alloantigen-induced lymphoproliferation. The in vitro production of a primary cytotoxic lymphocyte response, and the functional capacity of macrophages, including cytostatic and phagocytic activities, are inhibited. Injections of nontoxic microgram quantities of momorcharins into mice reduced the delayed-type hypersensitivity response, humoral antibody formation in response to ovine erythrocytes, and thioglycollate-induced macrophage migration in vivo. However, there is little effect on the activation of natural killer cells in vivo.48

Charantin (Small RIP)

A 9.7-kDa peptide designated as charantin from bitter gourd seeds displays N-terminal sequence similarity to the 7.8-kDa napin-like bitter gourd peptide. Charantin reacts positively in the N-glycosidase assay and inhibits cell-free translation in a rabbit reticulocyte lysate system with an IC50 of 400 nM.49

50

RIBOSOME-INACTIVATING PROTEINS

Momorcochin

A 32-kDa glycoprotein with an abundance of Asp(D) and Glu(E) residues and an absence of Cys(C) residues was isolated from Momordica cochinchinensis root tubers. The protein is capable of inducing mid-term abortion in mice.50

Cochinin B

The sequence of the first 20 N-terminal amino acid residues of the 28-kDa cochinin B from Momordica cochinchinensis seeds is homologous to type I RIPs from other Momordica species. Cochinin B inhibits protein synthesis in the cell-free rabbit reticulocyte lysate system with an IC50 of 0.36 nM and exhibits N-glycosidase activity. It manifests cytotoxicity against human cervical epithelial carcinoma (HeLa), human embryonic kidney (HEK293) and human small cell lung cancer (NCI-H187) cells with IC50 of 16.9, 114, and 574 nM, respectively.51

Momorgrosvin

Momorgrosvin, a 27.7-kDa glycoprotein with a pI of about 9 from Momordica grosvenorii seeds, exhibits N-terminal amino acid sequence homology to other Momordica RIPs. It suppresses protein synthesis in the rabbit reticulocyte lysate system with an IC50 of 0.3 nM and displays RNA N-glycosidase activity giving rise to the diagnostic Endo’s band at a concentration as low as 9 nM.52

MOR-I and MOR-II

MOR-I and MOR-II from Marah oreganus (manroot) seeds are monomeric proteins exhibiting similarity to trichosanthin and crossreactivity with goat anti-trichosanthin polyclonal serum. They display molecular weights in the vicinity of 28 kDa and pI values over 8.8 and have in common a conserved N-terminal amino acid sequence. MOR-I and MOR-II inhibit cell-free protein synthesis with IC50s of 0.063 and 0.071 nM, respectively, and manifest moderate pH and temperature stability.53 Cucurmosin has been crystallized using polyethylene glycol as a precipitant. The crystals belong to space group P2(1)2(1)2(1) and have unit-cell parameters a = 41.91, b = 59. 48, c = 98.78 A. There is one molecule in the asymmetric unit. The diffraction data to 3.0 A resolution were collected on a MAR Research image-plate detector.

Moschatin

Moschatin is a 29-kDa protein with a pI of 9.4 and rRNA N-glycosidase activity. It curtails protein synthesis in the rabbit reticulocyte lysate with an IC50 of 0.26 nM. The immunotoxin formed from moschatin and anti-human melanoma monoclonal antibody Ng76 suppresses the growth of targeted melanoma cells M21 with an IC50 of 0.04 nM, 1500 times lower than that of free moschatin. The results implied that moschatin could be used as a new potential anticancer agent.54

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

51

Cucurbita Moschata RIP

Cucurbita moschata (pumpkin) RIP is a 30-kDa glycoprotein with antibacterial, antifungal, and superoxide dismutase activities. It reduces protein synthesis by a rabbit reticulocyte lysate with an IC50 of 0.035 nM. Protein synthesis by HeLa, HT29, and JM cells is inhibited with IC50 values in the 100 nM range. It deadenylates hsDNA and other polynucleotidic substrates, and depurinated yeast rRNA. The RIP crossreacts weakly with an antiserum against dianthin 32, but not with antisera against other RIPs.55

Pepocin

Pepocin from Cucurbita pepo fruits possesses a molecular weight of 26 kDa and a pI of about 9.9. It inhibited protein synthesis in a rabbit reticulocyte lysate with an IC50 of 15.4 pM, depurinated 28S rRNA, and was also active on wheat-germ and Escherichia coli ribosomes. Pepocin is localized in intercellular spaces in the sarcocarp and leaves.56

Sechiumin

The 27-kDa sechiumin from Sechium edule seeds inhibits protein synthesis in rabbit reticulocyte lysate and intact HeLa cells with an IC50 of 0.7 nM and 5000 nM, respectively. Sechiumin has a highly specific RNA N-glycosidase activity toward 28S rRNA. It displays approximately 60%, 62.5%, and 64.8% sequence similarity to luffin-a and trichosanthin. Glu160 and Arg163 are amino acid residues at the active site of sechiumin. The recombinant sechiumin was obtained as an insoluble protein, and the preparation of the active soluble form was achieved by renaturing the denatured protein. These studies suggest that the recombinant sechiumin could be used for the preparation of immunotoxin as a potential cancer chemotherapeutic agent.57

Trichoanguin

Trichoanguin from Trichosanthes anguina seeds, a 35-kDa glycoprotein with a pI of 9.1, exhibits 55%, 48%, 36%, and 34% amino acid sequence identity to those of trichosanthin, alphamomorcharin, ricin A-chain, and abrin A-chain, respectively. Potential glycosylation sites are located at Asn-51, Asn-65, Asn-201, and Asn-226. Its tertiary structure closely resembles those of trichosanthin and alpha-momorcharin. Its Cys-32 residue, which in all likelihood lies on the protein surface, could be conjugated with antibodies to form immunoconjugates for cancer therapy. It inhibits protein synthesis of rabbit reticulocyte lysate and HeLa cells with an IC50 of 0.08 nM and 6 uM, respectively.58

Trichobitacin

Trichobitacin inhibites HIV-1-induced syncytial cell formation and curtails HIV-1 p24 antigen expression and the number of HIV antigen positive cells in acutely, but not chronically, HIV-1 infected culture.59

52

RIBOSOME-INACTIVATING PROTEINS

Trichokirin

A 27-kDa basic glycoprotein designated as trichokirin from Trichosanthes kirilowii seeds, shows Notable N-terminal sequence homology to momordin. Trichokirin was conjugated, with the cleavable dimethyl 3, 3′-dithiobispropionimidate cross-linking reagent, to a monoclonal antibody directed against the Thy 1.2 antigen. This immunotoxin selectively destroys leukemia cells expressing the Thy 1.2 antigen. Ammonium chloride enhances the cytotoxicity of ricin A-chain immunotoxins but lowered that of the trichokirin immunotoxin, signifying that they employed different mechanisms to gain entry into cells. In vivo investigations disclosed that the trichokirin immunotoxin was superior to the ricin A-chain immunotoxins in pharmacokinetic properties for in vivo applications.60

Trichokirin-S1 (Small RIP)

Two stable strains of hybridomas (1 F11 and 2A5) produce highly specific monoclonal antibodies against trichokirin-S1, a small RIP from Trichosanthes kirilowii seeds. An immunoaffinity method using 2A5-coupled Sepharose 4B for the purification of trichokirin-S1 has been reported.61

Trichomaglin

Trichomaglin from Trichosanthes lepiniate root tubers, which has a molecular weight of 24.673 kDa and an isoelectric point of 5.8, inhibits protein synthesis in rabbit reticulocyte lysate with an IC50 of 10.1 nM. It manifests RNA N-glycosidase and abortifacient activities.62

Trichosanthin

Trichosanthin is a 26-kDa nonglycoprotein with immunosuppressive and abortifacient activities from Trichosanthes kirilowii root. It inhibits ConA-induced transformation in lymphocytes isolated from mouse spleens.63 The interaction between trichosanthin and recombinant ribosomal protein L10a was demonstrated by an in vitro binding assay. Kinetic analysis employing surface plasmon resonance technology discloses that L10a had a high affinity (K(D) = 7.78 nM) for trichosanthin. Using trichosanthin mutants, the correlation of this specific association with the ribosome-inactivating activity of trichosanthin was demonstrable.64 Trichosanthin exerts an immunosuppressive action analogous to that of momorcharins. It also suppresses the growth of a murine malignant tumour (MBL-2), in vitro as well as in vivo.65

Karasurin-A, Karasurin-B, and Karasurin-C

Karasurin-B and karasurin-C from Trichosanthes kirilowii root tubers Maximovicz var. japonica Kitamura have 247 and 249 amino acid residues, respectively. They exhibit amino acid sequence homology to karasurin-A and slightly dissimilar pI values. All karasurins potently inhibit in vitro translation in the rabbit reticulocyte system.66

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

53

A genomic DNA clone of karasurin was isolated from Trichosanthes kirilowii var. japonica using the polymerase chain reaction. The deduced amino acid sequence is consistent with sequences of karasurin-A and karasurin-C except for a putative signal peptide and extra C-terminal amino acids, neither of which exists in the natural protein. Recombinant karasurin was synthesized in Escherichia coli, in which the cloned karasurin gene was expressed under the control of the trc promoter.67

Trichosanthrip (Small RIP)

Trichosanthrip, an 11-kDa protein from mature Trichosanthes kirilowii seeds with a sequence similar to its other Cucurbitaceae counterparts, suppresses cell-free protein synthesis with an IC50 of 1.6 ng/ml. It displayed N-glycosidase activity toward 28 S rRNA.68

Euphorbiaceae RIPs Croton Tiglium and Jatropha Curcas Protein Synthesis Inhibitors

Proteins extracted from Croton tiglium and Jatropha curcas seeds were resolved into three major peaks (I, II, and III) by gel filtration. The crude protein extract and peak I and peak II from both seeds inhibits protein synthesis by a reticulocyte lysate, with peak II from both species demonstrating the highest inhibitory potency. However, there is no effect on protein synthesis by Ehrlich ascites cells in vitro.69

Curcin

Recombinant curcin, RIP from Jatropha curcas seeds, inhibits the growth of NCL-H446, SGC7901, and S180 tumor cells at 5 ug/ml.70

Gelonin

Gelonin is nontoxic to intact cells. Its isolation, characterization, and preparation of cytotoxic complexes with concanavalin A have been reported.71 Gelonin cross-linked with monoclonal antibodies by employing 2-iminothiolane is more cytotoxic than conjugates prepared using N-succinimidyl-3-(2-pyridylthio) propionate (SPDP) alone. Modification of a single amino group in gelonin with 2-iminothiolane brought about a 25–50% reduction of immunoreactivity and a 60–70% loss of protein synthesis inhibitory activity. Modification of two to three amino groups further hampers both immunoreactivity and protein synthesis inhibitory activity. Both N-succinimidyl 6-[3-(2-pyridyldithio) propionamido] hexanoate (long chain-SPDP) and SPDP modifications exert more pronounced effects on immunoreactivity and protein synthesis inhibitory activity compared to the similar ratio of 2-iminothiolane modification(s). Thus the positive charge plays an important role in the immunological as well as the protein synthesis inhibitory effect of gelonin.72

54

RIBOSOME-INACTIVATING PROTEINS

Lychnin

Lychnin from Lychnis chalcedonica seeds possesses 234 amino acid residues and all of the conserved amino acids in the RIP active site (Tyr69, Tyr119, Glu170, Arg173, and Trp203). A disulfide bridge is formed between Cys32 and Cys115. Cys214, which exists in the thiol form, is amenable for linking carrier molecules to generate immunotoxins and other conjugates.73

Mapalmin

Mapalmin, an RIP from Manihot palmata seeds, is a glycoprotein with an alkaline isoelectric point. It lowers protein synthesis by a rabbit reticulocyte lysate and phenylalanine polymerization by isolated ribosomes.25 The ribbon structures of representatives of the Caryophyllaceae, Cucurbitaceae, and Euphorbiaceae families are presented in Figure 4.1. Type 1 RIPs from other Families

Stirpe has written a comprehensive account of type 1 RIPs from different families.11

Amaranthin (Family Amaranthaceae)

Amaranthin from Amaranthus viridis leaves has a molecular weight of about 30 kDa and a pI of 9.8. It inhibits translation with an IC50 of 25 pM and tobacco mosaic virus infection on Nicotiana glutinosa leaves N-glycosidase activity on rRNA was present.74

Basella Rubra RIPs (Family Basellaceae) and Bougainvillea Spectabilis RIP (Family Nyctaginaceae)

RIPs from Basella rubra seeds and Bougainvillea spectabilis leaves inhibit protein synthesis in a cell-free system and by various cell lines. The RIPs display polynucleotide: adenosine glycosidase activity, releasing adenine not only from rat hepatic ribosomes but also from Escherichia coli rRNA, polyadenylic acid, herring sperm DNA, and artichoke mottled crinkle virus genomic RNA. Basella rubra RIPs have toxicity to mice similar to that of most type 1 RIPs with an LD50 at or lower than 8 mg/kg body weight. Bougainvillea spectabilis RIP had an LD50 above 32 mg/kg. The Basella RIPs, but not Bougainvillea spectabilis RIPs, cross-reacted with antisera against dianthin 32, momordin I and momorcochin-S. Both reduced infection of Nicotiana benthamiana by AMCV.75

Beta Vulgaris RIP (Family Chamnopodiaceae)

BE27 and BE29 are two forms of beetin, a virus-inducible type 1 RIP isolated from sugar beet (Beta vulgaris) leaves. Although beetin is present in adult plants but not in germ or young plants – suggesting

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

Dianthin (1RL0)

Bryodin (1BRY)

Beta-Luffin (1NIO)

Luffaculin (2OQA)

Saporin (3HIS)

Gelonin (3KU0)

Beta-Momorcharin (1CF5)

Alpha-Momorcharin (1AHA)

Trichosanthin (1QD2)

55

Figure 4.1 3D structure of ribosome-inactivating proteins (RIPs). In pictures, the yellow arrows are showing the secondary structures of beta-sheet, the green arrows are showing the secondary structures of alpha-helix. The Protein Data Bank IDs (PDB IDs) are quoted after the names of all RIPs. All pictures are generated by Cn3D. (This figure also appears in the color plate section.)

developmental regulation of expression – the transcript is detectable throughout development. Administration of mediators of plant-acquired resistance, including hydrogen peroxide and salicylic acid, to the leaves up-regulated beetin expression by transcript induction only in adult plants. The findings signify a dual regulation of beetin expression at the post-transcriptional and transcriptional levels. Sugar beet ribosomes are resistant to beetin.76

56

RIBOSOME-INACTIVATING PROTEINS

Nigritins f1 and f2 (Family Caprifoliaceae)

Two highly basic type 1 RIPS with N-glycosidase activity have been found in elderberries (Sambucus nigra fruits). Both proteins potently inhibit protein synthesis in rabbit reticulocyte lysates but have no activity against plant ribosomes. Nigritin f1 (molecular weight 24 kDa) is present in both immature and mature intact elderberries at nearly the same proportion with reference to total fruit protein. Nigritin f2 (molecular weight 23.5 kDa) is inducible and present only in mature intact elderberries.77

Charybdin (Family Hyacinthaceae)

Charybdin from Charybdis maritima bulbs is a 29-kDa protein that possesses Val at position 79 in lieu of the conserved Tyr as the active site residue. This replacement may account for the reduction in translation inhibitory activity (IC50 = 7.2 nM) compared with other RIPS. The fold of the protein comprises two structural domains: an alpha plus beta N-terminal domain (encompassing residues 4–190) and a mainly alpha-helical C-terminal domain (consisting of residues 191–257), with the active site located in the interface between the two domains and composed of residues Val79, Tyr117, Glu167, and Arg170.78

Iris RIP (Family Iridaceae)

Recombinant iris RIP from tobacco (Nicotiana tabacum cv. Samsun NN) leaves expressed under the control of the 35S cauliflower mosaic virus promoter is identical to the native protein from iris bulbs in molecular structure and RNA N-glycosidase activity. The tobacco transformants do not exhibit phenotypic side-effects, indicating that ectopically expressed iris RIP is not toxic to tobacco cells. The in planta antiviral activity of the RIP was demonstrated using a tobacco mosaic virus bioassay.79

Camphorin (Family Lauraceae)

Camphorin is a 23-kDa glycoprotein from Cinnamomum camphora seeds. It co-exists in the seeds with cinnamomin, a 61-kDa type 2 RIP with three isoforms. Camphorin is capable of cleaving supercoiled double-stranded DNA into nicked and linear forms80 and displays superoxide dismutase activity.81

Mirabilis Jalapa RIP (Family Nyctaginaceae)

Mirabilis antiviral protein (MAP) from Mirabilis jalapa root tubers has demonstrated abortifacient activity in pregnant mice, an inhibitory effect on cell-free protein synthesis, and an antiproliferative effect on tumor cells.82

Barley (Family Poaceae)

A 30-kDa RIP, a 26-kDa chitinase, and a 32-kDa (1–3)-beta-glucanase from barley (Hordeum vulgare) seeds demonstrates synergistic antifungal activity. There was a differential accumulation

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

57

of their mRNAs during seed development and germination. Chitinase mRNA accumulates in aleurone cells during late seed development and early germination. The RIP mRNA reaches high levels only in the starchy endosperm during late seed development. The glucanase mRNA is present at low levels during seed development and increases in aleurone and seedling tissues during germination.83

Wheat (Family Poaceae)

Three forms of the wheat protein synthesis inhibitor tritin (tritin 1, 2, and 3) identical in molecular weight and enzymatic properties were found. Antiserum against the most abundant form, tritin 2, crossreacts with tritin 1, tritin 3, and with barley and rye RIPs, but not with their counterparts in other plants.84 Algal RIPs Lamjapin (Family Laminariaceae)

Lamjapin from the marine alga Laminaria japonica has a molecular weight of approximately 36 kDa, and inhibits protein synthesis in rabbit reticulocyte lysate with an IC50 of 0.69 nM. It depurinates rat ribosomal RNA at multiple sites. It deadenylates specifically at the site A20 of the synthetic oligoribonucleotide (35-mer) substrate that mimics the rat ribosomal 28S RNA. However, it cannot hydrolyze the N–C glycosidic bond of guanosine, cytidine, or uridine at the corresponding site of the A20 of three mutant sarcin/ricin domain RNAs. Lamjapin exhibits the same base and position requirement as the RIPS from higher plants.85 Fungal RIPs Clavin (Family Trichocomaceae)

Clavin from the filamentous fungus Aspergillus clavatus IFO 8605 induces specific cleavage of ribosomal and synthetic RNA and suppresses protein synthesis in cell-free and cellular systems. When selectively targeted to a tumor cell antigen by coupling to a monoclonal antibody, clavin inhibits protein synthesis at nanomolar levels. Pharmacokinetics analysis in mice reveals an elimination half-life of 7.4 h. Hepatotoxicity and renal toxicity are either minimal or undetectable. Clavin induces a late and only slight antibody response in mice.86

Flammin and Velin (Family Tricholomataceae)

The 30-kDa flammin and 19-kDa velin from Flammulina velutipes inhibits translation in a rabbit reticulocyte lysate system with an IC50 of 1.4 and 2.5 nM, respectively. Flammin exhibits only slight resemblance in N-terminal sequence to some angiosperm type 1 RIPs including alpha-momorcharin, beta-momorcharin, and trichosanthin, but there is no similarity to other mushroom RIPs. Velin manifests limited sequence homology to the A chain of abrin, a type 2 angiosperm RIP.87

58

RIBOSOME-INACTIVATING PROTEINS

Flammulin (Family Tricholomataceae)

The 40-kDa flammulin from Flammulina velutipes inhibits cell-free translation in a rabbit reticulocyte lysate system with an IC50 of 0.25 nM.88

Hypsin (Family Tricholomataceae)

The 20-kDa hypsin from Hypsizigus marmoreus inhibits translation in the rabbit reticulocyte lysate system with an IC50 of 7 nM and HIV-1 reverse transcriptase activity with an IC50 of 8 uM. It impedes mycelial growth in Mycosphaerella arachidicola, Physalospora piricola, Fusarium oxysporum, and Botrytis cinerea with an IC50 of 2.7, 2.5, 14.2, and 0.06 uM, respectively. Antiproliferative activity against mouse leukemia cells and human leukemia and hepatoma cells was observed. About 60% of the translation-inhibitory activity remained after heating at 100 °C for 10 min. No reduction of translation-inhibitory activity is detected after brief treatment with trypsin.89

Lyophyllin (Family Tricholomataceae)

The 20-kDa lyophyllin from Lyophyllum shimeji with some N-terminal sequence similarity to those of plant ribosome-inactivating proteins exhibits antifungal activity against Physalospora piricola and Coprinus comatus. It inhibits translation in rabbit reticulocyte lysate with an IC50 of 1 nM, thymidine uptake by murine splenocytes with an IC50 of 1 mM, and HIV-1 reverse transcriptase activity with an IC50 of 7.9 nM. A synergism in antifungal activities of Lyophyllum antifungal protein and lyophyllin against P. piricola was demonstrable.90

Marmorin (Family Tricholomataceae)

Marmorin, a 9,567-Da RIP with a novel N-terminal sequence from Hypsizigus marmoreus, inhibits translation in the rabbit reticulocyte lysate system, proliferation of breast cancer MCF-7 cells and hepatoma Hep G2 cells, HIV-1 reverse transcriptase activity, and with an IC50 of 0.7 nM, 5 uM, 0.15 uM, and 30 uM, respectively. Compared to RIPs from garden pea, bitter gourd, hairy gourd, ridge gourd, and the mushroom Flammulina velutipes, marmorin was more potent in its antiproliferative activity toward hepatoma (HepG2) and breast cancer (MCF-7) cells, similar in inhibitory potency toward HIV-1 reverse transcriptase (with the exception that it was more potent than ridge gourd RIP and bitter gourd RIP), and less potent in translation-inhibitory potency.91

Pleuturegin (Family Pleurotaceae)

The 38-kDa pleuturegin from Pleurotus tuber-regium demonstrates an N-terminal sequence that is different from those of RIPs from other mushroom RIPs with known N-terminal sequences.

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

59

It inhibits translation in a cell-free rabbit reticulocyte lysate system with an IC50 of 0.5 nM Pleuturegin. 92

Velutin (Family Tricholomataceae)

The 13.8-kDa velutin from Flammulina velutipes inhibits HIV-1 reverse transcriptase, betaglucosidase and beta-glucuronidase. Its N-terminal sequence exhibits some similarity to those of plant RIPs.93

Volvarin (Family Plutaceae)

The 29-kDa mushroom RIP, designated volvarin, exhibits a potent inhibitory action on protein synthesis in the rabbit reticulocyte lysate system with an IC50 value of 0.5 nM. It acts as an N-glycosidase that depurinates rRNA from rabbit reticulocyte lysate, releasing a characteristic RNA fragment (Endo’s band) after treatment with aniline. It also exerts deoxyribonuclease activity on supercoiled SV-40 DNA and demonstrates a strong abortifacient effect in mice.94

Crystal Structures of RIPs

Single crystals of the RIP saporin from Saponaria officinalis seeds grown by the vapor-diffusion method using ammonium sulfate as precipitant are tetragonal, space group P4122 (P4322), with cell dimensions a = b = 67.53 and c = 119. 67 A, and diffract to 2.0 A resolution on a rotating-anode X-ray source. The asymmetric unit contains one molecule, corresponding to a volume of the asymmetric unit per unit mass (Vm) of 2.38 A3 Da-1.95

Production of Recombinant RIPs

The production of recombinant dianthin in Escherichia coli has been described. 96 Purification and functional activity studies demonstrate that recombinant gelonin expressed in E. coli is identical to the natural protein.97 Mirabilis expansa RIP ME1 has also been produced in E. coli.98

Immunoreactivity of RIPs Effects of Thiolation on the Immunoreactivity of the RIP Gelonin

The cross-linking agent N-succinimidyl-3-(2-pyridyldithio) propionate (SPDP) has been used for attaching gelonin via its epsilon-NH2 group to its carrier antibodies or hormones for immunotoxin or hormonotoxin, respectively. The consequence of this modification on the immunoreactivity of gelonin has been examined. Even one or two modifications, representing 1/20 to 1/10 of available epsilon-NH2 groups in gelonin brings about a 75% decline in immunoreactivity. Further modifications result in greater reduction.99

60

RIBOSOME-INACTIVATING PROTEINS

A complement-mediated liposome immune lysis assay employing entrapped calcein was developed for gelonin with a detection range from 3 ng to 60 ng. Gelonin was covalently coupled to DPPE, and then adsorbed on to the liposome surface. These antigen-bearing liposomes were lysed upon incubation with gelonin antibody in the presence of a guinea pig complement. Ricin A chain but not intact ricin crossreacts with gelonin antibody. The assay may be utilized for detecting and quantitating ricin A chain.100

Other Activities of Ribosome-inactivating Proteins

None of the three abortifacient proteins affect lipogenic activity in isolated rat adipocytes, plasmaglucose level in fasting mice, luteinizing hormone-induced testosterone production in isolated rat Leydig cells, or corticotropin-induced corticosterone production in isolated rat adrenal decapsular cells. The results suggest that the functions of adipocytes, adrenal decapsular cells, Leydig cells, and pancreatic beta cells are not greatly affected.101 After intravenous administration into normotensive rats via the external jugular vein, the RIPs saporin, trichosanthin, gelonin, and momordin evoke a mild hypotensive response while luffin and agrostin are inactive. The hypotensive response, however, lacks dose dependence. The RIPs trichosanthin, momordin, and gelonin do not affect the blood pressure response to angiotensin I. However, alpha-Momorcharin, beta-momorcharin, and trichosanthin induce morphological changes in hepatocytes including increased formation of cytoplasmic blebs and a reduction of microvilli on plasma membrane and increase the secretion of isocitrate dehydrogenase, glutamate-pyruvate transaminase, and lactate dehydrogenase. 102 In immature mice induced to ovulate by pregnant serum gonadotropin and human chorionic gonadotropin, luffaculin, luffin-a, luffin-b, and momorcochin elicit a reduction in the circulating titer of estradiol-17 beta. Both luffaculin and momorcochin bring about an increase in the number of degenerated oocytes. Only momorcochin causes a decrease in the ovarian weight and the number of ovulated oocytes.103 Trichosanthin does not affect the number of maturing follicles, corpora lutea, and ovulated oocytes, or the ovarian weight. However, there is an increased incidence of follicular atresia and degeneration of ovulated oocytes, and a lowering of serum estradiol-17 beta and progesterone levels.104 Ribosome inactivating proteins also display antitumor,105, 106 anti-insect,107, 108 antifungal,109 antiviral, 110 neuronotoxic,111 and translation inhibitory112 activities.

Conclusion

From the foregoing account it can be seen that type 1 RIPs are produced by plants in a variety of  families. A sizeable number of type 1 RIPs has been isolated from Cucurbitaceae plants. A lesser number has been reported from plants belonging to Caryophyllaceae and Euphorbiaceae families. Plants in Amaranthaceae, Basellaceae, Caprifoliaceae, Chamnopodiaceae, Hyacinthaceae, Iridaceae, Lauraceae, Nyctaginaceae, and Poaceae families, algae in Laminariaceae family and fungi in Tricholomataceae, Plurotaceae, and Plutaceae families also produce type 1 RIPs (Table 4.1). Attempts have been made to unravel the mechanisms of protein synthesis inhibitory, antitumor, antifungal, antiviral, anti-insect, and neuronotoxic activities of RIPs.3, 105–112

61

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

Table 4.1

Summary of RIPs mentioned in this article.

RIP Name Caryophyllaceae family Agrostin 2,5, and 6 Dianthin 30 Dianthin 32 Gypsophilin Petroglaucin 1 Petroglaucin Saporin L1,L2, R1-3, S5, S6, S8, and S9 Saponaria ocymoides RIP Vaccaria pyramidata RIP Cucurbitaceae family Bryodin Colocin 1 and colocin 2 Cucumis melo RIP Cucurmosin Hispin Lagenin Luffangulin 19 K, 15 K, and 9 K-PSI Luffin a Luffacylin LuffinS (1) LuffinS (2) LuffinS (3) Luffin P1 Balsamin MRK29 MAP30 Alpha- and betamomorcharins Charantin Momorcochin 5 Cochinin B Momorgrosvin MOR-I MOR-II Moschatin Pepocin Sechiumin Trichoanguin Trichobitacin Trichokirin Trichokirin-S1

Species Name

Inhibitory activity on cell-free protein synthesis IC50 (nM)

Agrostemma githago Dianthus caryophyllus Dianthus caryophyllus Gypsophila elegans Petrocoptis glaucifolia Petrocoptis grandiflora

0.57–0.60 0.31 0.11 0.04 (Rat liver) subnanomolar subnanomolar

Saponaria officinalis

0.037–0.86

19–22

Saponaria ocymoides

below 1 nM

23

Vaccaria pyramidata

below 1 nM

23

Bryonia dioica

0.12

24

Citrullus colocynthis Cucumis melo Cucurbita moschata Benincasa hispida Lagenaria siceraria Luffa cylindrica

0.04 and 0.13 – – 0.165 0.21 3.5

25 26 27 28 29 30

Luffa cylindrica Luffa cylindrica Luffa cylindrica Luffa cylindrica Luffa cylindrica Luffa cylindrica Luffa cylindrica Momordica blasamina Momordica charantia Momordica charantia

0.065, 360, and 140 1 0.14 0.13 1 0.63 0.6 90.6 ng/ml – 3.3

31 34 37 38 38 38 40 42 44 45

Momordica charantia Momordica charantia Momordica cochinchinensis Momordica cochinchinensis Momordica grosvenorii Marah oreganus Marah oreganus Cucurbita moschata Cucurbita moschata Cucurbita pepo Sechium edule Trichosanthes anguina Trichosanthes kirilowii Trichosanthes kirilowii Trichosanthes kirilowii

– 0.4 0.12 0.36 0.3 0.063 0.071 0.26 0.035 0.0154 0.7 0.08 – 0.113 0.7

Reference 15 16 16 17 18 18

46–48 49 50 51 52 53 53 54 55 56 57 58 59 60 61

62

RIBOSOME-INACTIVATING PROTEINS

Table 4.1

(Continued)

RIP Name

Species Name

Inhibitory activity on cell-free protein synthesis IC50 (nM)

Trichomaglin Trichosanthin Karasurin-A, -B and -C Trichosanthrip

Trichosanthes lepiniate Trichosanthes kirilowii

10.1 0.25

Trichosanthes kirilowii Trichosanthes kirilowii

– 1.6 ng/ml

66 68

Euphorbiaceae family Croton tiglium RIP Jatropha curcas Curcin 2,3 and 4 Gelonin Lychnin Mapalmin

Croton tiglium Jatropha curcas Jatropha curcas Gelonium multiflorum Lychnis chalcedonica Manihot palmata

– – 0.09–0.19 0.4 0.17 0.05

69 69 70 71,72 73 25

Reference 62 63,64

Other families Amaranthin Basella rubra RIP Beta vulgaris RIP Nigritins f1 and f2 Charybdin Iris RIP Camphorin MAP Hordeum vulgare RIP Tritin 1, 2 and 3

Amaranthus viridis Basella rubra Beta vulgaris Sambucus nigra Charybdis maritima Iris hollandica Cinnamomum caphora Mirabilis jalapa

0.025 0.057 – – 7.2 – – 0.19

74 75 76 77 78 79 81 82

Hordeum vulgare Triticum vulgaris

– –

83 84

Algal RIP Lamjapin

Laminaria japonica

0.69

85

Fungal RIP Clavin Flammin Velin Flammulin Hypsin Lyophyllin Marmorin Pleuturegin Velutin Volvarin

Aspergillus clavatus Flammulina velutipes Flammulina velutipes Flammulina velutipes Hypsizigus marmoreus Lyophyllum shimeji Hypsizigus marmoreus Pleurotus tuber-regium Flammulina velutipes Volvariella volvacea

0.14 1.4 2.5 0.25 7 1 0.7 0.5 0.29 0.5

86 87 87 88 89 90 91 92 93 94

“–” No information. Some of the information presented in this table is derived from references 113 and 114.

References 1. 2. 3.

Reyes AG, Anne J, Mejia A. Ribosome-inactivating proteins with an emphasis on bacterial RIPs and their potential medical applications. Future Microbiol. 2012;7:705–717. Puri M, Kaur I, Perugini MA, Gupta RC. Ribosome-inactivating proteins: current status and biomedical applications. Drug Discov Today. 2012;17:774–783. Kaur I, Gupta RC, Puri M. Ribosome inactivating proteins from plants inhibiting viruses. Virol Sin. 2011;26:357–365.

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

63

Fang EF, Ng TB, Shaw PC, Wong RN. Recent progress in medicinal investigations on trichosanthin and other ribosome inactivating proteins from the plant genus Trichosanthes. Curr Med Chem. 2011;18:4410–4417. Ng TB, Wong JH, Wang H. Recent progress in research on ribosome inactivating proteins. Curr Protein Pept Sci. 2010;11:37–53. Puri M, Kaur I, Kanwar RK, et al. Ribosome inactivating proteins (RIPs) from Momordica charantia for anti viral therapy. Curr Mol Med. 2009;9:1080–1094. Kozlov Iu V, Sudarkina O, Kurmanova AG. [Ribosome-inactivating lectins from plants]. Mol Biol. (Mosk) 2006;40:711–723. Stirpe F, Battelli MG. Ribosome-inactivating proteins: progress and problems. Cell Mol Life Sci. 2006;63:1850–1866. Robertus JD, Monzingo AF. The structure of ribosome inactivating proteins. Mini Reviews in Medicinal Chemistry. 2004;4:477–486. Park SW, Vepachedu R, Sharma N, Vivanco JM. Ribosome-inactivating proteins in plant biology. Planta. 2004;219: 1093–1096. Stirpe F. Ribosome-inactivating proteins. Toxicon. 2004;44:371–383. Battelli MG. Cytotoxicity and toxicity to animals and humans of ribosome-inactivating proteins. Mini Reviews in Medicinal Chemistry. 2004;4:513–521. Roberts LM, Lord JM. Ribosome-inactivating proteins: Entry into mammalian cells and intracellular routing. Mini Reviews in Medicinal Chemistry. 2004;4:505–512. Girbes T, Ferreras JM, Arias FJ, Stirpe F. Description, distribution, activity and phylogenetic relationship of ribosomeinactivating proteins in plants, fungi and bacteria. Mini Reviews in Medicinal Chemistry. 2004;4:461–476. Chiu LC, Ooi VE, Sun SS. Induction of apoptosis by a ribosome-inactivating protein from Agrostemma githago is associated with down-regulation of anti-apoptotic bcl-2 protein expression. Int J Oncol. 2001;19:137–141. Stirpe F, Williams DG, Onyon LJ, et al. Dianthins, ribosome-damaging proteins with anti-viral properties from Dianthus caryophyllus L. (carnation). Biochem J. 1981;195:399–405. Yoshinari S, Koresawa S, Yokota S, et al. Gypsophilin, a new type 1 ribosome-inactivating protein from Gypsophila elegans: purification, enzymatic characterization, and subcellular localization. Biosci Biotechnol Biochem. 1997;61:324–331. Arias FJ, Rojo MA, Ferreras JM, et al. Isolation and characterization of two new N-glycosidase type-1 ribosome-inactivating proteins, unrelated in amino-acid sequence, from Petrocoptis species. Planta. 1994;194:487–491. Lappi DA, Esch FS, Barbieri L, et al. Characterization of a Saponaria officinalis seed ribosome-inactivating protein: immunoreactivity and sequence homologies. Biochem Biophys Res Commun. 1985;129:934–942. Poma A, Zarivi O, Bianchini S, Spano L. The plant ribosome inactivating protein saporin induces micronucleus formation in peripheral human lymphocytes in vitro. Toxicol Lett. 1999;105:67–73. Wiley RG. Targeting toxins to neural antigens and receptors. Methods Mol Biol. 2001;166:267–276. Gunhan E, Swe M, Palazoglu M, et al. Expression and purification of cysteine introduced recombinant saporin. Protein Expr Purif. 2008;58:203–209. Bolognesi A, Olivieri F, Battelli MG, et al. Ribosome-inactivating proteins (RNA N-glycosidases) from the seeds of Saponaria ocymoides and Vaccaria pyramidata. Eur J Biochem. 1995;228:935–940. Stirpe F, Barbieri L, Battelli MG, et al. Bryodin, a ribosome-inactivating protein from the roots of Bryonia dioica L. (white bryony). Biochem J. 1986;240:659–665. Bolognesi A, Barbieri L, Abbondanza A, et al. Purification and properties of new ribosome-inactivating proteins with RNA N-glycosidase activity. Biochim Biophys Acta. 1990;1087:293–302. Rojo MA, Arias FJ, Ferreras JM, et al. Partial characterization of the translational inhibitor present in seeds of Cucumis melo L. Biochem Soc Trans. 1992;20:313S. Hou X, Meehan EJ, Xie J, et al. Atomic resolution structure of cucurmosin, a novel type 1 ribosome-inactivating protein from the sarcocarp of Cucurbita moschata. J Struct Biol. 2008;164:81–87. Ng TB, Parkash A. Hispin, a novel ribosome inactivating protein with antifungal activity from hairy melon seeds. Protein Expr Purif. 2002;26:211–217. Wang HX, Ng TB. Lagenin, a novel ribosome-inactivating protein with ribonucleolytic activity from bottle gourd (Lagenaria siceraria) seeds. Life Sci. 2000;67:2631–2638. Wang H, Ng TB. Luffangulin, a novel ribosome inactivating peptide from ridge gourd (Luffa acutangula) seeds. Life Sci. 2002;70:899–906. Watanabe K, Minami Y, Funatsu G. Isolation and partial characterization of three protein-synthesis inhibitory proteins from the seeds of Luffa cylindrica. Agric Biol Chem. 1990;54:2085–2092. Singh RC, Alam A, Singh V. Purification,characterization and chemical modification studies on a translation inhibitor protein from Luffa cylindrica. Indian J Biochem Biophys. 2003;40:31–39. Wang HX, Ng TB. Studies on the anti-mitogenic, anti-phage and hypotensive effects of several ribosome inactivating proteins. Comp Biochem Physiol C Toxicol Pharmacol. 2001;128:359–366.

64 34. 35. 36.

37. 38. 39. 40. 41. 42. 43. 44. 45. 46.

47. 48. 49. 50. 51. 52. 53. 54.

55. 56. 57. 58. 59. 60.

RIBOSOME-INACTIVATING PROTEINS

Ng TB, Wong RN, Yeung HW. Two proteins with ribosome-inactivating, cytotoxic and abortifacient activities from seeds of Luffa cylindrica roem (Cucurbitaceae). Biochem Int. 1992;27:197–207. Islam MR, Nishida H, Funatsu G. Complete amino acid sequence of luffin-a, a ribosome-inactivating protein from the seeds of sponge gourd (Luffa cylindrica). Agric Biol Chem. 1990;54:2967–2978. Minami Y, Islam MR, Funatsu G. Chemical modifications of momordin-a and luffin-a, ribosome-inactivating proteins from the seeds of Momordica charantia and Luffa cylindrica: involvement of His140, Tyr165, and Lys231 in the protein-synthesis inhibitory activity. Biosci Biotechnol Biochem. 1998;62:959–964. Parkash A, Ng TB, Tso WW. Isolation and characterization of luffacylin, a ribosome inactivating peptide with anti-fungal activity from sponge gourd (Luffa cylindrica) seeds. Peptides. 2002;23:1019–1024. Xiong CY, Zhang ZC. Isolation, purification and characterization of a group of novel small molecular ribosome inactivating protein – Luffin S from seeds of luffa cylindrica. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao (Shanghai). 1998;30:142–146. Gao W, Ling J, Zhong X, et al. Luffin-S–a small novel ribosome-inactivating protein from Luffa cylindrica. Characterization and mechanism studies. FEBS Lett. 1994;347:257–260. Li F, Yang XX, Xia HC, et al. Purification and characterization of Luffin P1, a ribosome-inactivating peptide from the seeds of luffa cylindrica. Peptides. 2003;24:799–805. Wang R, Gan C, Gao W, et al. A novel recombinant immunotoxin with the smallest ribosome-inactivating protein Luffin P1: T-cell cytotoxicity and prolongation of allograft survival. J Cell Mol Med. 2010;14:578–586. Kaur I, Yadav SK, Hariprasad G, et al. Balsamin, a novel ribosome-inactivating protein from the seeds of Balsam apple Momordica balsamina. Amino Acids. 2012;43:973–981. Ortigao M, Better M. Momordin II, a ribosome inactivating protein from Momordica balsamina, is homologous to other plant proteins. Nucleic Acids Res. 1992;20:4662. Jiratchariyakul W, Wiwat C, Vongsakul M, et al. HIV inhibitor from Thai bitter gourd. Planta Med. 2001;67:350–353. Lee-Huang S, Huang PL, Chen HC, et al. Anti-HIV and anti-tumor activities of recombinant MAP30 from bitter melon. Gene. 1995;161:151–156. Fong WP, Poon YT, Wong TM, et al. A highly efficient procedure for purifying the ribosome-inactivating proteins alpha- and beta-momorcharins from Momordica charantia seeds, N-terminal sequence comparison and establishment of their N-glycosidase activity. Life Sci. 1996;59:901–909. Kimura Y, Minami Y, Tokuda T, et al. Primary structures of N-linked oligosaccharides of momordin-a, a ribosomeinactivating protein from Momordica charantia seeds. Agric Biol Chem. 1991;55:2031–2036. Leung SO, Yeung HW, Leung KN. The immunosuppressive activities of two abortifacient proteins isolated from the seeds of bitter melon (Momordica charantia). Immunopharmacology. 1987;13:159–171. Parkash A, Ng TB, Tso WW. Purification and characterization of charantin, a napin-like ribosome-inactivating peptide from bitter gourd (Momordica charantia) seeds. J Pept Res. 2002;59:197–202. Yeung HW, Ng TB, Li WW, Cheung WK. Partial chemical characterization of alpha- and beta-momorcharins. Planta Med. 1987;53:164–166. Chuethong J, Oda K, Sakurai H, et al. Cochinin B, a novel ribosome-inactivating protein from the seeds of Momordica cochinchinensis. Biol Pharm Bull. 2007;30:428–432. Tsang KY, Ng TB. Isolation and characterization of a new ribosome inactivating protein, momorgrosvin, from seeds of the monk’s fruit Momordica grosvenorii. Life Sci. 2001;68:773–784. Shih NJ, McDonald KA, Girbes T, et al. Ribosome-inactivating proteins (RIPs) of wild Oregon cucumber (Marah oreganus). Biol Chem. 1998;379:721–725. Xia HC, Li F, Li Z, Zhang ZC. Purification and characterization of Moschatin, a novel type I ribosome-inactivating protein from the mature seeds of pumpkin (Cucurbita moschata), and preparation of its immunotoxin against human melanoma cells. Cell Res. 2003;13:369–374. Barbieri L, Polito L, Bolognesi A, et al. Ribosome-inactivating proteins in edible plants and purification and characterization of a new ribosome-inactivating protein from Cucurbita moschata. Biochim Biophys Acta. 2006;1760:783–792. Yoshinari S, Yokota S, Sawamoto H, et al. Purification, characterization and subcellular localization of a type-1 ribosomeinactivating protein from the sarcocarp of Cucurbita pepo. Eur J Biochem. 1996;242:585–591. Wu TH, Chow LP, Lin JY. Sechiumin, a ribosome-inactivating protein from the edible gourd, Sechium edule Swartz–purification, characterization, molecular cloning and expression. Eur J Biochem. 1998;255:400–408. Chow LP, Chou MH, Ho CY, et al. Purification, characterization and molecular cloning of trichoanguin, a novel type Iribosome-inactivating protein from the seeds of Trichosanthes anguina. Biochem J. 1999;338 ( Pt 1):211–219. Zheng YT, Ben KL, Jin SW. Anti-HIV-1 activity of trichobitacin, a novel ribosome-inactivating protein. Acta Pharmacol Sin. 2000;21:179–182. Casellas P, Dussossoy D, Falasca AI, et al. Trichokirin, a ribosome-inactivating protein from the seeds of Trichosanthes kirilowii Maximowicz. Purification, partial characterization and use for preparation of immunotoxins. Eur J Biochem. 1988;176:581–588.

RIBOSOME-INACTIVATING PROTEINS IN CARYOPHYLLACEAE

61.

62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80.

81. 82. 83. 84. 85. 86.

87.

65

Yang XX, Li F, Hu WG, et al. Preparation and preliminary application of monoclonal antibodies against Trichokirin-S1, a small ribosome-inactivating peptide from the seeds of Trichosanthes kirilowii. Acta Biochim Biophys Sin. (Shanghai) 2005;37:447–452. Chen R, Xu YZ, Wu J, et al. Purification and characterization of trichomaglin–a novel ribosome-inactivating protein with abortifacient activity. Biochem Mol Biol Int. 1999;47:185–193. Yeung HW, Ng TB, Wong NS, Li WW. Isolation and characterization of an abortifacient protein, momorcochin, from root tubers of Momordica cochinchinensis (family cucurbitaceae). Int J Pept Protein Res. 1987;30:135–140. Xia X, Hou F, Li J, Nie H. Ribosomal protein L10a, a bridge between trichosanthin and the ribosome. Biochem Biophys Res Commun. 2005;336:281–286. Leung KN, Yeung HW, Leung SO. The immunomodulatory and antitumor activities of trichosanthin-an abortifacient protein isolated from tian-hua-fen (Trichosanthes kirilowii). Asian Pac J Allergy Immunol. 1986;4:111–120. Kondo T, Mizukami H, Takeda T, Ogihara Y. Amino acid sequences and ribosome-inactivating activities of karasurin-B and karasurin-C. Biol Pharm Bull. 1996;19:1485–1489. Mizukami H, Iida K, Kondo T, Ogihara Y. Cloning and bacterial expression of a gene encoding ribosome-inactivating proteins, karasurin-A and karasurin-C, from Trichosanthes kirilowii var. japonica. Biol Pharm Bull. 1997;20:711–713. Shu SH, Xie GZ, Guo XL, Wang M. Purification and characterization of a novel ribosome-inactivating protein from seeds of Trichosanthes kirilowii Maxim. Protein Expr Purif. 2009;67:120–125. Stirpe F, Pession-Brizzi A, Lorenzoni E, et al. Studies on the proteins from the seeds of Croton tiglium and of Jatropha curcas. Toxic properties and inhibition of protein synthesis in vitro. Biochem J. 1976;156:1–6. Luo MJ, Yang XY, Liu WX, et al. Expression, purification and anti-tumor activity of curcin. Acta Biochim Biophys Sin. (Shanghai) 2006;38:663–668. Stirpe F, Olsnes S, Pihl A. Gelonin, a new inhibitor of protein synthesis, nontoxic to intact cells. Isolation, characterization, and preparation of cytotoxic complexes with concanavalin A. J Biol Chem. 1980;255:6947–6953. Singh RC, Alam A, Singh V. Role of positive charge of lysine residue on ribosome-inactivating property of gelonin. Indian J Biochem Biophys. 2001;38:309–312. Chambery A, de Donato A, Bolognesi A, et al. Sequence determination of lychnin, a type 1 ribosome-inactivating protein from Lychnis chalcedonica seeds. Biol Chem. 2006;387:1261–1266. Kwon SY, An CS, Liu JR, Paek KH. A ribosome-inactivating protein from Amaranthus viridis. Biosci Biotechnol Biochem. 1997;61:1613–1614. Bolognesi A, Polito L, Olivieri F, et al. New ribosome-inactivating proteins with polynucleotide:adenosine glycosidase and antiviral activities from Basella rubra L. and bougainvillea spectabilis Willd. Planta. 1997;203:422–429. Iglesias R, Perez Y, Citores L, et al. Elicitor-dependent expression of the ribosome-inactivating protein beetin is developmentally regulated. J Exp Bot. 2008;59:1215–1223. de Benito FM, Iglesias R, Ferreras JM, et al. Constitutive and inducible type 1 ribosome-inactivating proteins (RIPs) in elderberry (Sambucus nigra L.). FEBS Lett. 1998;428:75–79. Touloupakis E, Gessmann R, Kavelaki K, et al. Isolation, characterization, sequencing and crystal structure of charybdin, a type 1 ribosome-inactivating protein from Charybdis maritima agg. FEBS J. 2006;273:2684–2692. Desmyter S, Vandenbussche F, Hao Q, et al. Type-1 ribosome-inactivating protein from iris bulbs: a useful agronomic tool to engineer virus resistance? Plant Mol Biol. 2003;51:567–576. Ling J, Liu WY, Wang TP. Simultaneous existence of two types of ribosome-inactivating proteins in the seeds of Cinnamonum camphora–characterization of the enzymatic activities of these cytotoxic proteins. Biochim Biophys Acta. 1995;1252: 15–22. Li XD, Chen WF, Liu WY, Wang GH. Large-scale preparation of two new ribosome-inactivating proteins–cinnamomin and camphorin from the seeds of Cinnamomum camphora. Protein Expr Purif. 1997;10:27–31. Wong RN, Ng TB, Chan SH, et al. Characterization of Mirabilis antiviral protein–a ribosome inactivating protein from Mirabilis jalapa L. Biochem Int. 1992;28:585–593. Leah R, Tommerup H, Svendsen I, Mundy J. Biochemical and molecular characterization of three barley seed proteins with antifungal properties. J Biol Chem. 1991;266:1564–1573. Reisbig RR, Bruland O. The protein synthesis inhibitors from wheat, barley, and rye have identical antigenic determinants. Biochem Biophys Res Commun. 1983;114:190–196. Liu RS, Yang JH, Liu WY. Isolation and enzymatic characterization of lamjapin, the first ribosome-inactivating protein from cryptogamic algal plant (Laminaria japonica A). Eur J Biochem. 2002;269:4746–4752. Parente D, Raucci G, Celano B, et al. Clavin, a type-1 ribosome-inactivating protein from Aspergillus clavatus IFO 8605. cDNA isolation, heterologous expression, biochemical and biological characterization of the recombinant protein. Eur J Biochem. 1996;239:272–280. Ng TB, Wang HX. Flammin and velin: new ribosome inactivating polypeptides from the mushroom Flammulina velutipes. Peptides. 2004;25:929–933.

66 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111.

112.

RIBOSOME-INACTIVATING PROTEINS

Wang HX, Ng TB. Flammulin: a novel ribosome-inactivating protein from fruiting bodies of the winter mushroom Flammulina velutipes. Biochem Cell Biol. 2000;78:699–702. Lam SK, Ng TB. Hypsin, a novel thermostable ribosome-inactivating protein with antifungal and antiproliferative activities from fruiting bodies of the edible mushroom Hypsizigus marmoreus. Biochem Biophys Res Commun. 2001;285:1071–1075. Lam SK, Ng TB. First simultaneous isolation of a ribosome inactivating protein and an antifungal protein from a mushroom (Lyophyllum shimeji) together with evidence for synergism of their antifungal effects. Arch Biochem Biophys. 2001;393:271–280. Wong JH, Wang HX, Ng TB. Marmorin, a new ribosome inactivating protein with antiproliferative and HIV-1 reverse transcriptase inhibitory activities from the mushroom Hypsizigus marmoreus. Appl Microbiol Biotechnol. 2008;81:669–674. Wang HX, Ng TB. Isolation of pleuturegin, a novel ribosome-inactivating protein from fresh sclerotia of the edible mushroom Pleurotus tuber-regium. Biochem Biophys Res Commun. 2001;288:718–721. Wang H, Ng TB. Isolation and characterization of velutin, a novel low-molecular-weight ribosome-inactivating protein from winter mushroom (Flammulina velutipes) fruiting bodies. Life Sci. 2001;68:2151–2158. Yao QZ, Yu MM, Ooi LS, et al. Isolation and Characterization of a Type 1 Ribosome-Inactivating Protein from Fruiting Bodies of the Edible Mushroom (Volvariella volvacea). J Agric Food Chem. 1998;46:788–792. Savino C, Federici L, Brancaccio A, et al. Crystallization and preliminary X-ray study of saporin, a ribosome-inactivating protein from Saponaria officinalis. Acta Crystallogr D Biol Crystallogr. 1998;54:636–638. Legname G, Gromo G, Lord JM, et al. Expression and activity of pre-dianthin 30 and dianthin 30. Biochem Biophys Res Commun. 1993;192:1230–1237. Rosenblum MG, Kohr WA, Beattie KL, et al. Amino acid sequence analysis, gene construction, cloning, and expression of gelonin, a toxin derived from Gelonium multiflorum. J Interferon Cytokine Res. 1995;15:547–555. Vepachedu R, Park SW, Sharma N, Vivanco JM. Bacterial expression and enzymatic activity analysis of ME1, a ribosomeinactivating protein from Mirabilis expansa. Protein Expr Purif. 2005;40:142–151. Singh V, Sairam MR. Effects of thiolation on the immunoreactivity of the ribosome-inactivating protein gelonin. Biochem J. 1989;263:417–423. Paul A, Madan S, Vasandani VM, et al. Liposome immune lysis assay (LILA) for gelonin. J Immunol Methods. 1992;148:151–158. Ng TB, Hon WK, Lo LH, et al . Effects of alpha-momorcharin, beta-momorcharin and alpha-trichosanthin on lipogenesis and testicular and adrenal steroidogenesis in vitro and plasma-glucose levels in vivo. J Ethnopharmacol. 1986;18:45–53. Ng TB, Liu WK, Tsao SW, Yeung HW. Effect of trichosanthin and momorcharins on isolated rat hepatocytes. J Ethnopharmacol. 1994;43:81–87. Ng TB, Chan WY, Yeung HW. Changes in ovulatory and steroidogenic responses in mice after administration of the ribosome inactivating proteins momorcochin, luffaculin and luffins. Gen Pharmacol. 1994;25:19–21. Ng TB, Chan WY, Sze LY, Yeung HW. Trichosanthin induces atresia of ovarian follicles and inhibits steroidogenesis in gonadotropin-primed immature mice. Gen Pharmacol. 1991;22:847–849. Xiong SD, Yu K, Liu XH, et al. Ribosome-inactivating proteins isolated from dietary bitter melon induce apoptosis and inhibit histone deacetylase-1 selectively in premalignant and malignant prostate cancer cells. Int J Cancer. 2009;125:774–782. Li M, Li X, Li JC. Possible mechanisms of trichosanthin-induced apoptosis of tumor cells. Anat Rec. (Hoboken) 2010;293:986–992. Wang S, Zhang Y, Liu H, et al. Molecular cloning and functional analysis of a recombinant ribosome-inactivating protein (alpha-momorcharin) from Momordica charantia. Appl Microbiol Biotechnol. 2012;96,4:939–950. He D, Zheng Y, Tam S. The anti-herpetic activity of trichosanthin via the nuclear factor-kappaB and p53 pathways. Life Sci. 2012;90:673–681. Tong WM, Sha O, Ng TB, et al. Different in vitro toxicity of ribosome-inactivating proteins (RIPs) on sensory neurons and Schwann cells. Neurosci Lett. 2012;524:89–94. Shahidi Noghabi S, Van Damme E, Smagghe G. Bioassays for insecticidal activity of iris ribosome-inactivating proteins expressed in tobacco plants. Commun Agric Appl Biol Sci. 2006;71:285–289. Bertholdo-Vargas LR, Martins JN, Bordin D, et al. Type 1 ribosome-inactivating proteins - entomotoxic, oxidative and genotoxic action on Anticarsia gemmatalis (Hubner) and Spodoptera frugiperda (J.E. Smith) (Lepidoptera: Noctuidae). J Insect Physiol. 2009;55:51–58. Mansouri S, Nourollahzadeh E, Hudak KA. Pokeweed antiviral protein depurinates the sarcin/ricin loop of the rRNA prior to binding of aminoacyl-tRNA to the ribosomal A-site. RNA. 2006;12:1683–1692.

5

Non-toxic Type 2 Ribosome-inactivating Proteins Pilar Jiménez1, Manuel José Gayoso2, and Tomás Girbés1 1 2

Facultad de Medicina and Centro de Investigación en Nutrición, Universidad de Valladolid, Spain Departamento de Farmacología, Biología Celular e Histología, Universidad de Valladolid, Spain

Introduction

The literature on plant ribosome-inactivating proteins (RIPs) has been largely reviewed, including within the chapters of this present book. Tribute must be paid to those scientists who expanded significantly the field of these proteins, most of them contributors to the present volume, and especially to Prof. Stirpe and Dr. Lappi. In general, it may be assumed that RIPs are present in bacteria, fungi, and algae,1–4 and a preliminary report indicates that they could also be present in mammals.5 Plant RIPs have traditionally been classified according to their molecular structure into three categories: types 1, 2, and 3 RIPs.3 Type 1 RIPs, such as PAP (Phytolacca antiviral protein), saporin, dianthin, gelonin, trichosanthin, and so on, consist of a single polypeptide chain with N-glycosidase activity; whereas type 2 RIPs, such as ricin, abrin, volkensin, and so on among the toxic proteins, and ebulin, nigrin, SNA I, IRAb, and so on among the non-toxic, contain two different polypeptide chains linked by a disulfide bridge, an A chain with N-glycosidase activity, and a B chain with sugar-binding activity able to bind to cell surface polysaccharides. The type 3 RIPs are composed of a polypeptide chain containing a type I RIP in the N-terminal end and a C-terminal domain whose function is unknown at present. The N-glycosidase activity on both viral RNA and genomic DNA could have a special biological role in the control of viral infection, which could explain the antiviral role found in many RIPs both in vitro and in vivo.6–9 In addition, it has been proposed that the reported N-glycosidase activity on the genomic DNA could serve in mechanisms of plant senescence.10 In the present chapter we will focus on type 2 RIPs that have much less in vivo toxicity on mice than ricin and therefore are known as non-toxic type 2 RIPs. Nonetheless, this does not exclude the possibility that some of these proteins could act efficiently on other animal models, for instance, insects. The goal of the present chapter is to review the non-toxic type 2 RIPs. The most studied plant family has been the Sambucaceae, in which a few dozen RIPs and structure-related lectins have been found.3, 4 Due to the complex mixture of both type 1 and 2 RIPs and type 2 RIP-related lectins found in Sambucus, this family seems a good model to study the distribution, gene expression, seasonal variations, and toxicity of these proteins. In addition, some other type 2 RIPs that are

Ribosome-inactivating Proteins: Ricin and Related Proteins, First Edition. Edited by Fiorenzo Stirpe and Douglas A. Lappi. © 2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.

67

68

RIBOSOME-INACTIVATING PROTEINS

non-toxic as compared with ricin and are found in other families will be considered. All proteins herein presented display glycosidase activity and inhibition of translation. Therefore we only comment on the relevant features of each. Family Sambuceae

Type 2 and 1 RIPs and lectins have been found and characterized in Sambucus (Table 5.1). All these proteins showed affinity for D-galactose (D-gal) and N-Ac-galactosamine (N-Ac-gal) and most of them were isolated by a general chromatography procedure consisting in a first step of affinity chromatography on acid-treated-Sepharose 6B, followed by a second step of gel filtration on Superdex 75. All Sambucus proteins have been analyzed by SDS-polyacrylamide gel electrophoresis and the subunits and Mr are listed in Table  5.1. In all cases the A chain displays the enzymatic activity and the B chain the lectin activity, both held together by disulfide bonds. Perhaps the most remarkable characteristic of the type 2 RIPs from Sambucus is that, unlike ricin and other highly toxic type 2 RIPs, they have a very low toxicity to animal cells and even intact animals (103–104-fold less toxic to mice) as compared with ricin. An outline of the processing of elder proteins from a theoretical ancestral gene containing the A and B chains of type 2 RIPs is shown in Figure 5.1. Truncations and mutations of that gene led to gaps in the A chain of precursors of both lectins, yielding polypeptides containing the B chain and a fragment of the A chain, which after processing yields the B-type native lectin.11, 12

Sambucus Ebulus (Dwarf Elder) Ebulin l

Ebulin l was isolated from the leaves of dwarf elder (Sambucus ebulus L.) harvested in the spring.13 It agglutinates human O + red blood cells at 51 μg/ml. Agglutination increased five-fold at 4 °C with respect to that observed at 22 °C.12 Ebulin l strongly inhibits protein synthesis in rabbit reticulocyte lysates and other mammalian systems (IC50 values of 0.1–0.28 nM) but is inactive on plant and bacterial cell-free systems. The cytotoxicity of ebulin l for HeLa cells (IC50 64.3 nM) is 60,000 times lower than the IC50 for ricin. The i.p. LD50 values are 2000 and 2.6 μg/kg for ebulin l and ricin respectively. Studies by fluid-phase uptake of ebulin l indicate that some differences in the sugar-binding ability of the B chain prevent ebulin l from following the internalization pathway of ricin.14 A cDNA obtained from total RNA coding for ebulin l gene, isolated from early-summer young leaves of dwarf elder, contains a 1692-bp open reading frame encoding a polypeptide of 564 amino acids, which is further processed to give the native protein.15 Ebulin l A and B chains share 34 and 48% amino acid identity with ricin;15 the ebulin l B chain shares nearly 80% identity with two ebulin l-related lectins named SELld (B–B structure) and SELlm (B structure).12 A 2.8 Å X-ray diffraction of ebulin crystals reveals that it binds lactose and D-gal in subdomain 1α in a similar manner to that of ricin.15 However, ricin binds D-gal-polysaccharides and D-gal in the subdomain 2γ and ebulin l binds only D-gal and not lactose. The reason for such differences seems to be related to the orientation of D-gal bound to subdomain 2γ of ebulin, which leads to a weaker binding of ebulin l to terminal galactosides on the plasma cell membrane.15 The presence of ebulin l is massive in shoots but only vestigial in senescent leaves.16 Conversely, SELld – an ebulin l B chain-related lectin with a B–B structure – is vestigial in shoots and increases dramatically in

APA cinnamomin porrectin foetidissimin foetidissimin II EHL IRAb IRAr MCA PMRIPm PMRIPt RCA ebulin l ebulin f ebulin r1 ebulin r2 SEAI nigrin b basic nigrin b SNAI SNAI´ SNRLP1 SNRLP2 nigrin f nigrin s sieboldin b SSA VAA VCA

seeds seeds seeds roots roots tubers bulbs

seeds leaves

seeds leaves fruits rhizome rhizome rhizome bark bark bark bark bark bark fruits seeds bark bark leaves leaves

Abrus precatorius Cinnamomum camphora Cinnamomum porrectum Cucurbita foetidissima Cucurbita foetidissima Eranthis hyemalis Iris holandica

Momordica charantia Poligonatum multiflorum

Ricinus communis Sambucus ebulus

138 000 61 000 64 500 63 000 62 000 62 000 65 000 68 000 120 000 65 000 240 000 130 000 56 000 56 000 56 000 56 000 140 000 58 000 64 000 136 000 67 000 68 000 62 000 58 000 57 300 60 000 115 432 120 000 66 000

Whole – 30 500 30 500 30 000 29 000 30 000 30 000 30 000 28 000 30 000 30 000 31 000 26 000 26 000 26 000 26 000 30 000 26 000 32 000 33 000 32 000 34 000 30 000 26 300 26 300 27 000 28 700 – 31 000

A chain – 30 500 33 500 33 000 33 000b 28 000 38 000 35 000 30 000 35 000 35 000 34 000 30 000 30 000 30 000 30 000 35 000 32 000 32 000 35 000 35 000 34 000 32 000 31 600 31 000 33 000 29 000 – 35 000

B chain

Molecular mass (SDS-PAGE)

D-gal/NAGal D-gal D-gal n.d. n.d. n.d. D-gal/GalNAc/Man D-gal/GalNAc/Man D-gal D-gal/galNAc D-gal/galNAc D-gal/Man D-gal D-gal D-gal D-gal Neu5Ac-gal/GalNAc D-gal/GalNAc n.a. Neu5Ac-gal/GalNAc Neu5Ac-gal/GalNAc n.a. n.a. D-gal D-gal Neu5Ac-gal/GalNAc Neu5Ac-gal/GalNAc D-gal D-gal/GalNAc

Sugar spec. – 124 22.4 3.5 10 – – – – – – – 3.2 1.6 1.34 0.56 6 254 90 100 0.2 100b – 1.3c – 2.6 – – –

yielda 0.05 – 1 26 251 100 0.1 –

IC50(nM)

a

Type 2 RIPs have been also detected in the bark of S. mexicana, S. racemosa, S. peruviana, and leaves of S. nigra (unpublished results). yield per 100 g of starting material. b mixture of SNRLP1 and SNLRP2. IC50: concentration required to inhibit translation to 50% of control, unless indicated translation was assayed with reduced RIPs in rabbit reticulocyte lysates. D-gal: D-galactose; Man: manose; GalNAc: N-acetyl-galactosamine; Ney5Ac: N-acetyl-neuraminic acid (sialic acid). The corresponding references are in Reference 3.

Viscum album Viscum album coloratum

Sambucus sieboldiana

Sambucus nigra

Protein

Tissue

Non-toxic type 2 RIPs.

Species

Table 5.1

Figure 5.1 Transformation of precursors into native proteins both type 2 RIPs and related lectins from Sambucus. L: leader peptide; A: A chain; B: B chain; C: connecting peptide.

NON-TOXIC TYPE 2 RIBOSOME-INACTIVATING PROTEINS

71

(a)

(b)

Figure 5.2 Effects of the intraperitoneal administration of 5 mg/kg body weight of ebulin f to Swiss mice. (a) A detail of the small intestine mucosa; (b) a detail of the crypts of the large intestine. Staining was carried out with hematoxilin-eosin. Arrows indicate cells with morphological stages of apoptosis. Scale bar: 20 μm. (This figure also appears in the color plate section.)

senescent leaves. In mature leaves there is nearly the same amount of both proteins. This behavior of ebulin l contrasts with published data, which indicate that type 1 RIPs accumulate in senescence.10 This suggests that type 1 and type 2 RIPs most probably play different biological roles in the plant. Ebulin f

Ebulin f was isolated from the green fruits of S. ebulus using the same protocol as for ebulin l.16 Ebulin f inhibits protein synthesis in a similar manner as ebulin l. Nonetheless, while the sugar-binding specificity is the same as that of ebulin l, ebulin f triggers full agglutination of human red blood cells at a concentration ten-fold lower than that of ebulin l. Ebulin f is somewhat more toxic to HeLa cells (IC50 17 nM) than ebulin l (IC50 64.3 nM).17 These differences support the notion that both ebulins are different proteins. Ebulin f accumulates in the green fruit and disappears completely with maturation. In green fruits, ebulin f can be polymerized with other ebulin f molecules, and even with lectins, to form high molecular weight aggregates that coexist with free forms of these proteins.16 Intraperitoneal administration of 5 mg/kg body weight of ebulin f to Swiss mice promoted a serious specific intestinal derangement, which particularly affected the small intestine with complete destruction of the Lieberkühn crypts (Figure 5.2) and elimination of the villi. The large intestine

72

RIBOSOME-INACTIVATING PROTEINS

was affected to a lesser extent. The LD50 was found close to 2.8 mg/kg body weight. At sub-lethal doses ebulin f triggers similar, but transient and less prominent, effects. Ebulins r1 and r2

Ebulins r1 and r2 were isolated from dwarf elder rhizomes by the general affinity procedure.18 They coexist with two forms of a monomeric lectin named SEAII-1 and SEAII-2. These proteins were further resolved by anion exchange chromatography on Mono-Q into two isoforms of rhizome ebulin, named r1 and ebulin r2, and two monomeric lectins.18 Ebulins r1 and r2 in their native forms inhibit protein synthesis in rabbit reticulocyte lysates with IC50 values of 59.6 and 0.6 nM, respectively.19 In their reduced forms they inhibit translation in the same system with IC50 values of 0.34 and 1.14. Both RIPs also display glycosylase activity on herring sperm DNA that is higher in their reduced form than in their native form.19 Both rhizome ebulins and ebulin l are different proteins as revealed by immunologic characterization.16 SEA

Sambucus ebulus agglutinin (SEAI) was first isolated from the bark of dwarf elder rhizomes by affinity chromatography on fetuin-agarose, ion-exchange on Mono-Q, and gel filtration on Superose 12HR, with a yield of 3.1 mg/g (fresh weight).20 It has also been isolated like ebulin l.21 The sugar-binding specificity of SEA is D-gal/N-Ac-gal. Recent research carried out in our laboratory indicates that it is in fact a type 2 RIP.21 The molecular cloning of the SEAI gene was carried out from genomic sequences encoding SEAI, amplified by PCR, which, upon sequencing, revealed that SEA is encoded by a genomic sequence of 1617 bp accounting for a precursor of 539 amino acids, which is further processed to give the native protein. SEA shares approximately 85% of amino acid identity with the two other type 2 RIPs from Sambucus, namely SNAI and SSA. SEAI inhibits protein synthesis in rabbit reticulocyte lysates with an IC50 of 10-9 M and promotes loss of viability of COLO 320 cells with an IC50 close to 6 x 10-7 M by the induction of apoptosis, as proved by the typical DNA laddering.21 SEAI interacts with the mucus of the intestinal goblet cells as well as with the endothelia of villus stroma, thus indicating that it has affinity for mucin.21

Sambucus Nigra (Elderberry; Black Elder) Nigrin b

Nigrin b is also a type 2 RIP that was isolated from the bark of S. nigra.22 Later, Peumans and coworkers used the name SNA V.23 Nigrin b was prepared by general affinity chromatography. The yield of nigrin b was 86 mg/kg of wet elderberry bark. A more recent and complex, but more efficient, procedure yielded 445 mg/kg of wet bark.24 Nigrin b promotes full agglutination of intact human red blood cells at 12.5 μg/ml,22 and displays specificity for D-gal/N-Ac-gal.23 The amino acid sequence was deduced from a cDNA encoding SNAV (nigrin b).23 Nigrin b displays 36% amino acid sequence identity with ricin and 33% amino acid sequence identity with abrin. The amino acid sequence of the A chain shows sequence homologies of 88% with the ebulin l A chain and 76% with the ebulin l B chain. Nigrin b also has similarity with the type 2 RIP SNAI and the lectins SNAlm and SNAld of 53%, 86%, and 72%, respectively.24 Reduced nigrin b inhibits protein synthesis in mammalian cell-free systems with nearly the same efficiency as ebulin l, and like ebulin l is much less active than the native protein.19 Its IC50 on HeLa cells is 5.35 x 10 -7 M.25 Nigrin b and the highly toxic type 2 RIP volkensin bind to HeLa cells with the same affinity (approx. 10-10 M) and have a similar number of binding sites (2 x 10-5/cell) which

73

NON-TOXIC TYPE 2 RIBOSOME-INACTIVATING PROTEINS

Ebulin/nigrin

Extracellular space

Ricin AB AB Plasma membrane

Binding Receptors

Cytosol

“Coated pits” Clatrin Internalization

Secretion Late endosomes

Lisosomes

Caveosome

Ribosomes

Early endosomes

Blocked at T < 18 °C

RER Golgi-trans

Brefeldin A

Golgi-cis Figure 5.3 Outline scheme of the intracellular pathways followed by ricin, nigrin, and ebulin. (This figure also appears in the color plate section.)

is two-log lower than ricin.26 But in contrast to volkensin and ricin, which stack in the perinuclear localization, nigrin b accumulates in cytoplasmic dots and the Golgi compartments.26 The most significant results compared with ricin are summarized in Figure 5.3. Initially, ricin and nigrin b follow the same pathway to reach the endosomes. From this point onward, a fraction of ricin follows the trans-Golgi network to the endoplasmic reticulum, where it is translocated to the cytosol.27 Passage through the endosomal pathway promotes the degradation of 79% of ricin and 94% of nigrin b and the remainder is expelled from the cell.28 Nigrin translocation to cytosol seems to be dependent on the accumulation of nigrin b in the endosomes.28, 29 The classical paradigm for the cellular action of type 2 RIPs on sensitive cells was that the protein synthesis inhibition was enough to trigger cell destruction. However, since the concentration of nigrin b required to inhibit translation is higher than that required to arrest growth,30 and following a recent study on ricin using microarray, qRT-PCR, and Western blot, the notion that the effects of ricin and perhaps nigrin b could be best explained by other processes, for instance the induction of the so-called unfolded protein response (UPR), is gaining support.31 The i.p. LD50 for Swiss mice was found to be 12 mg/kg body weight,29 while the i.v. was approximately 7 mg/kg (unpublished results). Oral administration of larger amounts of nigrin b did not trigger toxicity at all.24 The histological analysis revealed that i.v. administration of lethal doses of nigrin b led to a serious derangement of the intestine (Swiss and APC/Min + mice) which promoted

74

RIBOSOME-INACTIVATING PROTEINS

Figure 5.4 Electron microscopy of the bottom of a small intestine crypt in a nigrin b treated mouse. The preparation was performed with samples of tissue from Swiss mice after 16 h of intravenous administration of 16 mg/kg of nigrin b. The remains of apopototic cells (big arrow) coexist with less affected Paneth cells (small arrow). Scale bar: 4 μm.

profuse bleeding and further death with no apparent gross damage of other organs. The effects were seen 5 h after administration first in the small intestine, where nigrin b triggers apoptosis of the transient amplifying cells of the mid third of the Lieberkühn crypts. After this, the villi are also destroyed and the whole mucosa disappears; finally the crypts of the large intestine are also destroyed.24, 32 At sub-lethal doses (5 mg/kg body weight) the effects were similar but completely reversible 7–9 days after the administration of the RIP.32 Electron microscopy analysis of the effects of nigrin b 6 h after administration of 16 mg/kg nigrin b i.v. confirms the apoptopic effects of nigrin b in mice. Figure 5.4 shows different stages of apoptosis: condensation of the cytoplasm and nucleus, cytoplasmic blebs with clumps of chromatin, condensed chromatin in sickle-like disposition, and elimination of remains of nuclear chromatin and shrunken cytoplasm. Nigrin f

Nigrin f was isolated from fruits by the general affinity procedure.33, 34 The yield was 13 mg/kg from unripe fruits and 1.2 mg/kg from ripe fruits. Nigrin f coexists with a monomeric lectin, named SNAIV, which accounts for over 98% of the protein and also binds to acid-treated Sepharose 6B. Fruit ripening led to a seven-fold reduction of D-gal-binding proteins and to a ten-fold reduction of nigrin f. Nigrin f was more toxic than nigrin b for HeLa cells with an IC50 of 2.9 nM.17 A gene that codes for LECSNAVf (nigrin f) has been cloned and analysis of its sequence reveals that it contains a 1689-bp open reading frame encoding a polypeptide of 563 amino acids, which is processed to finally yield a two-chain protein of 538 amino acids which displays 88% sequence identity with nigrin b.35

NON-TOXIC TYPE 2 RIBOSOME-INACTIVATING PROTEINS

75

Nigrin s

Seeds of S. nigra also contain a type 2 RIP, as revealed by Western blot analysis with anti-nigrin b rabbit polyclonal antibodies.36 It was named nigrin s and showed the characteristic effects of type 2 RIPs of Sambucus. It coexists with, but is much less abundant than, the monomeric lectin SNAIII that was previously isolated.37 Basic Nigrin b

An uncommon two-chain RIP of basic nature has been obtained from bark of S. nigra.38 The procedure involved the use of a crude protein extract depleted of D-galactose-binding proteins by acid-treated Sepharose 6B affinity chromatography to remove SNAI, nigrin b, and SNAII. This extract was then subjected to several chromatography steps on SP-Sepharose Fast Flow, CM-Sepharose Fast Flow, Superdex 75 HiLoad, and finally Mono S. The yield of basic nigrin b was 0.9 g/kg of elderberry bark. Unfortunately, it was impossible to separate both subunits. The analysis of some tryptic peptides indicates that it is clearly an A–B* type 2 RIP, the B* chain being a truncated lectin lacking functional sugar-binding domains. Consequently basic nigrin b does not agglutinate human red blood cells, which resembles the case of SNLRP1 and SNLRP2.39 Despite its high activity on rabbit reticulocyte lysates (IC50 18 pg/ml of lysate) it is inactive on plant cell-free systems and on HeLa cells even at 10-6 M. Together with the classical N-glycosidase activity, it also displays topological activity on DNA strands. On the other hand, intraperitoneal administration of up to 40 mg/kg body weight is not toxic to mice. SNAI

The first lectin isolated from Sambucaceae was SNAI (Sambucus nigra agglutinin I). It was obtained from elderberry (S. nigra L.) bark by affinity chromatography on fetuin-agarose.40 It is a tetrameric protein which can form octamers by non-covalent association between the tetramers.23 The SNAI B chain has sugar-binding specific activity for GalNAc and NeuAc(α-2,6).40 The specificity of this protein for sialic acid allowed its profuse use for labeling the terminal Neu5Ac(α2,6)Gal/GalNAc of glycan chains. The gene encoding SNAI (LECSNAI) has been cloned and reveals that the lectin has nearly 54% sequence identity with both nigrin b and nigrin f. The molecular modeling of SNAI suggests that the spatial structures of its A and B chains are similar to those of ricin. SNAI accumulation in elder bark is lowest in autumn and highest in summer.24 The histological analysis indicates that SNAI accumulates in the protein bodies of the bark phloem-parenchyma cells, perhaps as nitrogen storage.41 SNAI′ A minor protein, that was named SNAI′ and whose amino acid sequence is very similar to SNAI, was isolated from elderberry bark.42 The procedure involved chromatography on fetuin-Sepharose 4B and Sephacryl 100. These operations were repeated and finally pure SNAI′ was obtained. It has nearly 72% identity with SNAI and displays the same inhibitory activity on rabbit reticulocyte lysates (IC50 of 150 ng/ml). SNAI′ has the same sugar specificity as SNAI. Molecular cloning of the gene coding for SNAI′, named LECSNAI′, revealed that the main difference between SNAI and SNAIʹ is that the latter has a tetrameric instead of octameric structure because it has one cysteine fewer in its B chain.42 SNLRP1 and SNLRP2

SNLRP (Sambucus nigra lectin-related protein) 1 and 2 are basic proteins that were isolated from a lectin-depleted extract from S. nigra bark using a combination of hydrophobic interaction chromatography, ion exchange chromatography, and gel filtration.39 Both proteins inhibit protein synthesis in rabbit reticulocyte lysates (IC50 0.5 μg/ml for a mixture of both proteins) less than basic nigrin b (IC50 18 pg/ml). Like basic nigrin b, SNLRPs 1 and 2 are devoid of sugar-binding activity and hence

76

RIBOSOME-INACTIVATING PROTEINS

lack the ability to agglutinate red blood cells. SNLRP 1 and 2 have 91% identity and the molecular modeling indicates that they display striking differences in amino acids of the sugar-binding domains that makes them unable to bind sugars. Tryptic peptide sequences obtained from basic nigrin b indicate that this protein has a high sequence homology (nearly 90%) with SNLRPs.38 Besides the differences in inhibitory activity on protein synthesis, basic nigrin b is a major protein in elder bark while SNLRPs are minor ones.38

Sambucus Sieboldiana (Japanese Elderberry) Sieboldin b

Sieboldin b was isolated from the bark of Japanese elderberry (S. sieboldiana) by affinity chromatography on acid-treated Sepharose 6B, gel filtration on Superdex 75, and Mono-Q chromatography.43 Sieboldin b inhibits protein synthesis in rabbit reticulocyte lysates with an IC50 of 0.015 nM, but is inactive in a cell-free system obtained from Triticum aestivum. Sieboldin b agglutinates rabbit red blood cells at 3.1 μg/ml. Sieboldin b shows greater affinity for N-Ac-gal than for D-gal. Its amino acid sequence has 89% identity with that of nigrin b. Its IC50 in HeLa cells is 12 nM (slightly lower than nigrin) and it is not toxic to mice by intraperitoneal administration up to 1.6 mg/kg body weight. The gene of sieboldin b contains an open reading frame encoding a polypeptide of 563 amino acids. Sieboldin b conserves the cysteine, forming the disulfide bridge between the A and B chains and the amino acids of active sites. SSA

SSA (S. sieboldiana agglutinin) was obtained from the bark of Japanese elderberry by affinity chromatography on fetuin-agarose.44 It is specific for NeuAc(α-2,6)/GalNAc and for this reason it was thought that the protein might be the equivalent of SNAI in S. sieboldiana. The free, stabilized A chain has an IC50 of 540 ng/ml on rabbit reticulocyte lysates,45 which is lower than other RIPs from Sambucus. The IC50 values for the inhibition of the in vivo protein synthesis carried out by HeLa cells were 55 μg/ml and 710 ng/ml for sieboldin b and SSA, respectively. The gene coding for SSA contains a 1710-bp open reading frame encoding a 570 amino acid polypeptide that, after processing, gives the native protein. It has a cysteine to form the disulfide bridge between the A and B chains and another cysteine that permits linking between two dimers to form a covalentlystabilized tetramer. Type 2 RIPs from Other Families Cinnamomin (Cinnamomum Camphora; Family Lauraceae)

Cinnamomin was isolated from mature seeds of Cinnamomum camphora by affinity chromatography on acid treated-Sepharose 4B, DEAE-cellulose, and Sephadex G100.46 Cinnamomin A chain exhibits similar RNA N-glycosidase activity in inhibiting in vitro protein synthesis compared with ricin, whereas the cytotoxicity to BA/F3beta cells of intact cinnamomin is markedly lower than intact ricin due to its B chain. ELISA and Scatchard analysis of the binding of the ricin and cinnamomin B chains indicates that the differences in toxicity come from the different affinities rather than the number of binding sites.47

NON-TOXIC TYPE 2 RIBOSOME-INACTIVATING PROTEINS

77

Porrectin (Cinnamomum Porrectum; Family Lauraceae)

Porrectin is a type 2 RIP present in seeds of the camphor tree (Cinnamomum porrectum) and was purified by affinity chromatography on acid-treated Sepharose 4B at 4–8 °C.48 The yield was 224 mg/ kg of seeds. Porrectin is present in three isoforms with the same Mr but different pI. Porrectin inhibited protein synthesis in a rabbit reticulocyte lysate (IC50 1.1 × 10-7 M), an activity that could be greatly enhanced by reduction of porrectin by 2-mercaptoethanol (IC50 1 × 10-9 M). The protein has rRNA N-glycosidase activity on rat liver ribosomes, like the RIPs. The in vivo toxicity of porrectin has not been studied.

EHL (Eranthis Hyemalis; Family Ranunculaceae)

The two-chain lectin EHL isolated from Eranthis hyemalis49 was shown to be a type 2 RIP.50 EHL displays protein synthesis inhibitory activity in rabbit reticulocyte lysates (full inhibition at 100 nM) and wheat germ cell-free systems (full inhibition at 500 nM); the reduction and alkylation of EHL reduced its inhibitory activity. EHL also displays antiviral activity against the alfalfa mosaic virus and larvicidal activity against the southern corn rootworm (Diabrotica undecimpunctata howardii). The in vivo toxicity on rodents has not been assessed and therefore EHL cannot be compared with the highly toxic RIP ricin and the non-toxic nigrin b.

Foetidissimin (Cucurbita Foetidissima; Family Cucurbitaceae)

Foetidissimin II is a type 2 RIP obtained from the roots of Cucurbita foetidissima that was purified by chromatography on Sephadex G25, Q-Sepharose, and finally Superdex 75 with a yield of 10 mg/kg dry root powder.51 The IC50 value of foetidissimin II on rabbit reticulocyte lysates was 251 × 10−9 M and it exhibited RNA-N-glycosidase activity on the rRNA. It is also cytotoxic on adenocarcinoma and erythroleukemia cells with IC50 values close to 70 nM. The in vivo toxicity was not assayed. MCA (Momordica Charantia; Family Cucurbitaceae)

Bitter gourd seeds contain a haemagglutinin which inhibits protein synthesis.52 It was found to be a galactose-specific lectin.53 The structure is of the type (A–B)2 and shares an amino acid sequence identity of 39 and 27% with the A and B chains of RCA respectively, and 47 and 35% with the A and B chains of ebulin l respectively.54 An i.p. administration of 1 mg/kg body weight has no effect on rats.52

RCA (Ricinus Communis; Family Euphorbiaceae)

Ricinus communis agglutinin was isolated together with ricin from castor bean (Ricinus communis) seeds,55, 56 and the amino acid sequence was compared with that of ricin, showing 93 and 84% amino acid sequence homology in the A and B chains respectively.57 RCA, like APA, was not toxic to mice up to 33.3 μg/kg body weight (i.p.) and was shown to be a true four-chain type 2 RIP.58

78

RIBOSOME-INACTIVATING PROTEINS

APA (Abrus Precatorious; Family Fabaceae)

Two agglutinins, named APAI and APAII, were purified from the seeds of Abrus precatorius by lactamyl-Sepharose affinity chromatography followed by gel filtration and DEAE-Sephacel chromatography.59 APAII was shown to fulfill all requirements to be considered a type 2 RIP but without the in vivo toxic effects to mice displayed by ricin, at least up to 33.3 μg/kg body weight.

VAA (Viscum Album; Family Viscaceae)

Viscum album agglutinin60 is an immunomodulatory plant lectin with alleged benefits in tumor therapy.61 The lectin induces aggregation of neutrophils, thrombocytes, and mononuclear cells, and promotes H2O2 release from neutrophils. It is believed that these activities are mediated by the sugar-binding B chain. It has an apparent Mr of approximately 120,000 and a tetrameric structure like RCA and APA. It was shown to be a true type 2 RIP, more toxic to mice than APA and RCA.55

PMRIPm and PMRIPt (Poligonatum Multiflorum; Family Liliaceae)

Polygonatum multiflorum L. (Solomon’s seal) leaves contain two type 2 RIPs, named PMRIPm and PMRIPt, which are specific for D-galactose/N-Ac-galactosamine.62 Both RIPs were purified by several steps of chromatography, first S Fast Flow, followed by mannose-Sepharose 4B, galactoseSepharose 4B, and finally phenyl-Sepharose. These RIPs exhibit differences in their agglutination activity, so PMRIPm is specific for Gal/N-Ac-gal and PMRIPt is specific only for N-Ac-gal. Both RIPs exhibit a very low cytotoxicity towards human and animal cells. Analysis of the genomic clones encoding both RIPs and molecular modeling revealed that they have a high degree of structural similarity to other type 2 RIPs. To our knowledge, the toxicity in vivo has not yet been studied.

IRAb and IRAr (Iris Holandica; Family Liliaceae)

Dutch iris (Iris hollandica var. Professor Blaauw) bulbs contain two closely related lectins called Iris agglutinin b (IRAb) and Iris agglutinin r (IRAr) which have been isolated and cloned.63 Both lectins share a high amino acid sequence similarity with other type 2 RIPs. They coexist with a type 1 RIP named IRIP with which IRAb and IRAr display 57 to 59% sequence identity. Sugar-binding studies and docking experiments revealed that IRAb and IRAr bind Gal/N-Ac-gal and mannose. The IRA RIPs display low toxicity to several cell lines as compared with viscumin.60 No assay of in vivo toxicity has been reported.

Uses of Nontoxic Type 2 RIPs Construction of Immunotoxins and Conjugates

Nigrin b and ebulin l have been conjugated with other proteins that increase the cytotoxicity of the RIPs.25 Among such proteins are human transferrin, two antihuman CD105 monoclonal antibodies, and dimeric mucin-binding lectins from S. ebulus. While conjugation of nigrin b and ebulin l to Tf

NON-TOXIC TYPE 2 RIBOSOME-INACTIVATING PROTEINS

79

did not affect greatly their antiribosomal inhibitory activity, the conjugates promoted inhibition of cell protein synthesis in HeLa cells (IC50 of 100 ng/ml);64 in contrast, both free forms of ebulin l and nigrin b showed much higher IC50 values, close to 4000 and 1800 ng/ml, respectively. CD105 (endoglin) is a TGF-β co-receptor highly expressed in proliferating endothelial cells of the new vasculature and is up-regulated by hypoxia.65 Therefore, it was considered a biomarker of proliferationdependent pathologies.66 Nigrin b and ebulin l have been used to construct immunotoxins containing antihuman endoglin (44G4). Both 44G4-ebulin l and 44G4-nigrin b immunotoxins displayed cytotoxicity in the picomolar range, while the RIPs alone kill target cells in the micromolar range.67 Immunofluorescence analysis indicated that 44G4-nigrin b accumulated in the perinuclear region.68 Conjugation of nigrin b with the dimeric lectins SELld and SELfd from S. ebulus, SNAI, or the isolated SNAI B chain increased the cytotoxicity of nigrin b to COLO 320 and HeLa cells by two to three orders of magnitude.69

Effects on Insects

It has been shown that administration of SNAI in the diet promoted insecticidal activity to Acyrthosiphon pisum, which was confirmed by experiments of feeding the tobacco aphid, Myzus nicotianae, and the beet armyworm, Spodoptera exigua, with transgenic tobacco expressing the gene coding for SNAI´.69 The toxic effects of SNAI seem to be exerted on the insect gut, since SNAI applied to cultures of midgut CF-203 insect cells triggers the apoptosis mediated by caspase-3.70, 71 The toxic effects were dependent on the sugar-binding activity of the B chain, since mutations in the sugar-binding sites strongly reduced the insecticidal activity.70 Further support for this idea comes from experiments in which similar apoptopic effects were seen upon administration of the monomeric lectin SNAII to CF-203 cells.72 Concluding Remarks

Research on non-toxic type 2 RIPs has received special attention due to their practical uses for experimental therapy and crop protection. The main feature of these proteins is their low toxicity as compared with ricin and related highly toxic type 2 RIPs. This characteristic makes them good candidates for the construction of more specific and less toxic immunotoxins and conjugates against specific antigens of tumor cells. RIPs from Sambucaceae do not inhibit protein synthesis, either in bacteria or in plants, which facilitates their cloning and expression and the construction of transgenic plants resistant to fungi and viruses. Their seasonal fluctuations and biological activities suggest a potential role for storage of nitrogen and an antipathogen role, which merit more research effort. Acknowledgment

Dedicated to Prof. Fiorenzo Stirpe, pioneer, master, and friend. References 1. Barbieri L, Battelli MG, Stirpe F. Ribosome-inactivating proteins from plants. Biochim Biophys Acta. 1993;1154:237–282. 2. Stirpe F. Ribosome-inactivating proteins. Toxicon. 2004;44:371–383.

80 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

24. 25. 26. 27. 28. 29. 30.

RIBOSOME-INACTIVATING PROTEINS

Girbés T, Ferreras JM, Arias FJ, Stirpe F. Description, distribution, activity and phylogenetic relationship of ribosome-inactivating proteins in plants, fungi and bacteria. Mini Reviews in Medicinal Chemistry. 2004;4:461–476. Ng TB, Wong JH, Wang H. Recent progress in research on ribosome inactivating proteins. Current Prot Pept Sci. 2010;11:37–53. Barbieri L, Valbonesi P, Bondioli M, et al. Adenine glycosylase activity in mammalian tissues: an equivalent of ribosomeinactivating proteins. FEBS Lett. 2001;505:196–197. Taylor S, Massiah A, Lomonossoff G, et al. Correlation between the activities of five ribosome-inactivating proteins in depurination of tobacco ribosomes and inhibition of tobacco mosaic virus infection. Plant J. 1994;5:827–835. Girbés T, de Torre C, Iglesias R, et al. RIP for viruses. Nature. 1996;379:777–778. Zoubenko O, Hudak K, Tumer NE. An-tpapmipivansa-ip. A. non-toxic pokeweed antiviral protein mutant inhibits pathogen infection via a novel salicylic acid-independent pathway. Plant Mol Biol. 2000;44:219–229. Iglesias R, Pérez Y, Citores L, et al. Elicitor-dependent expression of the ribosome-inactivating protein beetin is developmentally regulated by transcription and translation. J Exper Bot. 2008;59:1215–1223. Stirpe F, Barbieri L, Gorini P, et al. Activities associated with the presence of ribosome-inactivating proteins increase in senescent and stressed leaves. FEBS Lett. 1996;382:309–312. Ferreras JM, Citores L, Iglesias R, Jiménez P. Sambucus Ribosome-Inactivating Proteins and Lectins. In: Lord JM, Hartley MR, (eds.) Toxic Plant Proteins Plant Cell Monographs Vol. 18. Berlin: Springer, 2010:107–131. Citores L, Rojo MA, Jiménez P, et al. Transient occurrence of an ebulin-related D. galactose-lectin in shoots of Sambucus ebulus L. Phytochemistry. 2008;69:857–864. Girbés T, Citores L, Iglesias R, et al. Ebulin 1, a nontoxic novel type 2 ribosome-inactivating protein from Sambucus ebulus L. J Biol Chem. 1993;268:18195–18199. Svinth M, Steighardt J, Hernandez R, et al. Differences in cytotoxicity of native and engineered RIPs can be used to assess their ability to reach the cytoplasm. Biochem Biophys Res Commun. 1998;249:637–642. Pascal JM, Day PJ, Monzingo AF, et al. 2.8-A crystal structure of a nontoxic type-II ribosome-inactivating protein, ebulin l. Proteins. 2001;43:319–326. Rojo MA, Citores L, Arias FJ, et al. cDNA molecular cloning seasonal accumulation of an ebulin l-related dimeric lectin of dwarf, elder (Sambucus ebulus L) leaves. Intl J Biochem Cell Biol. 2003;35:1061–1065. Citores L, de Benito FM, Iglesias R, et al. Presence of polymerized free forms of the non-toxic type 2 ribosome-inactivating protein, ebulin a structurally related new homodimeric lectin in fruits, of Sambucus ebulus L. Planta. 1998;204:310–319. Citores L, de Benito FM, Iglesias R, et al. Characterization of a new non-toxic two-chain ribosome-inactivating protein a structurally-related lectin from rhizomes of dwarf elder (Sambucus ebulus L.). Cell Mol Biol. 1997;43:485–499. Barbieri L, Ciani M, Girbés T, et al. Enzymatic activity of toxic non-toxic type 2 ribosome-inactivating proteins. FEBS Lett. 2004;563:219–222. Nsimba-Lubaki M, Peumans WJ, Allen AK. Isolation and characterization of glycoprotein lectins from the bark of three species of elder, Sambucus ebulus, S. nigra and S. racemosa. Planta. 1986;168:113–118. Iglesias R, Citores L, Ferreras JM, et al. Sialic acid-binding dwarf elder four-chain lectin displays nucleic acid N-glycosidase activity. Biochimie. 2010;92:71–80. Girbés T, Citores L, Ferreras JM, et al. Isolation and partial characterization of nigrin b, a non-toxic novel type 2 ribosomeinactivating protein from the bark of Sambucus nigra L. Plant Mol Biol. 1993;22:1181–1186. Van Damme EJ, Barre A, Rougé P, et al. Characterization and molecular cloning of Sambucus nigra agglutinin V (nigrin b), a GalNAc-specific type-2 ribosome-inactivating protein from the bark of elderberry (Sambucus nigra). Eur J Biochem. 1996;237:505–513. Ferreras JM, Citores L, Iglesias R, et al. Occurrence and new procedure of preparation of nigrin, an antiribosomal lectin present in elderberry bark. Food Res Intl. 2011;44:2798–2805. Ferreras JM, Citores L, Iglesias R, et al. Use of ribosome-inactivating proteins from Sambucus for the construction of immunotoxins and conjugates for cancer therapy. Toxins. 2011;3:420–441. Muñoz R, Arias Y, Ferreras JM, et al. Sensitivity of cancer cell lines to the novel non-toxic type ribosome-inactivating protein nigrin b. Cancer Lett. 2001;167:163–169. Battelli MG, Musiani S, Buonamici L, et al. Interaction of volkensin with HeLa cells: binding, uptake, intracellular localization, degradation and exocytosis. Cell Mol Life Sci. 2004;61:1975–1984. Spooner RA, Lord JM. How ricin and Shiga toxin reach the cytosol of target cells: retrotranslocation from the endoplasmic reticulum. Current Top Microbiol Immunol. 2012;357:19–40. Battelli MG, Citores L, Buonamici L, et al. Toxicity and cytotoxicity of nigrin b, a two-chain ribosome-inactivating protein from Sambucus nigra: comparison with ricin. Arch Toxicol. 1997;71:360–364. Muñoz R, Arias Y, Ferreras JM, et al. Targeting a marker of the tumour neovasculature using a novel anti-human CD105immunotoxin containing the non-toxic type 2 ribosome-inactivating protein nigrin b. Cancer Lett. 2007;256:73–80.

NON-TOXIC TYPE 2 RIBOSOME-INACTIVATING PROTEINS

31. 32. 33. 34. 35. 36.

37. 38. 39. 40. 41. 42. 43.

44.

45.

46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

81

Horrix C, Raviv Z, Flescher E, et al. Plant ribosome-inactivating proteins type II induce the unfolded protein response in human cancer cells. Cell Mol Life Sci. 2011;68:1269–1281. Gayoso MJ, Muñoz R, Arias Y, et al. Specific dose-dependent damage of Lieberkühn crypts promoted by large doses of type ribosome-inactivating protein nigrin b intravenous injection to mice. Toxicol Appl Pharmacol. 2005;207:138–146. Citores L, de Benito FM, Iglesias R, et al. Isolation and characterization of a new non-toxic two chain ribosome-inactivating protein from fruits of elder elder (Sambucus nigra L.). J Exper Bot. 1996;47:1577–1585. Girbés T, Cítores L, de Benito FM, et al. A non-toxic two chain ribosome-inactivating protein coexists with a structurerelated monomeric lectin (SNA III) in elder (Sambucus nigra) fruits. Biochem J. 1996;315:343–344. Van Damme EJ, Roy S, Barre A, et al. The major elderberry (Sambucus nigra) fruit protein is a lectin derived from a truncated type ribosome-inactivating protein. Plant J. 1997;12:1251–1260. Citores L, Iglesias R, Muñoz R, et al. Elderberry (Sambucus nigra L.) seed proteins inhibit protein synthesis and display strong immunoreactivity with rabbit polyclonal antibodies raised against the type 2 ribosome-inactivating protein nigrin b. J Exp Bot. 1994;45:513–516. Peumans WJ, Kellens JT, Allen AK, Van Damme EJ. Isolation and characterization of a seed lectin from elderberry (Sambucus nigra L.) and its relationship to the bark lectins. Carbohydrate Res. 1991;213:7–17. de Benito FM, Citores L, Iglesias R, et al. Isolation and partial characterization of a novel and uncommon two-chain 64-kDa ribosome-inactivating protein from the bark of elder (Sambucus nigra L.). FEBS Lett. 1997;413:85–91. Van Damme EJ, Barre A, Rougé P, et al. Isolation and molecular cloning of a novel type 2 ribosome-inactivating protein with an inactive B chain from elderberry (Sambucus nigra) bark. J Biol Chem. 1997;272:8353–8360. Broekaert WF, Nsimba-Lubaki M, Peeters B, Peumans WJ. A lectin from elder (Sambucus nigra L.) bark. Biochem J. 1984;221:163–169. Greenwood JS, Stinisen HM, Peumans WJ, Chrispeels MJ. Sambucus nigra agglutinin is located in protein bodies in the phloem parenchyma of the bark. Planta. 1986;167:275–278. Van Damme EJ, Roy S, Barre A, et al. Elderberry (Sambucus nigra) bark contains two structurally different Neu5Ac(alpha2,6) Gal/GalNAc-binding type 2 ribosome-inactivating proteins. Eur J Biochem. 1997;245:648–655. Rojo MA, Yato M, Ishii-Minami N, et al. Isolation, cDNA cloning, biological properties, and carbohydrate binding specificity of sieboldin-b, a type II ribosome-inactivating protein from the bark of Japanese elderberry (Sambucus sieboldiana). Arch Biochem Biophys. 1997;340:185–194. Kaku H, Tanaka Y, Tazaki K, et al. Sialylated oligosaccharide-specific plant lectin from Japanese elderberry (Sambucus sieboldiana) bark tissue has a homologous structure to type II ribosome-inactivating proteins, ricin and abrin. cDNA cloning and molecular modeling study. J Biol Chem. 1996;271:1480–1485. Kaku H, Kaneko H, Minamihara N, et al. Elderberry bark lectins evolved to recognize Neu5Ac(α2,6)Gal/GalNAc sequence from a Gal/GalNAc binding lectin through the substitution of amino-acid residues critical for the binding to sialic acid. J Biochem. 2007;142:393–401. He WJ, Liu WY. Cinnamomin: a multifunctional type II ribosome-inactivating protein. Intl J Biochem Cell Biol. 2003;35:1021–1027. Wang BZ, Zou WG, Liu WY, Liu XY. The lower cytotoxicity of cinnamomin (a type II RIP) is due to its B-chain. Arch Biochem Biophys. 2006;451:91–96. Li XD, Liu WY, Niu CL. Purification of a new ribosome-inactivating protein from the seeds of of Cinnamomum porrectum and characterization of the RNA N-glycosidase activity of the toxic protein. Biol Chem. 1996;377:825–831. Cammue BP, Peeters B, Peumans WJ. Isolation and partial characterization of an N-acetylgalactosamine-specific lectin from winter-aconite (Eranthis hyemalis) root tubers. Biochem J. 1985;227:949–955. Kumar MA, Timm DE, Neet KE, et al. Characterization of the lectin from the bulbs of Eranthis hyemalis (winter aconite) as an inhibitor of protein synthesis. J Biol Chem. 1993;268:25176–25183. Zhang D, Halaweish FT. Isolation and characterization of ribosome-inactivating proteins from Cucurbitaceae. Chem Biodiv. 2007;4:431–442. Barbieri L, Lorenzoni E, Stirpe F. Inhibition of protein synthesis in vitro by a lectin from Momordica charantia and by other haemagglutinins. Biochem J. 1979;182:633–635. Mazumder T, Gaur N, Surolia A. The physicochemical properties of the galactose-specific lectin from Momordica charantia. Eur J Biochem. 1981;113:463–470. Chandran T, Sharma A, Vijayan M. Crystallization and preliminary X-ray studies of a galactose-specific lectin from the seeds of bitter gourd (Momordica charantia). Acta Crystallogr Sect F Struct Biol Cryst Commun. 2010;66:1037–1940. Nicolson GL, Blaustein J, Etzler ME. Characterization of two plant lectins from Ricinus communis and their quantitative interaction with a murine lymphoma. Biochem. 1974;13:196–204. Surolia A, Bachhawat BK, Vithyathil PJ, Podder SK. Unique subunit structure for Ricinus communis agglutinin. Indian J Biochem Biophys. 1978;15:248–250.

82 57. 58.

59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.

72.

RIBOSOME-INACTIVATING PROTEINS

Roberts LM, Lamb FI, Pappin DJ, Lord JM. The primary sequence of of Ricinus communis agglutinin. J Biol Chem. 1985;260:15682–15686. Citores L, Ferreras JM, Iglesias R, et al. Molecular mechanism of inhibition of mammalian protein synthesis by some fourchain agglutinins. Proposal of an extended classification of plant ribosome-inactivating proteins (rRNA N-glycosidases). FEBS Lett. 1993;329:59–62. Hegde R, Maiti TK, Podder SK. Purification and characterization of three toxins and two agglutinins from Abrus precatorius seed by using lactamyl-Sepharose affinity chromatography. Anal Biochem. 1991;194:101–109. Olsnes S, Refsnes K, Christensen TB, Pihl A. Studies on the structure and properties of the lectins from Abrus precatorius and Ricinus communis. Biochim Biophys Acta. 1975;405:1–10. Timoshenko AV, Cherenkevich SN, Gabius HJ. Viscum album agglutinin-induced aggregation of blood cells and the lectin effects on neutrophil function. Biomed Pharmacother. 1995;49:153–158. Van Damme EJ, Hao Q, Charels D, et al. Characterization and molecular cloning of two different type 2 ribosome-inactivating proteins from the monocotyledonous plant Polygonatum multiflorum. Eur J Biochem. 2000;267:2746–2759. Hao Q, Van Damme EJ, Hause B, et al. Iris bulbs express type 1 and type 2 ribosome-inactivating proteins with unusual properties. Plant Physiol. 2001;125:866–876. Citores L, Ferreras JM, Muñoz R, et al. Targeting cancer cells with transferrin conjugates containing the non-toxic type ribosome-inactivating proteins nigrin b or ebulin l. Cancer Lett. 2002;184:29–35. Bernabeu C, Lopez-Novoa JM, Quintanilla M. The emerging role of TGF-beta superfamily coreceptors in cancer. Biochim Biophys Acta. 2009;1792:954–973. Santibañez JF, Quintanilla M, Bernabeu C. TGF-β/TGF-β receptor system and its role in physiological and pathological conditions. Clin Sci (Lond). 2011;121:233–251. Benitez J, Ferreras JM, Muñoz R, et al. Cytotoxicity of an ebulin l-anti-human CD105 immunotoxin on mouse fibroblasts (L929) and rat myoblasts (L6E9) cells expressing human CD105. Med Chem. 2005;1:65–71. Muñoz R, Arias Y, Ferreras JM, et al. Targeting a marker of the tumour neovasculature using a novel anti-human CD105immunotoxin containing the non-toxic type 2 ribosome-inactivating protein nigrin b. Cancer Lett. 2007;256:73–80. Shahidi-Noghabi S, Van Damme EJ, Smagghe G. Expression of Sambucus nigra agglutinin (SNA-I′) from elderberry bark in transgenic tobacco plants results in enhanced resistance to different insect species. Transgenic Res. 2009;18:249–259. Shahidi-Noghabi S, Van Damme EJ, Iga M, Smagghe G. Exposure of insect midgut cells to Sambucus nigra L. agglutinins I and II causes cell death via caspase-dependent apoptosis. J Insect Physiol. 2010;56:1101–1107. Shahidi-Noghabi S, Van Damme EJ, Mahdian K, Smagghe G. Entomotoxic action of Sambucus nigra agglutinin I in Acyrthosiphon pisum aphids and Spodoptera exigua caterpillars through caspase-3-like-dependent apoptosis. Arch Insect Biochem Physiol. 2010;75:207–220. Shahidi-Noghabi S, Van Damme EJ, De Vos WH, Smagghe G. Internalization of Sambucus nigra agglutinins I and II in insect midgut CF-203 cells. Arch Insect Biochem Physiol. 2011;76:211–222.

6

The Intracellular Journey of Type 2 Ribosome-inactivating Proteins Robert A. Spooner and J. Michael Lord School of Life Sciences, University of Warwick, UK

Introduction

Type 2 RIPs are heterodimeric proteins in which a type 1 RIP (in this case, known as the A chain) is covalently linked by a disulfide bond directly to a cell-binding component (known as the B chain in plant type 2 RIPs) or indirectly to a pentameric cell-binding B chain complex (the bacterial Shiga toxin type 2 RIPs). The plant B chains are galactose-specific lectins, whereas the Shiga toxin (STx) B chain pentamer binds a membrane-embedded glycolipid. The B chain or pentameric B chain complex is thus responsible for target cell binding, the first step in cytotoxicity. This is followed by endocytosis, and intracellular trafficking in vesicular carriers, via early/sorting endosomes and the Golgi apparatus, to the lumen of the endoplasmic reticulum (ER) where the A and B chains are reductively separated, allowing the catalytically active A chains to cross the ER membrane to reach their cytosolic targets, the ribosomes. To date, there have been limited point-by-point comparisons of type 2 RIPs from any source, so general conclusions have been based largely upon findings with exemplars, typically ricin and STx. However, cytotoxicity is multifactorial, depending upon the efficiencies of cell binding and toxin internalization; the choice of retrograde trafficking route; and the rates of reductive separation of the A and B chain(s), A chain entry to the cytosol (dislocation), and recovery of catalytic activity of the A chain in the cytosol, as well as the catalytic rate of the A chain against its target ribosomes. Thus, the general conclusions hide a number of intriguing facts: for example, when examining toxicity of three type 2 plant RIPs from Adenia, the cellular uptake of lanceolin and stenodactylin was ~10 and ~40 fold greater, respectively, than that reported for volkensin, even though the cell-binding properties were similar.1, 2 Whilst the broad description of events required for cytotoxicity given above is correct, few molecular details are known for the trafficking of plant type 2 RIPs, even for the most well-characterized: ricin. This is in marked contrast to the depth of knowledge already acquired for STx trafficking. For the later stages in cytotoxicity, such as the ER translocation step, molecular mechanisms have been uncovered for ricin, and recently have started to emerge for STx. In this chapter we discuss current knowledge of these processes.

Ribosome-inactivating Proteins: Ricin and Related Proteins, First Edition. Edited by Fiorenzo Stirpe and Douglas A. Lappi. © 2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.

83

84

RIBOSOME-INACTIVATING PROTEINS

Cell Surface Events

Ricin is the exemplar for all other type 2 RIPs from plant sources, which have remarkably similar overall structures (Figure 6.1). It is produced in the developing endosperm of castor oil seeds from Ricinus communis,3 as a pre-proprotein that is subsequently matured by proteolytic cleavage. Mature holotoxin comprises the catalytically active A chain (RTA, red), disulfide-linked to its B chain (RTB, green) (Figure 6.1). Like all type 1 RIPs, RTA specifically depurinates 28S rRNA, thus halting protein synthesis.4 RTB binds terminal non-reducing (β1 → 4 linked) galactose residues.5 Thus, multiple glycoproteins and glycolipids at the cell surface may act as binding sites, making the identification of specific receptors arduous. N-glycosylated proteins have been proposed to act as the sole receptors for ricin toxicity.6 However, mammalian cells that cannot synthesize complex N-glycans on target proteins because of a lack of N-acetylglucosaminyl transferase 1 activity7 are only protected from a ricin challenge by approximately 20-fold.8 This suggests that ~5% of productive receptors (those whose binding results in cytotoxicity) are non-protein. Productive routing of ricin does not appear to depend upon recruitment of ricin-bound receptors to clathrin-coated pits, because arresting coated-pit formation at the cell surface does not alter ricin cytotoxicity, even though there is a 50% reduction in overall ricin internalization.9, 10 Productive routing also appears to be caveolae-independent.11 Since the effect of inhibiting individual entry pathways makes little difference overall, ricin is thought to use many productive receptors that can enter cells via multiple mechanisms. However, in a haploid mammalian cell screen for ricin resistance, a very strong

RTA

Abrin A

Abrin B

RTB

STxA

STxB pentamer

RTA

STxA1 STxA2

RTB

5x STxB

Figure 6.1 Crystal structures (above) and cartoon representations (below) of the A-B toxin ricin (left) compared with the A-B toxin abrin, underlining the similarity in structure (middle) and with the A-B5 toxin Shiga (STx) toxin. Crystal structures (PDB codes 2AAI,96 1ABR,97 1S5E,98 respectively) were viewed in RasMol and are shown from the side with the receptor-binding surfaces of the B chains (green) facing downwards. Arrowheads in the cartoons show the site of proteolytic cleavage that is required for activation of STx, separating the A chain (red) into A1 and A2 products which remain in close association, held by both noncovalent interactions and by a disulfide bond linking the two (orange). The A1 chains dislocate from the ER, leaving the A2 portion associated with the pentameric B ring. The A and B chains of ricin and abrin are held together by hydrophobic interactions and a disulfide bond. (This figure also appears in the color plate section.)

THE INTRACELLULAR JOURNEY OF TYPE 2 RIBOSOME-INACTIVATING PROTEINS

85

requirement for the cell surface orphan GPCR Gpr107 was reported for normal cytotoxicity,12 thus providing a candidate for a preferred ricin receptor. Similarly, the B chains of other plant type 2 RIPs bind exposed galactose residues. Of the four characterized pulchellin isoforms from seeds of Abrus pulchellus, subtle differences have been noted in these B chain-mediated cell surface interactions,13 and a comparison of ricin and abrin isoforms (from the seeds of A. precatorius) suggests that such subtle differences may account in part for differences in the kinetics of intoxication.14 To date, definitive receptors are unknown, although crosslinking experiments suggest the involvement of a 45-kDa protein candidate for abrin toxicity.15 Competition with various glycoproteins indicates that abrin receptors differ from those used by modeccin from Adenia digitata.16 For the type 2 plant RIPs, uncertainty over events at the cell surface means that trafficking pathways productive for cytotoxicity cannot be studied by examining the behavior of known receptors. Shiga toxin (STx) is synthesized by the bacterium Shigella, and the essentially identical Shigalike toxins by enterohemorrhagic strains of Escherichia coli. In contrast to the plant type 2 RIPs, the holotoxin comprises an A chain (STxA, red) that pierces the central channel of a doughnut-shaped ring of five B-chain subunits (STxB pentamer, various colors, Figure 6.1). STxA is cleaved by furin in the early stages of mammalian cell intracellular transport, giving an activated A1-chain disulfide linked to the A2 chain that remains associated with the B-chain pentamer.17 The STxA1 chain subsequently crosses the ER membrane to enter the cytosol18 where it specifically depurinates the large ribosomal subunit, in a catalytically identical process to that performed by RTA.19 Each STxB chain subunit can bind three molecules of its receptor, the glycolipid globotriaosylceramide or Gb3,20 and in doing so reorganizes the lipid membrane.21, 22 This cargo-induced promotion of membrane curvature and invagination also occurs upon binding of cholera toxin and polyoma viruses to cells, since these crosslink their receptors (the ganglioside lipid GM1) with the same spatial geometry.23 STx further regulates its own entry by stimulating activity of Syk kinase.24 In stark contrast, there is no suggestion that the binding of plant type 2 RIPs can force membrane curvature or regulate their own endocytosis.

Endosome to TGN Sorting

Endocytotic pathways for ricin appear to merge at the early/sorting endosome (Figure 6.2). At least a portion of an endosomal trafficking toxin must avoid sorting to the destructive environment of the lysosome, a process that appears to be inefficient for ricin since only approximately 5% of cell-surface bound ricin is transported to the Golgi.25 A candidate for this difficult step is access to retromer components. Mammalian retromer mediates retrograde transport between endosomes and the trans-Golgi network (TGN) and appears to sequester membrane cargo proteins from vacuolar endosomal membranes into retrograde transport intermediates, thereby preventing default delivery to lysosomes.26 The retromer coat is composed of a heterodimer of different sorting nexin (SNX) combinations, a Vps26–Vps29–Vps35 trimer and the kinase hVps34.27, 28 Roles for SNX2, SNX4, SNX8, and hVps34 have been identified for transport of ricin from early endosomes to the TGN.29, 30 A level of uncertainty exists for these early endosome events, which may indicate confounding cell-type specific effects. Thus, the role of the small GTPase dynamin acting as a scission agent is not clear, with expression of mutant dynamin having either no effect on ricin toxicity in COS7 cells,11 or else protecting HeLa cells about ten-fold from ricin challenge by reducing retrograde traffic to the Golgi stack.31 Docking with the TGN requires syntaxin 16.32

86

RIBOSOME-INACTIVATING PROTEINS

Ricin

STx

Receptor binding Endocytosis

Lysosomal sorting Endosomal sorting

PDI

BiP

TGN docking

Golgi trafficking

ER arrival Reductive separation Unfolding Dislocation Folding RTA

STxA1

Figure 6.2 Trafficking schemes. Ricin and STx bind their respective receptors at the plasma membrane and after internalization by endocytosis, traffic via early endosomes, the TGN and the Golgi stack to the ER, where the toxic polypeptides (A or A1 chains) are processed by ER chaperones (gray-blue) resulting in reductive separation from the holotoxin and maintenance of solubility prior to dislocation and recovery of activity in the cytosol. Despite the common stages in trafficking, and a common docking mechanism at the TGN for ricin and STx, the routes taken by these toxins are otherwise idiosyncratic. (This figure also appears in the color plate section.)

In marked contrast to the sparse description of endosomal sorting events for ricin, multiple details are known for STx:33, 34 indeed, the isolated STxB-chain pentamer has proved a very useful tool for defining endosomal components of the retrograde pathway, since recombinant STxB pentamer tagged with a “sulf” peptide can be labeled in vivo during endosome-to-Golgi passage35 by sulfotransferases that reside at steady state in the trans-Golgi cisternae.36 A demonstration that TGN docking requires syntaxin 16 suggests a common retrograde route is followed by ricin and STx.32 Strikingly, however, the requirement for sorting nexins differs from that of ricin with roles that promote retrograde trafficking for SNX137–39 as well as SNX239 and an inhibitory role for SNX8.30 These differences in requirements for retrograde trafficking are not confined to a comparison between ricin and STx. Endosomal sorting pathways, followed by cholera toxin and STx, can be separated at the endosomal level by use of the small molecule Exo2 which strongly blocks egress of  STx from early endosomes36 but has little or no effect on CTx trafficking.40 In addition, the clathrin associated Hcs70 co-chaperone RME-8 regulates endosomal trafficking of STx37 whilst having no effect on the transport of cholera toxin.41 In a genome-wide RNAi screen comparing cellular requirements for ricin and pseudomonas exotoxin (PE) intoxication, striking differences were seen even though factors required by both toxins are present from the endosomes to the ER,42 in part because pseudomonas exotoxin can utilize multiple retrograde trafficking routes.43

THE INTRACELLULAR JOURNEY OF TYPE 2 RIBOSOME-INACTIVATING PROTEINS

87

That different retrograde cargo molecules use different components of retromer complexes is underlined by a comparison of the trafficking of the endogenous cargo TGN38 and the exogenous cargo STxB, which utilize different TGN golgins; transport of TGN38 requires the TGN golgin GCC88 whereas that of Shiga toxin requires GCC185.44, 45 The bulk of Shiga toxin is transported from early endosomes to recycling endosomes en route to the TGN, whereas the bulk of TGN38 is transported from early endosomes to the TGN with only low levels detected in recycling endosomes, decisions that appear to be made by SNX1 and SNX2, respectively.46 In addition, a comparison of STx and epidermal growth factor trafficking revealed differences in the requirements for Rab GTPase activating proteins.47 Taken together, these studies identify different itineraries for the retrograde transport of different toxins and endogenous cargos and distinct regulators of the transport of these cargos. These findings suggest that the non-toxicity of the type 2 RIPs ebulin(s) and nigrin(s), purified from Sambucus tissues,48 may in part reflect failure of these proteins to access retrograde pathways. Small differences in protein structure have been suggested to underlie this inability to enter a productive intoxication pathway.49

From the Golgi to the ER

To aid description of the productive pathway, C-terminally tagged versions of RTA reconstituted into holotoxin with plant-derived RTB have been employed. One of these, RTA-sulf1, bears a tag that can be sulfated in vivo by the addition of [35S]O42− in a reaction catalyzed by the trans-Golgi cisternal residents TPST1 and TPST2,36 and so measures retrograde arrival of the toxin at the Golgi stack.50 Another, RTA-sulf2, bears the same sulfation signal and an additional set of motifs that can be N-glycosylated by oligo-saccharyl transferase complexes upon arrival in the ER lumen.50 RTA-sulf2-containing holotoxin has been used to demonstrate cytosolic delivery of a sulfated N-glycosylated RTA,51 thus defining a plasma membrane–Golgi–ER– cytosol retrograde trafficking route (Figure  6.2). Ricin negotiates the Golgi stack without using the COPI-dependent route that typically characterizes retrograde Golgi transport52, 53 and, under some circumstances, may bypass the Golgi altogether.54 The role of the Golgi stack in retrograde transport of plant type 2 RIPs is, therefore, unclear. Despite the remarkable structural similarities of ricin and nigrin,55 and co-localization of nigrin with Golgi structures on treated mammalian cells, treatment with brefeldin A (that protects cells from ricin by denying access to the retrograde route) does not appear to alter the intracellular location of nigrin.2 Thus, even though a proportion of nigrin can access the Golgi stack, it still remains non-toxic, suggesting either that it cannot access the ER from the Golgi, or that some other factor prevents its toxicity. Similarly, STx negotiates the Golgi stack in a COPI-independent manner. Golgi56–58 and Golgi functions are essential for STx toxicity, since subcellular surgery to remove the Golgi stack halts the toxin retrograde transport process.59 Nevertheless, a discrete Golgi stack is not necessary, since a derivative of the Arf-GEF inhibitor Exo2 that fuses Golgi and ER membranes and disrupts the remaining Golgi into small punctae has no measurable effect on STx toxicity.60 It appears that whilst plant and bacterial type 2 RIPs traffic from early endosomes to the trans-Golgi network and through the Golgi stack, each follows an idiosyncratic route (Figure 6.2). Nevertheless, for the toxic type 2 RIPs, the terminal destination of holotoxin is the ER lumen.50, 61–63

88

RIBOSOME-INACTIVATING PROTEINS

Reductive Separation and Destabilization of the Holotoxin Subunits

Proteins destined for the secretory pathway fold and mature in the ER, processes that are overseen by ER quality control surveillance, which may involve a combination of retrieval from post-ER compartments, ER retention, autophagy, and/or ER-associated protein degradation (ERAD) mechanisms.64 ERAD comprises a set of processes that recognize terminally-misfolded and orphan proteins and remove them from the ER (“dislocation”). Recognition of these substrates occurs in part by ER chaperones65 that maintain their solubility before removal. The dislocation process usually involves polyubiquitylation of the target protein on lysine residues as it is extruded through the “dislocon” to enter the cytosol.66, 67 This polyubiquitylation ultimately provides tags for interactions with the cytosolic proteasomes,68, 69 where the ERAD substrates are destroyed.70 The paucity of lysine residues in the A chains of some ER trafficking toxins71 suggested that, upon arrival in the ER, the A chains could enter the cytosol using an ERAD-type pathway, but largely avoid the ubiquitylation on lysine residues that results in proteasomal destruction. This in turn would suggest that reduction of the holotoxin to constituent A and B chains might occur in the ER, and that chaperone interactions in the ER might maintain the solubility of the A chains prior to an ERAD-like process of dislocation. The ER resident chaperone protein, disulfide isomerase (PDI), reduces the disulfide bond that links the catalytic A (RTA) and cell-binding B (RTB) chains of ricin holotoxin.63, 72, 73 Curiously, in in vitro experiments, the ability of the small reducing agent β-mercaptoethanol to reduce abrin holotoxins is influenced by the presence of lactose,14 with lactose binding of the B chain reducing the rate of interchain dissociation by 11- and 3-fold, respectively, for abrin I and abrin III. This suggests a remodeling of the holotoxin structure at the interface between A and B chains can be stimulated by substrate binding: the subtle differences noted for substrate binding between different plant type 2 RIPs may also influence toxicity by altering the rate of A chain release. When RTA is expressed directly in the ER of the yeast Saccharomyces cerevisiae, it dislocates, enters the cytosol, and destroys sufficient of its ribosomal targets to cause a severe growth defect.74 Under these conditions, deletion of either of the two known nucleotide exchange factors of Kar2p (the yeast equivalent of the mammalian ER chaperone BiP) reduces the growth defect, suggesting that there is a role for BiP in the maintenance of solubility of free RTA in the ER lumen.75 In mammalian cells, a role for EDEM has been proposed in transporting free RTA to the “dislocon” for ER removal.76 Roles have also been proposed for the Hsp90 family ER chaperone GRP94 in ricin toxicity to mammalian cells.77, 78 Reductive separation of RTA and RTB exposes a relatively hydrophobic patch of amino acids towards the C terminus of RTA. When RTA is tagged with fluorophores in different positions around the molecule, fluorophore tags in this relatively hydrophobic stretch are quenched in the presence of microsomal membranes pre-soaked with a lipopholic quencher, whilst fluorophores elsewhere are not.79 Thus, RTA makes non-random interactions with ER membranes and at least part of the hydrophobic C-terminal patch of RTA enters the non-polar core of the membrane. The process is temperature-dependent: as temperature increases, structural changes associated with membrane entry become more apparent.79 Isolated purified RTA is also temperature-sensitive in the absence of microsomes, and at 37 ºC it is relatively unstable and prone to aggregation.77 The key to ricin A chain toxicity, then, lies in destabilization of the holotoxin upon ER luminal arrival: reductive separation releases the catalytic A chain from its ER-targeting B chain and thermal instability of RTA stimulates chaperone interactions and an ordered insertion into the ER membrane. This membrane-embedded form is thought to mimic a misfolded protein that is then dislocated from the ER in an ERAD-like manner. The final recognition of misfolded RTA as a dislocatable

THE INTRACELLULAR JOURNEY OF TYPE 2 RIBOSOME-INACTIVATING PROTEINS

89

substrate appears to be by membrane-integral components of the HRD ubiquitin ligase complex that probably forms the dislocon.75, 80 Since the A1 and A2 chains of STx are disulfide-linked, then a reductive event is expected prior to dislocation of SLTxA1, although to our knowledge this has not yet been characterized. STxA1 interacts with a pre-assembled ER luminal protein complex containing the chaperones HEDJ, BiP, and GRP94.81, 82 The A1 chain of STx also possesses a relatively hydrophobic string of amino acids, which is required for its dislocation when expressed in the yeast ER.18 Since an artificial peptide based on this sequence interacts with lipid membranes,83, 84 it has been suggested that STxA also interacts with ER membranes. However, experimental testing reveals only a minor role for potential peptide interaction in the context of STxA1 dislocation.85

Dislocation Across the ER Membrane

The RTA dislocation process is largely unmapped in mammalian cells: there are suggestions that the translocon protein Sec61 is part of a dislocation channel.51, 76, 86 A role for the SEL IL component of ERAD dislocation complexes has been described;80 there is no obvious function of derlins, other components of mammalian dislocation systems.76 It may be that, in mammalian cells, RTA utilizes ERAD-associated proteins selectively. However, the severe growth defect caused by expressing RTA in the yeast ER lumen allowed us to examine the requirements for RTA dislocation by identifying single gene knockout strains that grew normally because they retained RTA in the ER rather than dislocating it.75 Recently, we have utilized a similar approach to address the requirements for STxA1 dislocation.85 RTA utilizes the integral membrane protein HRD E3 ubiquitin ligase complex for dislocation (Figure 6.3), but uses components of this complex selectively, with strong requirements for the multi-spanning Hrd1p ubiquitin ligase and its Hrd3p regulator, intermediate requirements for the Der1p subunit that appear to recognize misfolded luminal domains, and minor requirements for the Usa1p subunit that optimizes Hrd1p function and for the Cue1p and Ubx7p subunits that promote ubiquitylation of the dislocated substrate.75 Only a small proportion recovers activity in the cytosol, the remainder being degraded, but not by the proteasome. The extraction motor for RTA appears to be the Rpt4p subunit of the proteasomal cap, which has previously been shown to play a role in extraction of other substrates in conjunction with Cdc48p.69 A comparison with STxA1 dislocation reveals differential use of ERAD components.85 STxA1 has strong requirements for Hrd1p and Hrd3p, but unlike RTA also has strong requirements for Der1p, suggesting that recognition of STxA1 occurs principally by examining a luminal rather than membrane-embedded substrate (Figure 6.3). There is also a strong requirement for Usa1p, suggesting that an optimized dislocon is required. The major dislocating population is polyubiquitylated during dislocation, is extracted by Cdc48p and its ubiquitin handling co-factors, and is destroyed in the proteasome. However, for the proportion of toxin that regains catalytic activity in the cytosol, dislocation is independent of canonical ubiquitylation, even though the catalytic cysteine of Hrd1p is required. The extraction motor for the population of STxA1 that regains activity in the cytosol remains unknown. Thus, RTA appears to be dislocated promiscuously through non-optimized dislocons in a manner that either displaces the RING-H2 ubiquitin ligase domain of Hrd1p or avoids it altogether. In contrast, STxA1 appears to be dislocated through optimized dislocons in a manner that requires the RING-H2 ubiquitin ligase domain of Hrd1p.

90

RIBOSOME-INACTIVATING PROTEINS

RTA

STxA1

Der1

Hrd1

Hrd1

Ubx2

Cytosol

Usa1 Der1

Hrd3

Hrd3

ER lumen

H2

Proteasome

RTA

STxA1

Cdc48-Npl4

Proteasome

Figure 6.3 RTA and SLTxA1 expressed exogenously in the yeast ER lumen dislocon utilize components of the Hrd1p complex selectively. Strong requirements for membrane-associated members of the complex are indicated by a hard outline: intermediate requirements by a dashed outline. Dislocation of RTA requires the core components of the Hrd1p–Hrd3p dislocon, but not the Hrd1p E3 ubiquitin ligase activity encoded by its RING-H2 domain (H2). Extraction from the dislocon requires the Rpt4p subunit of the proteasome cap. For STxA1, the RING-H2 domain is required for extraction of both the bulk population by ERAD-enabled Cdc48 complexes that is destroyed by the proteasome, and also for the population that recovers activity in the cytosol, even though this latter appears to be a ubiquitin-independent process. (This figure also appears in the color plate section.)

Cytosolic Post-dislocation Events that Restore Catalytic Activity

Since the A chains are thought to enter the cytosol in a substantially disordered conformation, then recovery of catalytic activity is expected to require extensive folding in the cytosol (Figure 6.4). In mammalian cells, post-dislocation scrutiny by the cytosolic chaperone Hsc70 is required for RTA to gain a catalytic conformation,77 followed by specific depurination of the ribosomal targets87 and subsequent cell death. Since RTA carefully unfolded to a molten globule state can fold in the presence of substrate ribosomes,88 Hsc70 may aid solubility of dislocated RTA, allowing substratemediated refolding in vivo. Alternatively, Hsc70 may stabilize RTA in the cytosol by masking the hydrophobic patch that interacts with membranes. Hsc70 co-chaperones regulate the folding process, with some (HIP and BAG-2) promoting activation of cytosolic RTA, and others, such as the proteasome-engaging BAG-1, promoting inactivation. A proportion of RTA is passed from Hsc70, via the Hsc70–Hsp90 organizing protein HOP, to the Hsp90 chaperone. From here, the net fate of RTA is inactivation. In vitro, RTA can be ubiquitylated by mimicking the Hsc70–Hsp90 sequential triage, suggesting that Hsp90 interactions inactivate RTA by promoting cytosolic ubiquitylation.77 Although RTA is not ubiquitylated during dislocation, a low level of cytosolic ubiquitylation does occur via an unknown E3 ligase.75 Thus, a network of chaperones determines RTA fate in the cytosol by regulating the competing processes of folding and ubiquitin-tagging. For STx, experimental tests of proteasome inhibition give conflicting results, with a report that this results in a slight increase in cytotoxicity of the toxin, suggesting that there is a role for ubiquitin

THE INTRACELLULAR JOURNEY OF TYPE 2 RIBOSOME-INACTIVATING PROTEINS

91

Dislocated RTA

HSP40 HSC70

HOP HIP HSP90

CHIP BAG-2 BAG-1 Ub

Inactivation RTA

Activation

Figure 6.4 Post-dislocation scrutiny by a chaperone/co-chaperone network determines the cytosolic fate of RTA. Non-native dislocated RTA is recognized by Hsp40 and Hsc70 and from this chaperone-bound state, routes lead to activation (folding) and inactivation. HIP (the Hsc70-interacting protein) stabilizes the Hsc70:RTA interaction. Release of RTA from this complex by BAG family guanine nucleotide exchange factors can take place in the vicinity of the proteasome (via the interlaced ubiquitin-like domain of BAG-1) or away from the proteasome (via BAG-2), suggesting that inactivation may occur by proteasomal degradation of RTA. Transfer of RTA from Hsc70 to Hsp90 via the Hsc70–Hsp90 operating protein HOP leads to CHIP-mediated ubiquitylation (Ub) of RTA and subsequent inactivation. (This figure also appears in the color plate section.)

tagging in the fate of cytosolic STx,89 and a report that cytotoxicity is not affected by proteasomal inhibition, suggesting no obvious role for ubiquitin tagging.90 In retrospect, the model that low lysine content on the A chains of some protein toxins suggested ER trafficking, engagement with ERAD processes, and avoidance of canonical ubiquitylation and subsequent proteasomal degradation, was prescient.71 In this respect, a (testable) rationale may be made for the extremely low cytotoxicity of the Sambucus RIPs ebulin and nigrin: in contrast to

92

RIBOSOME-INACTIVATING PROTEINS

STxA1, RTA, abrin A, and other potently cytotoxic A chains, which typically retain two lysine residues per A chain, the nigrin and ebulin A chains have six and eight, respectively. This suggests that cytotoxicity may correlate with potential for polyubiquitylation during dislocation, leading to proteasomal destruction.

Concluding Remarks

Investigations have demonstrated retrograde trafficking for a wide range of type 2 RIPs, and examination of pre-dislocation, dislocation, and post-dislocation events has provided considerable mechanistic and molecular details for the exemplars ricin and STx. To date, there have been only limited point-by-point comparisons with other type 2 RIPs. Overall, it becomes clear that the common themes that have emerged give a broad description only: cell surface binding events are followed by internalization, retrograde trafficking of a proportion of toxin via endosomal compartments and the Golgi complex to the ER, destabilization and reduction of the toxin in the ER, A chain dislocation, and recovery of activity by folding processes in the cytosol. However, the details of each step vary: large differences are seen between the plant type 2 RIPs and the bacterial STx, but subtler differences between the plant type 2 RIPs result in a wide range of cytotoxicities, from the very toxic volkensin to the essentially non-toxic ebulins and nigrins. Subtle differences in galactose-binding at the cell surface may allow idiosyncratic selection of receptors that are productive for cytotoxicity, thus influencing subsequent trafficking pathways, resulting in complex decisions made at the endosomal sorting level. The finding that substrate binding also alters propensity for subunit reduction suggests that subtle differences in cell-surface events may have far-reaching consequences. Knowledge of ER, dislocation, and post-dislocation events has been gleaned from a comparison of the two exemplars ricin and STx only: not surprisingly, differences are seen with ER chaperone interactions and choice of ERAD components. Thus, the conclusion that emerges is that each type 2 RIP intoxicates mammalian cells in an idiosyncratic manner, within the broad confines of the overall trafficking scheme. The proportion of toxic RTA or STxA1 that recovers activity in the cytosol dislocates in an ubiquitin-independent manner. The yeast viral AB toxin K28 also dislocates from the ER in an ubiquitin-independent manner and its A chain recovers catalytic activity in the cytosol.91 Human hepatitis E virus ORF2 protein is initially inserted into the ER lumen and subsequently accumulates in the cytosol of infected cells.92 Similarly, the hepatitis b precore protein retrotranslocates from the ER to the cytosol without being destroyed.93 Thus, the fate of a protein that is dislocated by Hrd1p is not necessarily destruction. This suggests that toxic type 2 RIP A chains and some viral proteins have evolved to utilize ERAD to gain cytosolic access, but uncouple from the destructive later phases. However, calreticulin, a protein normally regarded as an ER resident, has been identified in mammalian cytosolic extracts where it appears to have a role in integrin binding. This cytosolic pool of calreticulin is derived from an ER population and so has presumably been dislocated in an ERAD-like manner and then been refolded in the cytosol rather than being destroyed.94 Proteins with a known stable structure, such as firefly and Renilla luciferases, can also be taken up macropinocytotically by dendritic cells and trafficked in vesicular structures to the ER, where they unfold to utilize ERAD components, subsequently refolding in the cytosol.95 These findings suggest that ERAD is simply a subset of fundamental retro-translocation events that result in substrate destruction. If so, the RIP A chains and viral proteins that refold in the cytosol might simply utilize selective components of a dislocation process rather than subvert ERAD. The key to uncoupling from the

THE INTRACELLULAR JOURNEY OF TYPE 2 RIBOSOME-INACTIVATING PROTEINS

93

final destructive stage of ERAD has been identified by study of toxins: avoidance of polyubiquitylation via the membrane-integral E3 ligase Hrd1. In turn, this permits bypass of Cdc48/p97 interactions for cytosolic extraction and subsequent proteasomal presentation.

Acknowledgements

This work was supported by Wellcome Trust Programme Grant 080566/Z/06/Z.

References 1. 2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

Battelli MG, Scicchitano V, Polito L, et al. Binding and intracellular routing of the plant-toxic lectins, lanceolin and stenodactylin. Biochim Biophys Acta. 2010;1800:1276–1282. Battelli MG, Musiani S, Buonamici L, et al. Interaction of volkensin with HeLa cells: binding, uptake, intracellular localization, degradation and exocytosis. Cell Mol Life Sci. 2004;61:1975–1984. Harley SM, Beevers H. Lectins in castor bean seedlings. Plant Physiol. 1986;80:1–6. Endo Y, Tsurugi K. RNA N-glycosidase activity of ricin A-chain. Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes. J Biol Chem. 1987;262:8128–8130. Olsnes S, Saltvedt E, Pihl A. Isolation and comparison of galactose-binding lectins from Abrus precatorius and Ricinus communis. J Biol Chem. 1974;249:803–810. Spilsberg B, Van Meer G, Sandvig K. Role of lipids in the retrograde pathway of ricin intoxication. Traffic. 2003;4:544–552. Reeves PJ, Callewaert N, Contreras R, Khorana HG. Structure and function in rhodopsin: high-level expression of rhodopsin with restricted and homogeneous N-glycosylation by a tetracycline-inducible N-acetylglucosaminyltransferase I-negative HEK293S stable mammalian cell line. Proc Natl Acad Sci U S A. 2002;99:13419–13424. Crispin M, Chang VT, Harvey DJ, et al. A human embryonic kidney 293 T cell line mutated at the Golgi alpha-mannosidase II locus. J Biol Chem. 2009;284:21684–21695. Moya M, Dautry-Varsat A, Goud B, et al. Inhibition of coated pit formation in Hep2 cells blocks the cytotoxicity of diphtheria toxin but not that of ricin toxin. J Cell Biol. 1985;101:548–559. Sandvig K, Olsnes S, Petersen OW, van Deurs B. Inhibition of endocytosis from coated pits by acidification of the cytosol. J Cell Biochem. 1988;36:73–81. Simpson JC, Smith DC, Roberts LM, Lord JM. Expression of mutant dynamin protects cells against diphtheria toxin but not against ricin. Exp Cell Res. 1998;239:293–300. Elling U, Taubenschmid J, Wirnsberger G, et al. Forward and reverse genetics through derivation of haploid mouse embryonic stem cells. Cell Stem Cell. 2011;9:563–574. Castilho PV, Goto LS, Roberts LM, Araujo AP. Isolation and characterization of four type 2 ribosome inactivating pulchellin isoforms from Abrus pulchellus seeds. FEBS Journal . 2008;275:948–959. Hegde R, Karande AA, Podder SK. The variants of the protein toxins abrin and ricin. A useful guide to understanding the processing events in the toxin transport. Eur J Biochem. 1993;215:411–419. Tsuzuki J, Wu HC. Receptors for a cytotoxic lectin, abrin and their role in cell intoxication. Biochim Biophys Acta. 1982;720:390–399. Olsnes S, Sandvig K, Eiklid K, Pihl A. Properties and action mechanism of the toxic lectin modeccin: interaction with cell lines resistant to modeccin, abrin, and ricin. Journal of supramolecular structure. J Supramol Struct. 1978;9:15–25. Garred O, van Deurs B, Sandvig K. Furin-induced cleavage and activation of Shiga toxin. J Biol Chem. 1995;270: 10817–10821. LaPointe P, Wei X, Gariepy J. A role for the protease-sensitive loop region of Shiga-like toxin 1 in the retrotranslocation of its A1 domain from the endoplasmic reticulum lumen. J Biol Chem. 2005;280:23310–23318. Endo Y, Tsurugi K, Yutsudo T, et al. Site of action of a Vero toxin (VT2) from Escherichia coli O157:H7 and of Shiga toxin on eukaryotic ribosomes. RNA N-glycosidase activity of the toxins. Eur J Biochem. 1988;171:45–50. Jacewicz M, Feldman HA, Donohue-Rolfe A, et al. Pathogenesis of Shigella diarrhea. XIV. Analysis of Shiga toxin receptors on cloned HeLa cells. J Infect Dis. 1989;159:881–889. Römer W, Berland L, Chambon V, et al. Shiga toxin induces tubular membrane invaginations for its uptake into cells. Nature. 2007;450:670–675.

94 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

41. 42. 43. 44. 45. 46. 47. 48. 49. 50.

RIBOSOME-INACTIVATING PROTEINS

Windschiegl B, Orth A, Römer W, et al. Lipid reorganization induced by Shiga toxin clustering on planar membranes. PLoS One. 2009;4:e6238. Ewers H, Romer W, Smith AE, et al. GM1 structure determines SV40-induced membrane invagination and infection. Nat Cell Biol. 2010;12:11–18; sup pp. 11–12. Lauvrak SU, Walchli S, Iversen TG, et al. Shiga toxin regulates its entry in a Syk-dependent manner. Mol Biol Cell. 2006;17:1096–1109. van Deurs B, Sandvig K, Petersen OW, et al. Estimation of the amount of internalized ricin that reaches the trans-Golgi network. J Cell Biol. 1988;106:253–267. Arighi CN, Hartnell LM, Aguilar RC, et al. Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. J Cell Biol. 2004;165:123–133. Bonifacino JS, Rojas R. Retrograde transport from endosomes to the trans-Golgi network. Nat Rev Mol Cell Biol. 2006;7:568–579. Bonifacino JS, Hurley JH. Retromer. Curr Opin Cell Biol. 2008;20:427–436. Skånland SS, Walchli S, Utskarpen A, et al. Phosphoinositide-regulated retrograde transport of ricin: crosstalk between hVps34 and sorting nexins. Traffic. 2007;8:297–309. Dyve AB, Bergan J, Utskarpen A, Sandvig K. Sorting nexin 8 regulates endosome-to-Golgi transport. Biochem Biophys Res Commun. 2009;390:109–114. Llorente A, Rapak A, Schmid SL, et al. Expression of mutant dynamin inhibits toxicity and transport of endocytosed ricin to the Golgi apparatus. J Cell Biol. 1998;140:553–563. Amessou M, Fradagrada A, Falguieres T, et al. Syntaxin 16 and syntaxin 5 are required for efficient retrograde transport of several exogenous and endogenous cargo proteins. J Cell Sci. 2007;120:1457–1468. Johannes L, Popoff V. Tracing the retrograde route in protein trafficking. Cell. 2008;135:1175–1187. Johannes L, Römer W. Shiga toxins--from cell biology to biomedical applications. Nat Rev Microbiol. 2010;8:105–116. Amessou M, Popoff V, Yelamos B, et al. Measuring retrograde transport to the trans-Golgi network. Curr Protoc Cell Biol. 2006;Chapter 15:Unit 15.10. Spooner RA, Watson P, Smith DC, et al. The secretion inhibitor Exo2 perturbs trafficking of Shiga toxin between endosomes and the trans-Golgi network. Biochem J. 2008;414:471–484. Popoff V, Mardones GA, Bai SK, et al. Analysis of articulation between clathrin and retromer in retrograde sorting on early endosomes. Traffic. 2009;10:1868–1880. Bujny MV, Popoff V, Johannes L, Cullen PJ. The retromer component sorting nexin-1 is required for efficient retrograde transport of Shiga toxin from early endosome to the trans Golgi network. J Cell Sci. 2007;120:2010–2021. Utskarpen A, Slagsvold HH, Dyve AB, et al. SNX1 and SNX2 mediate retrograde transport of Shiga toxin. Biochem Biophys Res Commun. 2007;358:566–570. Feng Y, Jadhav AP, Rodighiero C, et al. Retrograde transport of cholera toxin from the plasma membrane to the endoplasmic reticulum requires the trans-Golgi network but not the Golgi apparatus in Exo2-treated cells. EMBO Rep. 2004;5:596–601. Girard M, Poupon V, Blondeau F, McPherson PS. The DnaJ-domain protein RME-8 functions in endosomal trafficking. J Biol Chem. 2005;280:40135–40143. Moreau D, Kumar P, Wang SC, et al. Genome-wide RNAi screens identify genes required for Ricin and PE intoxications. Dev Cell. 2011;21:231–244. Smith DC, Spooner RA, Watson PD, et al. Internalized Pseudomonas exotoxin A can exploit multiple pathways to reach the endoplasmic reticulum. Traffic. 2006;7:379–393. Derby MC, Lieu ZZ, Brown D, et al. The trans-Golgi network golgin, GCC185, is required for endosome-to-Golgi transport and maintenance of Golgi structure. Traffic. 2007;8:758–773. Lieu ZZ, Derby MC, Teasdale RD, et al. The golgin GCC88 is required for efficient retrograde transport of cargo from the early endosomes to the trans-Golgi network. Mol Biol Cell. 2007;18:4979–4991. Lieu ZZ, Gleeson PA. Identification of different itineraries and retromer components for endosome-to-Golgi transport of TGN38 and Shiga toxin. Eur J Cell Biol. 2010;89:379–393. Fuchs E, Haas AK, Spooner RA, et al. Specific Rab GTPase-activating proteins define the Shiga toxin and epidermal growth factor uptake pathways. J Cell Biol. 2007;177:1133–1143. Girbes T, Ferreras JM, Arias FJ, et al. Non-toxic type 2 ribosome-inactivating proteins (RIPs) from Sambucus: occurrence, cellular and molecular activities and potential uses. Cell Mol Biol (Noisy-le-grand). 2003;49:537–545. Citores L, Munoz R, De Benito FM, et al. Differential sensitivity of HELA cells to the type 2 ribosome-inactivating proteins ebulin l, nigrin b and nigrin f as compared with ricin. Cell Mol Biol (Noisy-le-grand). 1996;42:473–476. Rapak A, Falnes PO, Olsnes S. Retrograde transport of mutant ricin to the endoplasmic reticulum with subsequent translocation to cytosol. Proc Natl Acad Sci USA. 1997;94:3783–3788.

THE INTRACELLULAR JOURNEY OF TYPE 2 RIBOSOME-INACTIVATING PROTEINS

51. 52. 53. 54.

55. 56. 57. 58.

59. 60. 61. 62. 63. 64. 65.

66. 67. 68.

69. 70. 71. 72. 73. 74. 75. 76.

95

Wesche J, Rapak A, Olsnes S. Dependence of ricin toxicity on translocation of the toxin A-chain from the endoplasmic reticulum to the cytosol. J Biol Chem. 1999;274:34443–34449. Chen A, AbuJarour RJ, Draper RK. Evidence that the transport of ricin to the cytoplasm is independent of both Rab6A and COPI. J Cell Sci.. 2003;116:3503–3510. Chen A, Hu T, Mikoryak C, Draper RK. Retrograde transport of protein toxins under conditions of COPI dysfunction. Biochim Biophys Acta. 2002;1589:124–139. Llorente A, Lauvrak SU, van Deurs B, Sandvig K. Induction of direct endosome to endoplasmic reticulum transport in Chinese hamster ovary (CHO) cells (LdlF) with a temperature-sensitive defect in epsilon-coatomer protein (epsilon-COP). J Biol Chem. 2003;278:35850–35855. Pascal JM, Day PJ, Monzingo AF, et al. 2.8-A crystal structure of a nontoxic type-II ribosome-inactivating protein, ebulin l. Proteins. 2001;43:319–326. Girod A, Storrie B, Simpson JC, et al. Evidence for a COP-I-independent transport route from the Golgi complex to the endoplasmic reticulum. Nat Cell Biol. 1999;1:423–430. White J, Johannes L, Mallard F, et al. Rab6 coordinates a novel Golgi to ER retrograde transport pathway in live cells. J Cell Biol. 1999;147:743–760. Jackson ME, Simpson JC, Girod A, et al. The KDEL retrieval system is exploited by Pseudomonas exotoxin A, but not by Shiga-like toxin-1, during retrograde transport from the Golgi complex to the endoplasmic reticulum. J Cell Sci. 1999;112 ( Pt 4):467–475. McKenzie J, Johannes L, Taguchi T, Sheff D. Passage through the Golgi is necessary for Shiga toxin B subunit to reach the endoplasmic reticulum. FEBS J. 2009;276:1581–1595. Guetzoyan LJ, Spooner RA, Boal F, et al. Fine tuning Exo2, a small molecule inhibitor of secretion and retrograde trafficking pathways in mammalian cells. Mol Biosyst. 2010;6:2030–2038. Sandvig K, Ryd M, Garred O, et al. Retrograde transport from the Golgi complex to the ER of both Shiga toxin and the nontoxic Shiga B-fragment is regulated by butyric acid and cAMP. J Cell Biol. 1994;126:53–64. Wales R, Chaddock JA, Roberts LM, Lord JM. Addition of an ER retention signal to the ricin A chain increases the cytotoxicity of the holotoxin. Exp Cell Res. 1992;203:1–4. Spooner RA, Watson PD, Marsden CJ, et al. Protein disulphide-isomerase reduces ricin to its A and B chains in the endoplasmic reticulum. Biochem J. 2004;383:285–293. Ellgaard L, Molinari M, Helenius A. Setting the standards: quality control in the secretory pathway. Science. 1999;286:1882–1888. Brodsky JL, Werner ED, Dubas ME, et al. The requirement for molecular chaperones during endoplasmic reticulum-associated protein degradation demonstrates that protein export and import are mechanistically distinct. J Biol Chem. 1999;274:3453–3460. Carvalho P, Goder V, Rapoport TA. Distinct ubiquitin-ligase complexes define convergent pathways for the degradation of ER proteins. Cell. 2006;126:361–373. Denic V, Quan EM, Weissman JS. A luminal surveillance complex that selects misfolded glycoproteins for ER-associated degradation. Cell. 2006;126:349–359. Elkabetz Y, Shapira I, Rabinovich E, Bar-Nun S. Distinct steps in dislocation of luminal endoplasmic reticulum-associated degradation substrates: roles of endoplamic reticulum-bound p97/Cdc68p and proteasome. J Biol Chem. 2004;279: 3980–3989. Lipson C, Alalouf G, Bajorek M, et al. A proteasomal ATPase contributes to dislocation of endoplasmic reticulum-associated degradation (ERAD) substrates. J Biol Chem. 2008;283:7166–7175. Brodsky JL, McCracken AA. ER protein quality control and proteasome-mediated protein degradation. Semin Cell Dev Biol. 1999;10:507–513. Hazes B, Read RJ. Accumulating evidence suggests that several AB-toxins subvert the endoplasmic reticulum-associated protein degradation pathway to enter target cells. Biochemistry. 1997;36:11051–11054. Pasetto M, Barison E, Castagna M, et al. Reductive activation of type 2 ribosome-inactivating proteins is promoted by transmembrane thioredoxin-related protein. J Biol Chem. 2012;287:7367–7373. Bellisola G, Fracasso G, Ippoliti R, et al. Reductive activation of ricin and ricin A-chain immunotoxins by protein disulfide isomerase and thioredoxin reductase. Biochem Pharmacol. 2004;67:1721–1731. Allen SC, Moore KA, Marsden CJ, et al. The isolation and characterization of temperature-dependent ricin A chain molecules in Saccharomyces cerevisiae. FEBS Journal. 2007;274:5586–5599. Li S, Spooner RA, Allen SC, et al. Folding-competent and folding-defective forms of ricin A chain have different fates after retrotranslocation from the endoplasmic reticulum. Mol Biol Cell. 2010;21:2543–2554. Slominska-Wojewodzka M, Gregers TF, Walchli S, Sandvig K. EDEM is involved in retrotranslocation of ricin from the endoplasmic reticulum to the cytosol. Mol Biol Cell. 2006;17:1664–1675.

96 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87.

88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98.

RIBOSOME-INACTIVATING PROTEINS

Spooner RA, Hart PJ, Cook JP, et al. Cytosolic chaperones influence the fate of a toxin dislocated from the endoplasmic reticulum. Proc Natl Acad Sci USA. 2008;105:17408–17413. Taylor M, Navarro-Garcia F, Huerta J, et al. Hsp90 is required for transfer of the cholera toxin A1 subunit from the endoplasmic reticulum to the cytosol. J Biol Chem. 2010;285:31261–31267. Mayerhofer PU, Cook JP, Wahlman J, et al. Ricin A chain insertion into endoplasmic reticulum membranes is triggered by a temperature increase to 37 {degrees}C. J Biol Chem. 2009;284:10232–10242. Redmann V, Oresic K, Tortorella LL, et al. Dislocation of ricin toxin A chains in human cells utilizes selective cellular factors. J Biol Chem. 2011;286:21231–21238. Yu M, Haslam DB. Shiga toxin is transported from the endoplasmic reticulum following interaction with the luminal chaperone HEDJ/ERdj3. Infect Immun. 2005;73:2524–2532. Falguieres T, Johannes L. Shiga toxin B-subunit binds to the chaperone BiP and the nucleolar protein B23. Biol Cell. 2006;98:125–134. Saleh MT, Ferguson J, Boggs JM, Gariepy J. Insertion and orientation of a synthetic peptide representing the C-terminus of the A1 domain of Shiga toxin into phospholipid membranes. Biochemistry. 1996;35:9325–9334. Menikh A, Saleh MT, Gariepy J, Boggs JM. Orientation in lipid bilayers of a synthetic peptide representing the C-terminus of the A1 domain of shiga toxin. A polarized ATR-FTIR study. Biochemistry. 1997;36:15865–15872. Li S, Spooner RA, Hampton RY, et al. Cytosolic entry of shiga-like toxin a chain from the yeast endoplasmic reticulum requires catalytically active hrd1p. PLoS One. 2012;7:e41119. Simpson JC, Roberts LM, Romisch K, et al. Ricin A chain utilises the endoplasmic reticulum-associated protein degradation pathway to enter the cytosol of yeast. FEBS Lett. 1999;459:80–84. Endo Y, Mitsui K, Motizuki M, Tsurugi K. The mechanism of action of ricin and related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification in 28 S ribosomal RNA caused by the toxins. J Biol Chem. 1987;262:5908–5912. Argent RH, Parrott AM, Day PJ, et al. Ribosome-mediated folding of partially unfolded ricin A-chain. J Biol Chem. 2000;275:9263–9269. Tam PJ, Lingwood CA. Membrane cytosolic translocation of verotoxin A1 subunit in target cells. Microbiology. 2007;153:2700–2710. Aletrari MO, McKibbin C, Williams H, et al. Eeyarestatin 1 interferes with both retrograde and anterograde intracellular trafficking pathways. PLoS One. 2011;6:e22713. Heiligenstein S, Eisfeld K, Sendzik T, et al. Retrotranslocation of a viral A/B toxin from the yeast endoplasmic reticulum is independent of ubiquitination and ERAD. Embo J. 2006;25:4717–4727. Surjit M, Jameel S, Lal SK. Cytoplasmic localization of the ORF2 protein of hepatitis E virus is dependent on its ability to undergo retrotranslocation from the endoplasmic reticulum. J Virol. 2007;81:3339–3345. Duriez M, Rossignol JM, Sitterlin D. The hepatitis B virus precore protein is retrotransported from endoplasmic reticulum (ER) to cytosol through the ER-associated degradation pathway. J Biol Chem. 2008;283:32352–32360. Afshar N, Black BE, Paschal BM. Retrotranslocation of the chaperone calreticulin from the endoplasmic reticulum lumen to the cytosol. Mol Cell Biol. 2005;25:8844–8853. Giodini A, Cresswell P. Hsp90-mediated cytosolic refolding of exogenous proteins internalized by dendritic cells. Embo J. 2008;27:201–211. Rutenber E, Katzin BJ, Ernst S, et al. Crystallographic refinement of ricin to 2.5 A. Proteins. 1991;10:240–250. Tahirov TH, Lu TH, Liaw YC, et al. Crystal structure of abrin-a at 2.14 A. J Mol Biol. 1995;250:354–367. O’Neal CJ, Amaya EI, Jobling MG, et al. Crystal structures of an intrinsically active cholera toxin mutant yield insight into the toxin activation mechanism. Biochemistry. 2004;43:3772–3782.

7

Shiga Toxins: The Ribosome-inactivating Proteins from Pathogenic Bacteria Maurizio Brigotti Dipartimento di Medicina Specialistica, Università di Bologna, Italy

Introduction

Adhesiveness, invasiveness, and delivery of toxins are considered to be the main weapons of the bacteria responsible for diseases in humans and animals. The former virulence factors allow pathogenic bacteria to interact with eukaryotic cells, to multiply and to penetrate host tissues, whereas the production of toxins induces specific effects on target cells (exotoxins) and broad or systemic effects in the host (endotoxins). The interplay between these virulence factors is of prime importance in the development of infectious diseases. Stxs (Shiga toxins) are produced by various bacteria including Shigella dysenteriae type 1, which is responsible for bacillary dysentery in humans, and by a restricted subset of Escherichia coli strains, hence called STEC (Shiga toxin-producing E. coli), which have a causative role in the pathogenesis of hemorrhagic colitis and HUS (hemolytic uremic syndrome) in children.1 Interestingly, these bacterial toxins and the RIPs (ribosome-inactivating proteins) from plants were found to share the same enzymatic mechanism of action on ribosomes (see below) and hence Stxs are considered the bacterial branch of this large family of enzymes. Recently, another bacterial RIP from the soil bacterium Streptomyces coelicolor has been described.2 This protein, however, differs from Stxs since it is not toxic to intact cells and is apparently not related to human diseases. Purification of Shiga Toxins

The involvement of Stxs in the pathogenic processes of the above-mentioned human diseases was established after the purification to homogeneity of the cytotoxin produced by Shigella. This was accomplished by Olsnes and colleagues,3, 4 starting from bacterial culture media or bacterial lysates, through a laborious purification procedure based on repeated chromatographic steps, sucrose gradient centrifugation, and non-denaturating polyacrylamide gel electrophoresis. The addition of carrier proteins (rabbit hemoglobin, bovine serum albumin) and radioiodination of the partially purified toxin were exploited to avoid loss of toxin activity and to facilitate detection. Since then, many different methods have been devised enabling isolation of higher amounts of purified Stxs and preserving their cytotoxic and enzymatic activities. To date, the most simple and useful methods are Ribosome-inactivating Proteins: Ricin and Related Proteins, First Edition. Edited by Fiorenzo Stirpe and Douglas A. Lappi. © 2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.

97

98

RIBOSOME-INACTIVATING PROTEINS

those based on the capture of the toxins present in crude bacterial lysates or in culture media with the immobilized specific receptor. Typical examples are the affinity chromatographies with globotriose-fractogel,5 P1 glycoprotein,6, 7 and Synsorb pK.8 These methods are not time-consuming and induce only minimal perturbation of the toxic sample. This preserves the cytotoxic activity (binding specificity for target cells and intracellular enzymatic activity) and also the known ability of Stxs to be recognized by PMN (polymorphonuclear leukocytes) in the blood stream of the infected hosts. The latter interaction is probably a crucial step in the pathogenesis of HUS and/or in the innate response of the host, as recently reviewed.9 Proper folding is required for Stxs to be recognized by PMN and it has been suggested that single-step purification procedures would yield Stxs with the proper conformation, whereas more complicated multi-step methods may induce partial unfolding of Stxs and loss of binding to PMN.9, 10 It is worth noting that toxicity is preserved in partially unfolded Stxs, that only lose the binding activity for PMN.10

Structure and Mechanism of Action of Shiga Toxins The AB5 Structure

Suitable and functional molecular models have often been re-proposed during evolution, as in the structure of some bipartite bacterial exotoxins. The classical AB5 model comprises an enzymatic A chain non-covalently bound to a pentamer of B chains which bind cell receptors. Cholera toxin, E. coli heat-labile toxin, and all the Stxs variants share this structure, although the binding specificity as well as the type of enzymatic activity greatly differ among them. The Stxs family comprises several toxic molecules: Stx (Shiga toxin) is the single prototype produced by Shigella, whereas numerous cytotoxins are elaborated by STEC, even though two variants prevail: Stx1 (Shiga toxin 1) and Stx2 (Shiga toxin 2).1 The former is immunologically indistinguishable from Stx since it differs by a single amino acid residue located in the A chain (Thr45 in Stx, Ser45 in Stx1).11 Conversely, Stx2 is not neutralized by antibodies to Stx since it has less than 60% similarity with the prototype toxin, both at protein and at gene levels.12 Stx1 and Stx2 are phage-encoded and require induction for full expression,13, 14 whereas Stx is encoded by genes located on the Shigella chromosome.11 Crystallographic studies have detailed the structure of Stxs variants and the reciprocal interactions between A (32 kDa) and B chains (7.7 KDa; Figure 7.1).15 The B chains, each folded in a single α-helix, are arranged in a pentamer which form a ring delimiting an opening. In this space, a single α-helix belonging to the C-terminus moiety of the A chain deeply interacts with the five B subunit α-helices forming a non-covalent bridge between the two different chains of the holotoxin. The structure is reinforced with antiparallel β-sheets formed by strands belonging to each B subunit that, on the outside of the B pentamer, encircle the A/B contact regions (Figure 7.1). It is of note that the arrangement and the fold of the B-chain pentamer of Stxs are similar to those of the B subunits of the E. coli heat-labile toxin. Stx A chain harbors the active site in a cleft closely resembling the one in the A subunit of ricin, the well-known plant RIP from the seeds of Ricinus communis.15, 16 The folding of ricin A chain and of a portion of Stx A chain (called A1 fragment, see below) are very similar, since they possess about 150 structurally equivalent amino acid residues and more than 20% of them are identical.15, 16 More importantly, there are several invariant amino acid residues present in the active sites of the two toxins (Tyr-77, Val-78, Ser-112, Tyr-114, Glu-167, Ala-168, Arg-170, and Trp-203; numbers as in Stx) and this represents a further evolutionary convergence between molecules belonging to quite different biological kingdoms.15, 16 A comprehensive review of the structural behaviors of RIPs is given in Chapter 8.

SHIGA TOXINS: THE RIBOSOME-INACTIVATING PROTEINS FROM PATHOGENIC BACTERIA

99

A chain

B chains Figure 7.1 Detailed structure of Stx1 depicted as ribbon diagram. Reproduced with permission15 with modifications. The fracture between A1 and A2 is due to proteolysis.

Intracellular Activation

Both Stx1 and Stx2 share the requirement of intracellular activation to fully express the toxic activity. The reduction of a single disulfide bond connecting A and B subunits in ricin allows the A chain to reach the cellular targets.17 In the case of Stxs the A/B chain interactions are non-covalent and the disulfide bond in fact connects two fragments of the A subunit arising from the intracellular proteolytic activity (Figure 7.2). In eukaryotic cells, during endocytosis, the membrane-anchored protease furin cleaves the A chain of Stxs18 forming two different fragments: A1 (~28 kDa) endowed with the enzymatic activity and A2 (~4 kDa) interacting with the B chains (Figure 7.2). The reduction of the disulfide bond is mandatory for the disengagement of the active site-containing A1 moiety and, in the case of Stx1, also for boosting the enzymatic activity. In fact, analysis of the kinetic constants of the enzymatic reaction catalyzed on ribosomes (see below) revealed similar substrate affinities for holotoxin, isolated A chain, and A1 fragment (Km ~ 1 μM), whereas the Kcat of the A1 fragment was 100 and 1000-fold higher than those of the A chain and holotoxin, respectively.19

Ribosomes and DNA as Intracellular Targets

Stxs have been classified as RNA N-glycosidases acting on eukaryotic ribosomes, even though the term glycosylases was proposed to indicate their action on nuclear DNA (Figure 7.3). Both the enzymatic activities involve the cleavage of a glycosidic bond connecting the base adenine to the sugar of the corresponding nucleic acid. Stxs express their activity on 28S rRNA in the large

100

RIBOSOME-INACTIVATING PROTEINS

B

B

A2

A1

B

B

Figure 7.2

S–S

B

Schematic structure of the A/B interactions in Stxs. (This figure also appears in the color plate section.)

Adenine

Stxs

X 28S rRNA

Adenine Adenine

Adenine X

X

X

Adenine

Stxs

X

DNA Figure 7.3 section.)

Enzymatic activity of Stxs on ribosomal RNA and nuclear DNA. (This figure also appears in the color plate

ribosomal subunit, specifically removing a single adenine residue20 located in a conserved stemloop region21 recognized by elongation factors during protein synthesis.22 The ribosomal functional impairment is thus related to the inhibition of the binding of elongation factors to ribosomes and the consequence is the irreversible arrest of translation in cells. The specific injury and the particular functional impairment induced by ricin and Stxs acting on ribosomes are indistinguishable, as the plant and bacterial toxins share the same mechanism of action. A more detailed description of this topic is presented in Chapter 2. Shortly after the discovery of the mechanism of action on ribosomes, a method for the detection of the adenine released by the enzymatic activity of these toxins was devised. The base can be readily detected by HPLC (high pressure liquid chromatography) after conversion into its fluorescent etheno-derivative.23 By this method, the glycosylase activity of plant ribosome-inactivating proteins24, 25 and Stxs26 on isolated DNA in vitro was demonstrated. Although these results provided an important contribution to the field, the effect on nuclear DNA in whole cells and the relationship with the ribotoxic activity were only pinpointed several years later. The time-course of the enzymatic activities of Stx1, Stx2, and ricin acting on ribosomes and DNA was measured in human endothelial cells.27, 28 The two bacterial toxin variants showed overlapping time-courses of ribosome inactivation, while the ribotoxic activity of ricin was shown to occur later, probably reflecting different modes of internalization. The glycosylase activity on DNA was expressed at the same time (ricin) or 3 hours after (Stx1, Stx2) the inhibition of translation, even though the damage to nuclear DNA was not secondary to the inhibition of protein synthesis, nor to apoptosis.27, 28 In fact, nuclear

SHIGA TOXINS: THE RIBOSOME-INACTIVATING PROTEINS FROM PATHOGENIC BACTERIA

101

DNA damage peaked several hours before the onset of the earliest markers of apoptosis in intoxicated cells,27, 28 thus indicating that the ordered apoptotic DNA digestion with formation of discrete fragments is clearly distinguishable from the primary and direct DNA damage induced by the toxins. It should be noted that nuclear DNA damage was quantified by a sensitive and simple method, the “halo assay,” performed in alkaline conditions, which allows the conversion of DNA apurinic sites in single strand DNA breaks to be detected by microscopic observation.29, 30 The glycosylase activity of Stxs appears rather broad in specificity as these enzymes are capable of removing multiple adenines from DNA in vitro26, 31 and in whole cells.27 Their activity is very similar to that of DNA repair enzymes that preserve DNA homeostasis in the nucleus by removing inappropriate, mismatched, or damaged bases. There are few significant examples of DNA mismatch repair enzymes, either in bacteria (MutY) or in man (MYH), which specifically recognize the normal base adenine as well as Stxs do. By comparing the kinetics of Stx1 acting on DNA with those of known DNA glycosylases acting on adenine in mismatch, it is noteworthy that the specificity constant (Kcat/Km) of Stx1 (4.2 μM-1/min-1)32 fits within the range of values (0.5–39 μM-1/min-1) obtained with MutY and MYH,33, 34 showing that the bacterial toxin is as efficient in removing adenine from DNA as other known DNA glycosylases. However, the consequences of these enzymatic actions are quite dissimilar, since Stxs would induce toxic, mutagenic, and cancerogenic effects rather than DNA repair, as previously demonstrated in the case of plant RIP.35 The situation is complicated by the ability of Stxs to inhibit the repair of oxidative and alkylative DNA lesions via a mechanism involving impairment of BER (base excision repair).36 It has been shown that some plant RIPs35 remove adenines from poly(ADP-ribose), a polymer produced by the enzymatic activity of PARP (poly(ADP-ribose) polymerase) and involved in signaling DNA damage and in orchestrating DNA repair.37, 38 As observed in vitro,35 an extensive depurination might promote early degradation of poly(ADP-ribose) polymer and, if this happens in the nucleus of intoxicated cells, the PARP-dependent nuclear DNA repair processes could be affected. These authors also demonstrated the transforming activity of plant RIP for 3 T3 fibroblasts.35 Thus, the genotoxicity of Stxs on mammalian cells would be the result of direct (DNA damaging activity) and indirect effects (DNA repair inhibition), being also influenced by the presence of other DNA targeting species.

Binding Properties of Shiga Toxins

The binding properties of the two main Stxs variants are rather specific and involve a multivalent interaction of their B chains to neutral glycosphingolipids, mainly globotriaosylceramide (Gb3Cer) and to a lesser extent to globotetraosylceramide (Gb4Cer). Gb3Cer is expressed on the membrane of few human cells, that is endothelial cells lining the microvasculature of the kidney, brain, and intestine; mesangial cells; glomerular epithelial cells; and tubular epithelial cells in the kidney.39–42 The receptor distribution explains the involvement of the gut in hemorrhagic colitis and of kidney and brain in HUS, as consequences of human infections by the widely diffused STEC. Stxs bind to the membrane of sensitive cells and are then endocytosed by different pathways involving lipid rafts, the latter defined as membrane microdomains in which cholesterol, glycosphingolipids, sphingomyelin, dipalmitoylphosphatidylcholine, and several proteins are assembled.43 To reach their intracellular targets, Stxs present in endosomes must be retrogradely transported through the Golgi apparatus to the endoplasmic reticulum where the active A1 fragment is translocated to the cytosol and to the nucleus.44–48 Although Gb3cer is required as the Stxs receptor, glycosphingolipids in general are important for the transport of Stxs from endosomes to Golgi once the toxin has bound the specific receptor.49 Moreover, in vertebrates, glycosphingolipids are composed of a sphingosine

102

RIBOSOME-INACTIVATING PROTEINS

core linked to a fatty acid which differs in chain length (mostly C16 to C24) and degree of desaturation.43 This heterogeneity in GB3Cer lipoforms might condition the intracellular transport and destination of Stxs, since different fatty chains would mediate different interactions with other membrane components in lipid drafts.50, 51 Chapter 2 of this book compares and contrasts ricin and Stxs, giving the molecular details of trafficking in eukaryotic cells.

Role of Shiga Toxins in the Pathogenesis of HUS

HUS is the life-threatening sequela of intestinal infections caused by STEC and, in the Western world, is the leading cause of acute renal failure in early childhood.52–54 Typical STEC strains, such as O157:H7, live in the gut of ruminants, particularly cattle, without provoking any symptoms in animals.55 Therefore, hemorrhagic colitis and HUS should be considered zoonoses, with vegetables or undercooked bovine meat contaminated with fecal bovine specimens being the main vehicles of the diseases.52–54

Pathogenesis of HUS

The hallmark of HUS is the presence of thrombotic microvascular lesions confined to few organs and in particular to the renal glomeruli, the gastrointestinal tract, and the brain. Microvascular endothelial injury is considered the most important pathogenic event in HUS, and the targeted endothelial cell subtypes are known to express Gb3Cer receptors for Stxs.1 It is recognized that the toxins have a causative role in HUS and this is consistent with data obtained in patients and in vitro at both cellular and molecular levels.1, 53 The histopathology of HUS in humans has been investigated in detail, and the main features observed in renal glomeruli include changes in capillary wall thickness, swelling and detachment of endothelial cells from the basement membrane, and deposition of fibrin and platelets with congested rather than ischemic glomeruli.56, 57 This explains the acute renal failure, the first component of the triad characterizing HUS, closely related to the presence of microthrombi, which compromise blood supply by narrowing or occluding the capillary lumen.58, 59 In HUS, platelets are found activated and massively engaged in microthrombi, as well as aggregated in blood and thus removed by the reticuloendothelial system. This large consumption of platelets is the cause of the severe thrombocytopenia seen in HUS patients, the second component of the triad.58, 59 HUS patients complete the triad by developing hemolytic anemia, whose main features are the formation of fragmented erythrocytes, often with shapes resembling helmets, that are removed from the blood by the reticuloendothelial system. The lesions to red blood cells are mechanically induced, since these cells are forced to pass through the partially occluded capillaries in renal glomeruli.58, 59 In keeping with the notion that Stxs are the main actors in the development of HUS, many investigations have been carried out to shed more light on the mechanisms underlying the pathophysiology of the syndrome on the basis of the toxin actions on targeted organs and cells. After the consumption of food contaminated by STEC and a short incubation time (about 3 days) patients develop watery diarrhea with cramping abdominal pain (Figure  7.4).1, 53 Interestingly, this non-specific symptom is not related to the action of Stxs, but rather to the particular and intimate mode of adhesion of these E. coli strains to the epithelial cells of the gut. STEC interacting with human intestinal mucosa are capable of translocating various bacterial

SHIGA TOXINS: THE RIBOSOME-INACTIVATING PROTEINS FROM PATHOGENIC BACTERIA

Days

103

Stxs involvement + main role

HUS

10

Bloody diarrhea

5

+ definite role

Watery diarrhea abdominal pain

3

– no role

STEC ingestion

0

Figure 7.4 Time-course of STEC infections in humans and role of Stxs in the related diseases. HUS (hemolytic uremic syndrome); STEC (Shiga toxin-producing E. coli).

effectors into enterocytes through an injection device called the Type III secretion system.60 This leads to intimate adhesion to the gut, and the deriving, attaching and effacing lesions cause profound modifications and perturbations in epithelial cells that lose brush borders and accumulate actin-derived pedestals beneath the site of membrane contact.1, 59 The derangement of the intestinal lining alters the mechanisms of absorption of the bowel, thus explaining the watery diarrhea. In a short time (1–2 days), a large proportion of infected patients may develop bloody diarrhea that is the common prodromal symptom in most HUS cases (Figure 7.4).1, 53 In this case the pathogenetic process involves Stxs: these are produced by the non-invasive bacteria and translocate through the gut epithelium reaching the lamina propria, although human enterocytes apparently do not possess specific Stxs receptors.61 Passage of the toxins through epithelial tight junctions loosened by the adherence of STEC and/or by the opposite passage of PMN in the inflamed bowel have been observed.62–64 Recently, a new Gb3Cer-independent mechanism explaining transcellular transcytosis and centered on the actin-driven macropynocitotic activity of human enterocytes has been described.65 Interestingly, Stxs uptake by these cells also stimulates apical secretion and depletion of galectin-3, impairing the functions of transporters and structural brush border proteins, thus contributing to the onset of diarrhea.66 Once in the gut lamina propria and absorbed in the circulation, Stxs find Gb3Cer on intestinal endothelial cells, which in turn become intoxicated.61, 62 Mucosal and submucosal edema, hemorrhage with focal areas of necrosis, and thrombotic microangiopathy are the histopathological changes observed in the bowel of patients with hemorrhagic colitis and are related to the action of toxins on gut endothelia.56, 67 After about a week from the onset of bloody diarrhea, HUS develops in a low proportion (~15%) of patients (Figure 7.4).

Linking Shiga Toxins to the Pathogenesis of HUS

During the journey from the gut to the kidney and the brain, Stxs may encounter several blood components. These interactions are believed to be important in the pathogenesis of HUS for two different reasons. On the one hand, Stxs may exploit binding to blood cells to be shuttled into the

104

RIBOSOME-INACTIVATING PROTEINS

ultimate target cells. On the other hand, the challenge to circulating cells may induce responses related to the subsequent pathogenetic steps, such as the intoxication of endothelial cells or the prothrombotic state observed in HUS. Indeed, Stxs have been found to bind in vitro to red cells,68 monocytes,69 platelets,70–72 and PMN.73–75 While the interaction with the former cells is related to the presence on their membrane of Gb3Cer lipoforms similar to those on target endothelial cells, PMN interact with Stxs through a low affinity unknown receptor.73 The transfer of Stxs from PMN to the high affinity Gb3Cer receptors of human endothelial cells has been demonstrated in vitro.73, 76 Moreover, PMN-bound Stxs have been detected in patients with HUS,77, 78 hence the role of these cells in transporting the toxins in blood has been envisaged, although is as yet unproven, in HUS patients. Strikingly, the A chain of ricin and two different single-chain RIPs interact with the same PMN receptor, indicating a specific role for these cells in the recognition of foreign toxins.79 Monocytes triggered by Stxs showed up-regulation and secretion of IL-1 and TNF-α,69 two potent inflammatory mediators that, in turn, enhance Gb3Cer expression on endothelial cells.80, 81 This would render renal endothelial lining more susceptible to toxin action, and so HUS is probably fomenting in the kidney during the prodromal gastrointestinal phase. Platelet activation and aggregation induced by Stxs have been observed.72 as well as binding of Stxs to platelets activated by other stimuli.71 In both cases, these interactions appear to be involved in the development of thrombocytopenia (massive engagement in damaged body sites or removal of platelet aggregates by the reticuloendothelial system) and in microthrombi formation. Recently, however, Stxs have also been found on blood cell complexes.82, 83 This is particularly intriguing, since two STEC virulence factors added to human blood, such as LPS from the widely diffused O157:H7 STEC strain and Stx2, induce the formation of platelet–monocyte or platelet–neutrophil aggregates containing activated thrombocytes and leukocytes.83 Moreover, platelet- or monocyte-derived microparticles are generated upon these treatments and, interestingly, the combined (Stx2, O157LPS) challenge leads to higher amounts of microparticles expressing tissue factor, a key activator involved in fibrin generation and thus related to thrombogenesis.83 Indeed, HUS patients show higher platelet–leukocyte aggregates in blood and increased levels of plasmatic and aggregate-borne tissue factor than do healthy children.83 Stx2 has been shown to be present on these circulating aggregates and this further highlights the direct involvement of these bacterial toxins in the prothrombotic states observed during the natural course of HUS.83 Stxs also have a clear-cut role in directly activating the complement system via the alternative pathway and in delaying the protecting action of factor H for host cells.84 This was confirmed in HUS patients who have circulating platelets, leukocytes, and microparticles bearing activated complement components on their surfaces.82 The same study also confirmed the involvement of Stxs and O157LPS in this phenomenon.82 This is in keeping with old and recent unexplained evidence on the activation of the complement system in HUS patients.85, 86 Complement is a key weapon of innate immunity, deeply related to the inflammation response, thus activation of complement by Stxs may well induce direct destruction of kidney tissues and indirect inflammatory injuries. It should be noted that a small number of sporadic or familial HUS cases are caused by known genetically-driven defects (atypical HUS) of key complement regulators (loss of function mutations) or of C3 convertase components (gain of function mutations).87 The former mutations reduce the defenses of endothelial cells against complement attacks, whereas the second mutations result in hyperactive C3 convertase, the step at which the alternative or classical complement pathways converge.87 A novel therapy for atypical HUS was developed by the administration of a monoclonal

SHIGA TOXINS: THE RIBOSOME-INACTIVATING PROTEINS FROM PATHOGENIC BACTERIA

105

antibody to C5 (eculizumab) preventing the generation of the pro-inflammatory peptide C5a and the formation of the membrane-attack complex.88, 89 Since the involvement of complement might represent a common thread between atypical HUS and typical STEC-induced HUS, treatment with eculizumab was proposed in a recent German outbreak (see below).90 Nevertheless, the fact that HUS develops in a relatively low proportion of STEC-infected children suggests that genetic or acquired factors concur with STEC virulence factors in triggering HUS. One can hypothesize that particular polymorphisms of normal genes regulating complement rather than genetic defects are involved in typical STEC-induced HUS. However, it should be borne in mind that the pathogenesis of typical HUS is centered on the interaction between Stxs and endothelial cells endowed with the specific toxin receptor. Therefore, explanations of what happens during typical HUS on the basis of non-endothelial cell responses to Stxs or LPS also need to take account of the actions of the toxin on the endothelial lining of target organs. Stxs act on sensitive human endothelial cells triggering an array of responses: intoxicated cells might undergo apoptosis, produce pro-inflammatory cytokines, or up-regulate their adhesion molecules. Microvascular endothelial cells treated with Stxs express P selectin, an adhesion molecule implicated in platelet deposition, also being a specific high affinity ligand for C3.91 Moreover, Stxs promote C3 activation that results in C3a generation and the related pro-inflammatory response.91 Other findings have revealed a more complicated inflammatory scenario, since endothelial cells challenged with Stxs up-regulate the expression of several pro-inflammatory cytokine messengers and of the corresponding proteins. This has been demonstrated in different laboratories as recently reviewed9 and, in particular, in a straightforward study with human endothelial cells in which microarray analysis of thousands of genes showed the up-regulation of about 25 pro-inflammatory messengers after treatment with both Stx1 and Stx2.92 The upregulating events were also demonstrated at the protein level, and this is not obvious since Stxs are strong inhibitors of translation.92 How can the toxic effect induced by Stxs be related to these up-regulating effects? Endothelial cells after intoxication even by small concentrations of Stxs suffer from two different stresses: ribotoxic stress and ER (endoplasmic reticulum) stress. Ribotoxic stress is a direct consequence of the depurinating activity of 28S RNA by Stxs, as well as by other RIPs such as ricin.93 On the other hand, ER-stress is probably initiated by the production of misfolded or incomplete proteins after toxin action on the protein synthesis machinery.94 The former stress leads to the activation of MAPK (mitogen-activated protein kinases) cascades culminating in the transcription of the above mentioned pro-inflammatory cytokine messengers through activation of proper transcription factors (NF-kB and AP-1).9, 94 The same signaling pathway might also be important in regulating apoptosis of intoxicated cells as does a prolonged ER-stress.94 Indeed, endothelial cells might choose to participate in their own demise when multiple signaling pathways, such as ribotoxic stress, ER-stress, DNA damage, interaction with Gb3Cer (B chain), or with Bcl-2 (Stx1 A chain),28, 48, 94–96 converge in the induction of apoptosis. Thus, Stxs contribute to the development of HUS by eliciting a plethora of effects on endothelial and non-endothelial cells that can be categorized as: (i) pro-apoptotic effects, (ii) pro-thrombotic effects, and (iii) pro-inflammatory effects. Stxs seem to be a multi-faceted agonist stimulating eukaryotic cells to devise multiple concurrent and interacting responses involved in the pathogenesis of this syndrome. Estimating the relative importance of the different Stxs-induced pathways, their relationships, and the precise hierarchy would help further understanding of the pathogenesis of HUS and, in particular, the obscure mechanisms underlying the transition between hemorrhagic colitis and HUS. The relevance of Stxs in human pathology was exemplified dramatically during the 2011 German outbreak discussed below.

106

RIBOSOME-INACTIVATING PROTEINS

The 2011 German STEC Outbreak

In May–July 2011, a public health issue arose from the sudden appearance of a large STEC outbreak that caused a remarkable number of deaths among adults in Germany and other European countries. The unusual STEC strain responsible for the outbreak (E. coli O104:H4) was improperly re-named by the media as a killer bacterium, leading to widespread alarm in the population. Although STEC has been studied for many decades by scientists, this strain displayed an unexpectedly high level of virulence, likely related to an unusual combination of virulence determinants. The German strain produced Stx2 and exhibited a particular pattern of adherence to human intestinal epithelial cells, called enteroaggregative adhesion.97, 98 This pattern differed considerably from that caused by the interaction of the known STEC strains with human intestine. A similar combination was found in the 1990s in a different E. coli strain (O111:H2) that caused a more confined HUS outbreak in French children.99, 100 In the third week of May 2011, the Robert Koch Institute in Germany noted an abnormally high proportion of patients with HUS and bloody diarrhea caused by STEC.98 Since then, and during the fully evolved outbreak, many attempts have been undertaken by the German Public Health Authorities to identify the source and vehicle of the infection. Initially, cucumbers were incriminated and were consequently banned in several European countries with considerable economic consequences. The unusual combination of virulence factors in the etiologic agent of the outbreak98 was identified in Denmark by Flemming Scheutz (World Health Organization, Collaborating Centre for Reference and Research on Escherichia and Klebsiella, Copenhagen) at the end of May 2011 in a traveling German patient.97 A few days later, the European Union Reference Laboratory for E. coli, directed by Alfredo Caprioli (Istituto Superiore di Sanità, Rome, Italy), developed a novel real-time PCR procedure for the detection of the German strain in food.97 The new test was shared with the National Reference Laboratories of the different European Countries for food testing. The actions of the European Public Health Authorities appeared to be rapid, efficient, interactive, and integrated. By virtue of this synergism and of more carefully conducted analysis, a correlation between the consumption of fenugreek sprouts and this food-borne STEC outbreak emerged, hence exonerating cucumbers.101 Overall the German outbreak caused about 4000 STEC-induced cases of diarrhea and 22% of these patients developed HUS, with an unusually high incidence, as well as an unexpectedly high mortality (about 50 cases).102 Moreover, close to 50% of HUS patients developed neurologic complications and needed dialysis103 and, most importantly, 88% of the patients were adult (median 42 years) and 68% of them were female.104 Since HUS is the main cause of acute renal failure in early childhood, the epidemic profile of the German outbreak appears idiosyncratic.104 Horizontal genetic exchange with acquisition of the prophage encoding Stx2 by an enteroaggregative E. coli O104:H4 strain, otherwise able to cause only diarrhea, was found to be the mechanism allowing the emergence of this exceptionally virulent strain. 102 The features of E. coli O104:H4 stunned the experts, since enteroaggregative E. coli are a well known group of human pathogens. Actually, enteroaggregative E. coli are a frequent cause of persistent watery diarrhea in infants and children in developing countries105 and have been involved in rare outbreaks in adults,97 always without inducing renal involvement. The lesson to scientists and Public Health Authorities arising from this severe outbreak is two-fold: first, the extreme plasticity of E. coli genome can allow other less virulent strains to acquire the genes encoding for Stxs; second, Stxs emerge as pivotal toxins in human pathology. In this light, to reduce the threat in the near future, it will be of paramount importance to set up international prevention programs aimed at rapidly identifying new emerging pathogenic bacteria producing Stxs.

SHIGA TOXINS: THE RIBOSOME-INACTIVATING PROTEINS FROM PATHOGENIC BACTERIA

107

REFERENCES 1. Paton JC, Paton AW. Pathogenesis and diagnosis of Shiga toxin-producing Escherichia coli infections. Clin Microbiol Rev. 1998;11:450–479. 2. Reyes AG, Geukens N, Gutschoven P, et al. The Streptomyces coelicolor genome encodes a type I ribosome-inactivating protein. Microbiology. 2010;156:3021–3030. 3. Olsnes S, Eiklid K. Isolation and characterization of Shigella shigae cytotoxin. J Biol Chem. 1980;255:284–289. 4. Olsnes S, Reisbig R, Eiklid K. Subunit structure of Shigella cytotoxin. J Biol Chem. 1981;256:8732–8738. 5. Ryd M, Alfredsson H, Blomberg L, et al. Purification of Shiga toxin by alpha-D-galactose-(1----4)-beta-D-galactose-(1---4)-beta-D-glucose-(1- ---) receptor ligand-based chromatography. FEBS Lett. 1989;258:320–322. 6. Acheson DW, Jacewicz M, Kane AV, et al. One step high yield affinity purification of shiga-like toxin II variants and quantitation using enzyme linked immunosorbent assays. Microb Pathog. 1993;14:57–66. 7. Donohue-Rolfe A, Acheson DW, Kane AV, Keusch GT. Purification of Shiga toxin and Shiga-like toxins I and II by receptor analog affinity chromatography with immobilized P1 glycoprotein and production of cross-reactive monoclonal antibodies. Infect Immun. 1989;57:3888–3893. 8. Mulvey G, Vanmaele R, Mrazek M, et al. Affinity purification of Shiga-like toxin I and Shiga-like toxin II. J Microbiol Meth. 1998;32:247–252. 9. Brigotti M. The Interactions of Human Neutrophils with Shiga Toxins and Related Plant Toxins: Danger or Safety? Toxins. 2012;4:157–190. 10. Brigotti M, Carnicelli D, Arfilli V, et al. Change in conformation with reduction of alpha-helix content causes loss of neutrophil binding activity in fully cytotoxic Shiga toxin 1. J Biol Chem. 2011;286:34514–34521. 11. Strockbine NA, Jackson MP, Sung LM, et al. Cloning and sequencing of the genes for Shiga toxin from Shigella dysenteriae type 1. J Bacteriol. 1988;170:1116–1122. 12. Jackson MP, Newland JW, Holmes RK, O’Brien AD. Nucleotide sequence analysis of the structural genes for Shiga-like toxin I encoded by bacteriophage 933 J from Escherichia coli. Microb Pathog. 1987;2:147–153. 13. Scotland SM, Smith HR, Willshaw GA, Rowe B. Vero cytotoxin production in strain of Escherichia coli is determined by genes carried on bacteriophage. Lancet. 1983;2:216. 14. O’Brien AD, Newland JW, Miller SF, et al. Shiga-like toxin-converting phages from Escherichia coli strains that cause hemorrhagic colitis or infantile diarrhea. Science. 1984;226:694–696. 15. Fraser ME, Chernaia MM, Kozlov YV, James MN. Crystal structure of the holotoxin from Shigella dysenteriae at 2.5 A resolution. Nat Struct Biol. 1994;1:59–64. 16. Katzin BJ, Collins EJ, Robertus JD. Structure of ricin A-chain at 2.5 A. Proteins. 1991;10:251–259. 17. Olsnes S, Pihl A. Different biological properties of the two constituent peptide chains of ricin, a toxic protein inhibiting protein synthesis. Biochemistry. 1973;12:3121–3126. 18. Garred O, van Deurs B, Sandvig K. Furin-induced cleavage and activation of Shiga toxin. J Biol Chem. 1995;270:10817–10821. 19. Brigotti M, Carnicelli D, Alvergna P, et al. The RNA-N-glycosidase activity of Shiga-like toxin I: kinetic parameters of the native and activated toxin. Toxicon. 1997;35:1431–1437. 20. Endo Y, Tsurugi K, Yutsudo T, et al. Site of action of a Vero toxin (VT2) from Escherichia coli O157:H7 and of Shiga toxin on eukaryotic ribosomes. RNA N-glycosidase activity of the toxins. Eur J Biochem. 1988;171:45–50. 21. Chan YL, Endo Y, Wool IG. The sequence of the nucleotides at the alpha-sarcin cleavage site in rat 28 S ribosomal ribonucleic acid. J Biol Chem. 1983;258:12768–12770. 22. Moazed D, Robertson JM, Noller HF. Interaction of elongation factors EF-G and EF-Tu with a conserved loop in 23S RNA. Nature. 1988;334:362–364. 23. Zamboni M, Brigotti M, Rambelli F, et al. High-pressure-liquid-chromatographic and fluorimetric methods for the determination of adenine released from ribosomes by ricin and gelonin. Biochem J. 1989;259:639–643. 24. Barbieri L, Gorini P, Valbonesi P, et al. Unexpected activity of saporins. Nature. 1994;372:624. 25. Barbieri L, Valbonesi P, Bonora E, et al. Polynucleotide:adenosine glycosidase activity of ribosome-inactivating proteins: effect on DNA, RNA and poly(A). Nucleic Acids Res. 1997;25:518–522. 26. Barbieri L, Valbonesi P, Brigotti M, et al. Shiga-like toxin I is a polynucleotide:adenosine glycosidase. Mol Microbiol. 1998;29:661–662. 27. Brigotti M, Alfieri R, Sestili P, et al. Damage to nuclear DNA induced by Shiga toxin 1 and ricin in human endothelial cells. FASEB J. 2002;16:365–372. 28. Brigotti M, Carnicelli D, Ravanelli E, et al. Molecular damage and induction of proinflammatory cytokines in human endothelial cells exposed to Shiga toxin 1, Shiga toxin 2, and alpha-sarcin. Infect Immun. 2007;75:2201–2207.

108 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

RIBOSOME-INACTIVATING PROTEINS

Sestili P. The fast-halo assay for the assessment of DNA damage at the single-cell level. Methods Mol Biol. 2009;521:517–533. Sestili P, Cantoni O. Osmotically driven radial diffusion of single-stranded DNA fragments on an agarose bed as a convenient measure of DNA strand scission. Free Radic Biol Med. 1999;26:1019–1026. Brigotti M, Barbieri L, Valbonesi P, et al. A rapid and sensitive method to measure the enzymatic activity of ribosomeinactivating proteins. Nucleic Acids Res. 1998;26:4306–4307. Brigotti M, Carnicelli D, Accorsi P, et al. 4-Aminopyrazolo[3,4-d]pyrimidine (4-APP) as a novel inhibitor of the RNA and DNA depurination induced by Shiga toxin 1. Nucleic Acids Res. 2000;28:2383–2388. Manuel RC, Lloyd RS. Cloning, overexpression, and biochemical characterization of the catalytic domain of MutY. Biochemistry. 1997;36:11140–11152. Shinmura K, Yamaguchi S, Saitoh T, et al. Adenine excisional repair function of MYH protein on the adenine:8-hydroxyguanine base pair in double-stranded DNA. Nucleic Acids Res. 2000;28:4912–4918. Barbieri L, Brigotti M, Perocco P, et al. Ribosome-inactivating proteins depurinate poly(ADP-ribosyl)ated poly(ADP-ribose) polymerase and have transforming activity for 3 T3 fibroblasts. FEBS Lett. 2003;538:178–182. Sestili P, Alfieri R, Carnicelli D, et al. Shiga toxin 1 and ricin inhibit the repair of H2O2-induced DNA single strand breaks in cultured mammalian cells. DNA Repair (Amst). 2005;4:271–277. D’Amours D, Desnoyers S, D’Silva I, Poirier GG. Poly(ADP-ribosyl)ation reactions in the regulation of nuclear functions. Biochem J. 1999;342 ( Pt 2):249–268. Luo X, Kraus WL. On PAR with PARP: Cellular stress signaling through poly(ADP-ribose) and PARP-1. Genes Dev. 2012;26:417–432. Hughes AK, Stricklett PK, Kohan DE. Cytotoxic effect of Shiga toxin-1 on human proximal tubule cells. Kidney Int. 1998;54:426–437. Hughes AK, Stricklett PK, Kohan DE. Shiga toxin-1 regulation of cytokine production by human glomerular epithelial cells. Nephron. 2001;88:14–23. Obrig TG, Louise CB, Lingwood CA, et al. Endothelial heterogeneity in Shiga toxin receptors and responses. J Biol Chem. 1993;268:15484–15488. Van Setten PA, van Hinsbergh VW, Van den Heuvel LP, et al. Verocytotoxin inhibits mitogenesis and protein synthesis in purified human glomerular mesangial cells without affecting cell viability: evidence for two distinct mechanisms. J Am Soc Nephrol. 1997;8:1877–1888. Bauwens A, Betz J, Meisen I, et al. Facing glycosphingolipid-Shiga toxin interaction: dire straits for endothelial cells of the human vasculature. Cell Mol Life Sci. 2012;70:425–457. Johannes L, Goud B. Facing inward from compartment shores: how many pathways were we looking for? Traffic. 2000;1:119–123. Spooner RA, Smith DC, Easton AJ, et al. Retrograde transport pathways utilised by viruses and protein toxins. Virol J. 2006;3:26. Sandvig K, Garred O, Prydz K, et al. Retrograde transport of endocytosed Shiga toxin to the endoplasmic reticulum. Nature. 1992;358:510–512. Sandvig K, Torgersen ML, Engedal N, et al. Protein toxins from plants and bacteria: probes for intracellular transport and tools in medicine. FEBS Lett. 2010;584:2626–2634. Suzuki A, Doi H, Matsuzawa F, et al. Bcl-2 antiapoptotic protein mediates verotoxin II-induced cell death: possible association between bcl-2 and tissue failure by E. coli O157:H7. Genes Dev. 2000;14:1734–1740. Raa H, Grimmer S, Schwudke D, et al. Glycosphingolipid requirements for endosome-to-Golgi transport of Shiga toxin. Traffic. 2009;10:868–882. Lingwood CA, Binnington B, Manis A, Branch DR. Globotriaosyl ceramide receptor function - where membrane structure and pathology intersect. FEBS Lett. 2010;584:1879–1886. Lingwood CA, Manis A, Mahfoud R, et al. New aspects of the regulation of glycosphingolipid receptor function. Chem Phys Lipids. 2010;163:27–35. Griffin PM, Tauxe RV. The epidemiology of infections caused by Escherichia coli O157:H7, other enterohemorrhagic E. coli, and the associated hemolytic uremic syndrome. Epidemiol Rev. 1991;13:60–98. Tarr PI, Gordon CA, Chandler WL. Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome. Lancet. 2005;365:1073–1086. Trompeter RS, Schwartz R, Chantler C, et al. Haemolytic-uraemic syndrome: an analysis of prognostic features. Arch Dis Child. 1983;58:101–105. Karmali MA, Gannon V, Sargeant JM. Verocytotoxin-producing Escherichia coli (VTEC). Vet Microbiol. 2010; 140:360–370. Richardson SE, Karmali MA, Becker LE, Smith CR. The histopathology of the hemolytic uremic syndrome associated with verocytotoxin-producing Escherichia coli infections. Hum Pathol. 1988;19:1102–1108.

SHIGA TOXINS: THE RIBOSOME-INACTIVATING PROTEINS FROM PATHOGENIC BACTERIA

57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80.

81.

82. 83. 84.

109

Inward CD, Howie AJ, Fitzpatrick MM, et al. Renal histopathology in fatal cases of diarrhoea-associated haemolytic uraemic syndrome. Pediatr Nephrol. 1997;11:556–559. Ray PE, Liu XH. Pathogenesis of Shiga toxin-induced hemolytic uremic syndrome. Pediatr Nephrol. 2001;16:823–839. Zoja C, Buelli S, Morigi M. Shiga toxin-associated hemolytic uremic syndrome: pathophysiology of endothelial dysfunction. Pediatr Nephrol. 2010;25:2231–2240. Frankel G, Phillips AD. Attaching effacing Escherichia coli and paradigms of Tir-triggered actin polymerization: getting off the pedestal. Cell Microbiol. 2008;10:549–556. Muthing J, Schweppe CH, Karch H, Friedrich AW. Shiga toxins, glycosphingolipid diversity, and endothelial cell injury. Thromb Haemost. 2009;101:252–264. Schuller S. Shiga toxin interaction with human intestinal epithelium. Toxins (Basel). 2011;3:626–639. Hurley BP, Thorpe CM, Acheson DW. Shiga toxin translocation across intestinal epithelial cells is enhanced by neutrophil transmigration. Infect Immun. 2001;69:6148–6155. Acheson DW, Moore R, De Breucker S, et al. Translocation of Shiga toxin across polarized intestinal cells in tissue culture. Infect Immun. 1996;64:3294–3300. Malyukova I, Murray KF, Zhu C, et al. Macropinocytosis in Shiga toxin 1 uptake by human intestinal epithelial cells and transcellular transcytosis. Am J Physiol Gastrointest Liver Physiol. 2009;296:G78–92. Laiko M, Murtazina R, Malyukova I, et al. Shiga toxin 1 interaction with enterocytes causes apical protein mistargeting through the depletion of intracellular galectin-3. Exp Cell Res. 2010;316:657–666. Bielaszewska M, Karch H. Consequences of enterohaemorrhagic Escherichia coli infection for the vascular endothelium. Thromb Haemost. 2005;94:312–318. Bitzan M, Richardson S, Huang C, et al. Evidence that verotoxins (Shiga-like toxins) from Escherichia coli bind to P blood group antigens of human erythrocytes in vitro. Infect Immun. 1994;62:3337–3347. van Setten PA, Monnens LA, Verstraten RG, et al. Effects of verocytotoxin-1 on nonadherent human monocytes: binding characteristics, protein synthesis, and induction of cytokine release. Blood. 1996;88:174–183. Cooling LL, Walker KE, Gille T, Koerner TA. Shiga toxin binds human platelets via globotriaosylceramide (Pk antigen) and a novel platelet glycosphingolipid. Infect Immun. 1998;66:4355–4366. Ghosh SA, Polanowska-Grabowska RK, Fujii J, et al. Shiga toxin binds to activated platelets. J Thromb Haemost. 2004;2:499–506. Karpman D, Papadopoulou D, Nilsson K, et al. Platelet activation by Shiga toxin and circulatory factors as a pathogenetic mechanism in the hemolytic uremic syndrome. Blood. 2001;97:3100–3108. Te Loo DM, Monnens LA, van Der Velden TJ, et al. Binding and transfer of verocytotoxin by polymorphonuclear leukocytes in hemolytic uremic syndrome. Blood. 2000;95:3396–3402. Tazzari PL, Ricci F, Carnicelli D, et al. Flow cytometry detection of Shiga toxins in the blood from children with hemolytic uremic syndrome. Cytometry B Clin Cytom. 2004;61:40–44. Brigotti M, Carnicelli D, Ravanelli E, et al. Interactions between Shiga toxins and human polymorphonuclear leukocytes. J Leukoc Biol. 2008;84:1019–1027. Brigotti M, Tazzari PL, Ravanelli E, et al. Endothelial damage induced by Shiga toxins delivered by neutrophils during transmigration. J Leukoc Biol. 2010;88:201–210. Brigotti M, Tazzari PL, Ravanelli E, et al. Clinical Relevance of Shiga Toxin Concentrations in the Blood of Patients With Hemolytic Uremic Syndrome. Pediatr Infect Dis J. 2011;30:486–490. Te Loo DM, van Hinsbergh VW, van den Heuvel LP, Monnens LA. Detection of verocytotoxin bound to circulating polymorphonuclear leukocytes of patients with hemolytic uremic syndrome. J Am Soc Nephrol. 2001;12:800–806. Arfilli V, Carnicelli D, Rocchi L, et al. Shiga toxin 1 and ricin A chain bind to human polymorphonuclear leucocytes through a common receptor. Biochem J. 2010;432:173–180. van de Kar NC, Monnens LA, Karmali MA, van Hinsbergh VW. Tumor necrosis factor and interleukin-1 induce expression of the verocytotoxin receptor globotriaosylceramide on human endothelial cells: implications for the pathogenesis of the hemolytic uremic syndrome. Blood. 1992;80:2755–2764. Louise CB, Obrig TG. Shiga toxin-associated hemolytic-uremic syndrome: combined cytotoxic effects of Shiga toxin, interleukin-1 beta, and tumor necrosis factor alpha on human vascular endothelial cells in vitro. Infect Immun. 1991;59: 4173–4179. Stahl AL, Sartz L, Karpman D. Complement activation on platelet-leukocyte complexes and microparticles in enterohemorrhagic Escherichia coli-induced hemolytic uremic syndrome. Blood. 2011;117:5503–5513. Stahl AL, Sartz L, Nelsson A, et al. Shiga toxin and lipopolysaccharide induce platelet-leukocyte aggregates and tissue factor release, a thrombotic mechanism in hemolytic uremic syndrome. PLoS One. 2009;4:e6990. Orth D, Khan AB, Naim A, et al. Shiga toxin activates complement and binds factor H: Evidence for an active role of complement in hemolytic uremic syndrome. J Immunol. 2009;182:6394–6400.

110 85. 86. 87. 88. 89. 90. 91. 92. 93.

94. 95. 96. 97.

98. 99. 100. 101. 102. 103. 104. 105.

RIBOSOME-INACTIVATING PROTEINS

Monnens L, Molenaar J, Lambert PH, et al. The complement system in hemolytic-uremic syndrome in childhood. Clin Nephrol. 1980;13:168–171. Thurman JM, Marians R, Emlen W, et al. Alternative pathway of complement in children with diarrhea-associated hemolytic uremic syndrome. Clin J Am Soc Nephrol. 2009;4:1920–1924. Noris M, Remuzzi G. Thrombotic microangiopathy after kidney transplantation. Am J Transplant. 2010;10:1517–1523. Nurnberger J, Philipp T, Witzke O, et al. Eculizumab for atypical hemolytic-uremic syndrome. N Engl J Med. 2009;360:542–544. Gruppo RA, Rother RP. Eculizumab for congenital atypical hemolytic-uremic syndrome. N Engl J Med. 2009;360:544–546. Lapeyraque AL, Malina M, Fremeaux-Bacchi V, et al. Eculizumab in severe Shiga-toxin-associated HUS. N Engl J Med. 2011;364:2561–2563. Morigi M, Galbusera M, Gastoldi S, et al. Alternative pathway activation of complement by Shiga toxin promotes exuberant C3a formation that triggers microvascular thrombosis. J Immunol. 2011;187:172–180. Matussek A, Lauber J, Bergau A, et al. Molecular and functional analysis of Shiga toxin-induced response patterns in human vascular endothelial cells. Blood. 2003;102:1323–1332. Iordanov MS, Pribnow D, Magun JL, et al. Ribotoxic stress response: activation of the stress-activated protein kinase JNK1 by inhibitors of the peptidyl transferase reaction and by sequence-specific RNA damage to the alpha-sarcin/ricin loop in the 28S rRNA. Mol Cell Biol. 1997;17:3373–3381. Tesh VL. Activation of cell stress response pathways by Shiga toxins. Cell Microbiol. 2012;14:1–9. Tesh VL. The induction of apoptosis by Shiga toxins and ricin. Curr Top Microbiol Immunol. 2012;357:137–178. Mangeney M, Lingwood CA, Taga S, et al. Apoptosis induced in Burkitt’s lymphoma cells via Gb3/CD77, a glycolipid antigen. Cancer Res. 1993;53:5314–5319. Scheutz F, Nielsen EM, Frimodt-Moller J, et al. Characteristics of the enteroaggregative Shiga toxin/verotoxin-producing Escherichia coli O104:H4 strain causing the outbreak of haemolytic uraemic syndrome in Germany, May to June 2011. Euro Surveill. 2011;16. Bielaszewska M, Mellmann A, Zhang W, et al. Characterisation of the Escherichia coli strain associated with an outbreak of haemolytic uraemic syndrome in Germany, 2011: a microbiological study. Lancet Infect Dis. 2011;11:671–676. Morabito S, Karch H, Mariani-Kurkdjian P, et al. Enteroaggregative, Shiga toxin-producing Escherichia coli O111:H2 associated with an outbreak of hemolytic-uremic syndrome. J Clin Microbiol. 1998;36:840–842. Morabito S, Karch H, Schmidt H, et al. Molecular characterisation of verocytotoxin-producing Escherichia coli of serogroup O111 from different countries. J Med Microbiol. 1999;48:891–896. Buchholz U, Bernard H, Werber D, et al. German outbreak of Escherichia coli O104:H4 associated with sprouts. N Engl J Med. 2011;365:1763–1770. Rasko DA, Webster DR, Sahl JW, et al. Origins of the E. coli strain causing an outbreak of hemolytic-uremic syndrome in Germany. N Engl J Med. 2011;365:709–717. Group SWR. Experiences from the Shiga toxin-producing Escherichia coli O104:H4 outbreak in Germany and research needs in the field, Berlin, 28–29 November 2011. Euro Surveill. 2012;17. Frank C, Werber D, Cramer JP, et al. Epidemic profile of Shiga-toxin-producing Escherichia coli O104:H4 outbreak in Germany. N Engl J Med. 2011;365:1771–1780. Nataro JP, Kaper JB. Diarrheagenic Escherichia coli. Clin Microbiol Rev. 1998;11:142–201.

8

The Structure and Action of Ribosome-inactivating Proteins Jon D. Robertus and Arthur F. Monzingo Institute for Cellular and Molecular Biology, University of Texas at Austin, USA

Introduction

Protein synthesis is a complex, highly-regulated, multi-step process in all living cells.1 The process can be readily subverted by pathogens to attack host cells, or by host cells to attack pathogens. For example, it is known that the bacterial diphtheria toxin acts by inserting NAD into the active site of EF-2, inhibiting translation and killing the cell.2 The toxin is controlled by iron levels and is presumably used to aid the pathogen in obtaining that essential nutrient.3 In the obverse case, higher plants can defend against predation by bacteria, fungi, insects, or even mammals by attacking their translation systems. Probably the most widespread such method involves ribosome-inactivating proteins: RIPs. These use an N-glycosidase enzyme function to depurinate a crucial adenine in a highly conserved stem/loop of the large ribosomal subunit.4 Depurination of the ribosome inhibits protein synthesis and leads to cell death. RIPs fall into several structural classes.5 Type 1 RIPs are simple 30 kDa N-glycosidase enzymes, while type 2 have the glycosidase, called the A chain, complexed with one or more cell binding, or B, chains; the B chains are generally lectins. The type 2 RIPs are able to enter host cells much more effectively and are far more cytotoxic than type 1 RIPs. The type 2 RIP ricin is about 300,000 times more potent than its isolated A chain or the type 1 RIP pokeweed antiviral protein (PAP) when attacking cultured T-cells.6 Although type 1 RIPs are not cytotoxic, they still appear to play a defensive role in plants. One of the first discovered RIPs is the pokeweed antiviral protein, PAP, from Phytolacca americana; an extract from the plant was able to retard virus infection of abraded leaves. 7 It has been shown by electron microscopy that PAP is stored in the space between the cell wall and the plasma membrane; when the cell is compromised by a breach, and therefore susceptible to viral infection, PAP enters the cytoplasm, inhibits its own ribosomes, and shuts down viral replication.8 Type 2 RIPs are true cytotoxins and probably serve to protect the host. Ricin is stored in the seeds of the castor plant, and presumably its potent toxicity limits predation of the seeds by insects or small mammals. The evolutionary history of the RIP superfamily  is  unclear. The N-glycosidases, or A chains, from a wide range of plants are

Ribosome-inactivating Proteins: Ricin and Related Proteins, First Edition. Edited by Fiorenzo Stirpe and Douglas A. Lappi. © 2014 John Wiley & Sons, Inc. Published 2014 by John Wiley & Sons, Inc.

111

112

RIBOSOME-INACTIVATING PROTEINS

clearly homologous based on amino acid sequence and structure;9 it is also clear that these are homologs of such bacterial toxins as Shiga toxin or the Shiga-like toxins from Escherichia coli.10 It is plausible that the N-glycosidases evolved in bacteria and were passed to higher plants by lateral gene transfer. Some of these proteins associated with lectins to form the type 2 RIPs while others did not. The RIP N-glycosidases have been shown to depurinate a very specific adenine in rRNA.11, 12 It was originally shown that ricin attacks A4324 in rat 28S rRNA, but this specificity has been shown to be general to all the RIPs.4 The adenine is part of an invariant GAGA sequence within a conserved stem and loop structure called the sarcin/ricin loop (SRL), known to be essential for binding GTP-dependent translation factors and for ribosome assembly.13 The N-glycosidase enzymes have evolved to be exquisitely selective for their SRL target. Ricin attacks rabbit ribosomes with a kcat = 25/s and Km = 0.1 μM.14 This means that the specificity constant, kcat/Km, is >108 and suggests ricin depurination is essentially diffusion controlled. When the reaction rate is limited, as here, by the rate at which the enzyme can physically collide with its substrate, it is said to have reached catalytic perfection.15 Similar kinetic constants have been measured using other ribosome targets16 and for other RIPs like PAP.17 In addition to its potent ribosome N-glycosidation reaction, there have been assertions that RIPs carry out other reactions, acting as DNases or attacking single-stranded RNA. However, careful analysis, in particular examination of kinetic rates and conditions, show these claims are physiologically suspect. For example, a purported DNase activity was shown conclusively to be due to nuclease contamination. 18 Evaluating other work in this area is complicated by the fact that reaction rates are so slow for non-ribosomal substrates that steady-state kinetic conditions, such as substrate in great excess over enzyme, are rarely used. Schramm and his co-workers have done well-controlled quantitative experiments, and these show that depurination of naked RNA is barely detectable at physiological pH, even for SRL mimics.19 At pH 4, they are able to depurinate RNAs with specificity constants about 1000 times poorer than that for ribosomes. Similar, but less qualitative, observations have been made by others.20 Analysts have concluded that there is little evidence to support any major physiological role for these “extra” RIP actions. 21 The observed depurinations from non-ribosomal substrates may be simple, slow, side reactions by the RIP, which does clearly recognize RNA in its physiological role. 9 The RIP family is of great scientific and practical interest. The most well-known and thoroughly studied member of the family is ricin, derived from the castor plant, Ricinus communis. Ricin was first studied, and named, by Stillmark in 1888.22 Over the years there has been a great deal of effort expended to understand its great potency and to harness that toxicity for medically beneficial purposes. Ricin’s inherent toxicity was examined as an antitumor agent.23 Later, efforts were made to couple ricin A chain, other RIPs, or other enzyme toxins to tumor specific antibodies to create immunotoxins.24–26 Recently there has been considerable interest in understanding the exact mechanisms of ricin uptake, trafficking, and toxicity.27, 28 Ricin has also been the focus of recent public attention because of its use as a bioterrorist weapon.29–31 This has led to efforts to develop ricin neutralizing antibodies,32, 33 antiricin vaccines,34, 35 and small molecule ricin inhibitors.36–38 The scientific and practical interest in RIPs led to the chemical and structural characterization of many examples of this superfamily. There have been a number of reviews of the continuing exploration of the family over the years.5, 9, 39, 40 In this review we examine the structure of RIPs, the relationship to RIP mode of action, and its use in therapeutic and drug design.

THE STRUCTURE AND ACTION OF RIBOSOME-INACTIVATING PROTEINS

113

Ricin Structure

Ricin is the archetype of the RIP superfamily; it is the most well-known and widely studied member of the group. It was also the first RIP for which an X-ray structure was determined.41 Ricin is synthesized as a preproenzyme and is activated by proteolysis to form the classic disulfide-linked AB heterodimer.42 The ricin A chain, RTA, has 267 amino acid residues and a MW of 29 kDa and the B chain, RTB, has 262 residues and a MW of 28 kDa. The amino acid numbers correspond to the processed protein; a disulfide links C259 of RTA to C4 of RTB. Both RTA and RTB are N-glycosylated, as observed from the chemical sequencing of the proteins isolated from castor seeds.43, 44 As a consequence, the observed molecular weight of each chain is closer to 32 kDa, and the heterodimer tends to run at about 65 kDa on chromatography columns.40 The structure of ricin was solved by multiple isomorphous replacement crystallography in 1987.41 It was refined later to 2.5 Å,45 and the structure of RTA46 and RTB47 described in detail. Figure 8.1 shows a model of the intact protein. RTA, shown in blue, is a globular protein with substantial secondary structure. It is about 36% α-helix and 18% β-sheet. Formycin monophosphate, FMP, is an AMP analog, and the structure of ricin bound with FMP first revealed the active site region and the mode of substrate binding.48 RTB, in red, is an elongated structure composed of two domains that arose by gene duplication.49 It binds most strongly to galactosides, and the structure was solved in the presence of galactose. Two galactose molecules bind RTB, one to each domain, and these are shown as stick figures on either end of the dumbbell shaped protein. The disulfide bond linking RTA and RTB is shown as yellow bonds. Thermodynamic measures show that the association of RTA and RTB is largely hydrophobic in character, with a Ka ~ 3x106 (that is a free energy of association of about −37 KJ).50 Ricin can be N-glycosylated on both chains, although modification of RTA is often incomplete.51 The X-ray structure showed two glycosylation sites on RTB and these are indicated as NAG–NAG moieties shown as line structures. No glycosylation sites were seen on RTA;

Lac-2γ

FMP

Lac-1α Figure 8.1 The structure of ricin. The A chain, RTA, is colored blue and the B chain, RTB, is red. The disulfide bond linking the two is shown in yellow. Key ligands are shown as bonded molecules with carbon atoms colored cyan. FMP binds to, and marks the RTA active site. The two galactoside-binding sites of RTB are marked by lactose models. RTB glycosylation sites are marked by unlabeled sugar structures. (This figure also appears in the color plate section.)

114

RIBOSOME-INACTIVATING PROTEINS

and indeed if the sites were glycosylated, the protein would not have crystallized in the space group observed. That is, crystallization selected for ricin that was unmodified on RTA. In addition to ricin, the structures of several other type 2 plant RIPs have been solved. These include abrin (PDB code 1ABR),52 viscumin (1CE7),53 and ebulin (1HWN).54 These proteins are all homologs. If we compare their A chain sequences to RTA, the % identity/(% similarity) are 38/(53), 34/(49), and 36/ (57) respectively. If we compare the B chains to RTB, the values are 54/(72), 61/(74), and 48/(65) respectively. Not surprisingly, the three-dimensional structures are very similar with identical topological folds. When the heterodimers are superimposed on ricin, the RMS deviations of Cα atoms are 1.06, 0.77, and 1.48 Å respectively. In addition to their conservative folding pattern, these type 2 RIPs are known to have the same mechanism of action and so a description of the archetype ricin applies, in general, to all RIPs.

Ricin B Chain

As seen in Figure 8.1, RTB is a dumbbell shaped two-domain protein, seen originally from a low resolution structure.49 Subsequent structural analysis revealed that the evolutionary history of the lectin was more nuanced than simple gene duplication.47, 55 Each RTB domain, 1 and 2, is composed of three subdomains, labeled α, β, and γ. These domains assemble into a trefoil structure with pseudo threefold symmetry. Each subdomain conserves hydrophobic residues that form the core of the trefoil. The organization of domain 1 is shown in Figure 8.2a. Each subdomain contributes a conserved Trp that is the heart of the core, but other conserved aliphatic groups like Ile and Val also participate. The conformation of an individual subdomain, 1α, is shown in Figure 8.2b. The long C-terminal tail packs into the center of the trefoil domain structure, anchored by the conserved W37 in this subdomain. (b)

(a)

C39 C20 W57 W1β

D22 W49

W1γ

W1α

N46 Lac-1α

W57

Lac-1α

Figure 8.2 Structural elements of RTB. (a) A backbone drawing of domain 1 of RTB, viewed down the pseudo three-fold operator relating subdomains α, β, and γ. The central hydrophobic core contains conserved non-polar side chains shown as sticks; each subdomain contributes a Trp labeled W1α, W1β, and W1γ respectively. The galactose-binding site in domain 1 resides in subdomain γ and is indicated by the Lac-1γ. (b) The galactose-binding domain, 1γ, is shown. The non-polar face of galactose binds to W57. Specific hydrogen bonds are made between the sugar alcohols and RTB side chains D22 and N46. The view is roughly perpendicular to the domain pseudo three-fold, and the main anchoring W49 is shown.

THE STRUCTURE AND ACTION OF RIBOSOME-INACTIVATING PROTEINS

115

In addition to the conserved core hydrophobic residues, there are surface residues on the subdomain that can participate in galactoside binding. However, these are not conserved in all units, and one sees that only 1α and 2γ subdomains bind sugars in modern RTB, although the original structure presumably formed from a non-covalent association of three galactose-binding units. Figure 8.2b shows the disaccharide lactose binding to unit 1α. The sugar-binding site is shallow, with the side chain of W37 acting as a cover, and a kink in the polypeptide acting as the floor. Four hydrogen bonds are made between galactose and D22, Q35, and N46. The key protein residue is D22 that points to the sugar’s leading edge, directly between hydroxyls 3 and 4, and accepts hydrogen bonds from them. This is the mechanism of sugar binding specificity; glucose is epimeric with galactose at C4 and so cannot bind effectively to RTB. The binding constants for galactosides are fairly weak, with Kd values for lactose about 0.13 mM.56 Despite this poor binding to sugars, ricin binds tightly to cells for two reasons. (i) RTB has two independent binding sites, so their free energies of association are additive and the effective binding constant for ricin to a cell is much stronger than for each single site. (ii) The surface of eukaryotic cells is covered with galactosides; it has been shown that there are about 2 million sites for ricin on both erythrocyte and HeLa cells.57 As seen in Figure 8.1, the 1α and 2γ domains that retain sugar-binding capacity are on the extreme end of the protein; it is possible that this may optimize cell surface binding. In any case, RTB binds ricin to the target cell and facilitates uptake and toxicity. The proteins are endocytosed and undergo retrograde transport to the endoplasmic reticulum (ER), from where RTA enters the cytoplasm by subverting the ER associated protein degradation, or ERAD, pathway.58

Ricin A Chain

The A chains of the RIPs function as specific N-glycosidases, removing a key, invariant, adenine base from the SRL of the large ribosomal subunit.4, 12 As described in the introduction, this reaction has evolved to near perfection, being essentially limited by diffusion. The X-ray structure sheds light on the nature of this exquisite mechanism, allowing a description of the binding of substrate and identifying protein residues that are crucial to substrate binding and catalysis. Specificity Pocket

The first substrate analog shown in complex with RTA was the non-hydrolysable AMP analog formycin monophosphate, FMP.48 Its mode of binding is shown in Figure 8.3. The purine ring binds in a pocket that is called the specificity pocket. The substrate ring stacks with the side chain of Y80 and also contacts the side chain of Y123. The side chains of V81, I172, and I175 add to the hydrophobic character of the pocket. The orientation of Y80 is worthy of comment. In the native ricin structure, Y80 is rotated so as to block the specificity pocket. A substrate, or other ligand, must displace this side chain to bind into the specificity pocket; a number of complexes with substrate analogs and other ligands have been solved for various RIPs, which show Y80 can take up a number of related conformations depending on the nature of the ligand. Substrate binding clearly involves aromatic stacking of the substrate adenine between the side chains of invariant tyrosines 80 and 123. However, binding specificity is mediated by hydrogen bonds between the substrate ring and the protein; the major interactions are indicated in cartoon form in Figure 8.4a. The amido nitrogen of V81 donates a bond to N1 of the adenine ring. The exocyclic amine of adenine donates hydrogen bonds to the carbonyl oxygens of V81 and G121. A hydrogen bond is donated by the side chain of R180 to N3 of the adenine (this will be shown later to be mechanistically important).

116

RIBOSOME-INACTIVATING PROTEINS

G121

G121 V81

V81 Y80

I172

Y80

I172

Y123

Y123

E177

E177 R180

R180

Figure 8.3 FMP binding to RTA. This stereogram shows details of the mode of binding of the substrate analog FMP to the specificity pocket of the enzyme. The purine ring slips into a pocket defined by the side chains of tyrosines 80 and 123. Specific hydrogen bonds are formed, and shown as dashed lines.

(b)

(a)

G121

G121 Y123

Y123

O NH2 N

OH

N

N

O

HN

N

H N

V81

N R OH

R O

H2N

E177

N

P6 O

+ NH2

– O

O

– O

NH R180

E177

H2 N

NH 2

O

N

HN

V81

O + NH2

NH R180

Figure 8.4 Schematic of the RTA binding site. (a) Details of adenine binding to the active site are shown. (b) The binding of the pterin-based inhibitor family is shown. The inhibitor makes more and stronger (shorter) hydrogen bonds than does the natural substrate.

THE STRUCTURE AND ACTION OF RIBOSOME-INACTIVATING PROTEINS

117

Active Site Residues and Mechanism

The X-ray structure identified key active site residues of RTA, and by extension of all the RIP N-glycosidases. This allowed the role of these residues to be examined using site directed mutagenesis.16, 59–63 Table 8.1 summarizes the main findings of these studies; because mutants are characterized from differing assay systems, the second column, overall activity relative to wildtype RTA, may be the most useful in assessing the significance of a mutation. Steady-state kinetic parameters have been determined for a few key residues. The X-ray structure suggests that R180, invariant in all RIPs, bonds to N3 of the substrate adenine (see Figure 8.4a). If this residue is replaced by the long polar amino acid Gln, activity is reduced over 1000-fold. The Km for ribosomes is essentially unchanged, but kcat is reduced 1000-fold. This suggests that R180 plays a key catalytic role in the hydrolytic mechanism, probably partially or fully protonating the leaving adenine group.62 Another key residue is the RIP invariant E177. It has long been suspected that the depurination reaction, breaking the C1’–N9 bond of adenosine, would involve ribose oxacarbenium (that is ribose carbocation) character in the transition state, by analogy with glycohydrolases.64 This state is generally stabilized by a carboxylate, and the active site geometry made E177 the obvious choice for that catalytic role.16, 62 Consistent with that hypothesis, conversion to the neutral amide reduces activity 200-fold. Interestingly, conversion to Ala reduced activity only 20-fold63 to 60-fold.61 Structural analysis solved this puzzle by showing that removing the 177 side chain in the E177A mutant allowed a bystander residue, E208, to move in and partially compensate for the loss of the native carboxylate.61 The X-ray structures provide a geometric basis to hypothesize about the RIP reaction and transition state, but they are not definitive. Schramm and his coworkers have analyzed kinetic isotope effects on a variety of synthetic stem/loop nucleotide substrates to shed light on the reaction dynamics. They have clearly demonstrated that the transition state does indeed involve ribooxacarbenium character,65 has a high degree of ribose-adenine bond dissociation (>3 Å), and involves attack by an activated water.66 The detailed understanding of the transition state allowed the group to synthesize a variety of nucleotides that incorporated transition state geometry and electrostatics and that bound very tightly to RTA.66, 67 Table 8.1

Kinetic parameters of RTA mutants.

Protein Wild type RTA R180K R180Q R180H E177Q E177D E177A Y80F Y80S Y123F Y123S N209S W211F R134A R213A R258A

Relative activity

Km (μM)

Kcat min−1

Reference

100 38 5 Au; 2[AC] Sorghum bicolor: 14 Au; 1 [AB] Zea mays: 7 Au; 2 [AB]; 1[AC]; 1 [AD]

Poaceae species: Further details are required to establish the relationships between the Poaceae RIP genes

Signal peptide Pseudogene [AB]

[AΔB]

[ΔAB]

[AX]

[AΔx]

[ΔAX]

Clustered genes

* Assembly does not allow tracing possible RIP linkage Figure 9.3

Schematic representation of the RIP genes found in completed plant genomes.

only ortholog(s) with a CDS corresponding to a protein that is slightly truncated (as compared to the most closely related paralog/homolog) at either the N- or C-terminus but still covering a complete RIP domain, could be identified. In seven other cases no ortholog(s) with a (nearly) complete CDS could be retrieved. However, a complete CDS could be “reconstructed” by either a small change in

142

RIBOSOME-INACTIVATING PROTEINS

the nucleotide sequence (e.g., insertion/deletion/substitution of a single nucleotide to remove a frame shift or convert a stop codon in a sense codon) or removal of a transposable element. Due to the absence of a CDS, the latter genes can be considered pseudogenes. The rice RIP gene family can be subdivided in three main groups on the basis of the domain architecture of the protein products. Fifty-one genes encode proteins in which, besides the A domain, no other conserved domain can be recognized, and accordingly can be classified as a type [A] RIP. However, it should be emphasized that in many of these presumed type [A] RIPs the A domain is accompanied by short or middle long (up to 100 amino acid residues) extensions at the N- and/or C-terminus. Four other genes definitely display a chimeric domain architecture (3 [AC] and one [AP] form). At present the phylogeny of the Oryza RIP gene family has not been worked out in detail. However, a preliminary dendrogram (Figure 9.4) gives a good idea of (i) the complexity of the gene family, (ii) the occurrence of multiple clades, and (iii) the distribution all over the genome. For the sake of clarity, the identification of the genes is based on the position of the genes on the chromosomes and the domain architecture of the protein products. The dendrogram shown in Figure 9.4 differs quite substantially from that reported by Jiang et al. (2008), who made a genome-wide survey of the RIP domain in Oryza sativa, but confirms that genome-wide and tandem duplications are at the basis of the large RIP gene number. Similar analyses indicate that the Sorghum bicolor, Zea mays, Brachypodium distachyon, and Setaria italica RIP gene families are less complex. In contrast, the RIP gene families in Hordeum vulgare and Triticum aestivum might be of a complexity comparable to that of Oryza. Irrespective of the remaining uncertainty, one can reasonably conclude that the complexity of the RIP gene family in Poaceae species exceeds that found in dicots and date palms. Moreover, it seems that the evolution of the RIP gene family within the Poaceae followed a pathway well distinct from that of the dicots and the non-Poaceae monocots.

New Insights in the Overall Phylogeny of Plant RIPs

Genome and transcriptome sequence information deposited in the databases after completion of the phylogenetic analysis reported in our previous review,5 revealed that the evolutionary model proposed at that time needs to be updated and refined. To highlight the improvements, it is summarized here to what extent the major issues of our previous model built from the sequences available until 2009 still hold true or have to be updated or revised. Issue 1: it was proposed that an ancestor of modern seed plants developed the RIP domain at least 300 million years (Myr) ago. This conclusion still holds true. Issue 2: it was proposed that an ancestral RIP domain gave rise to (i) a direct lineage of type 1 RIPs (named primary type 1 RIPs that are still found in many monocots and at least one dicot), and (ii) an ancestral type 2 RIP through fusion with a duplicated (prokaryotic) ricin-B domain. The presumed direct lineage of type 1 RIPs might still exist but is not represented by the so-called primary type 1 RIP. Until recently, the only identified member of this lineage within the dicots was a presumed type 1 RIP from Populus sp. Now it appears that closely related type [A] proteins are expressed in several other dicots (e.g., Fagus sylvatica, Cannabis sativa, Elaeocarpus photiniifolius). Moreover, several species (e.g., cannabis, cacao, Elaeocarpus photiniifolius) express one or more chimeric proteins consisting of an N-terminal domain sharing high sequence similarity with the “primary type 1 RIP” fused to an unknown C-terminal domain. Though these chimeric proteins are to a certain extent reminiscent of the jasmonate responsive type [AC] RIP

UPDATED MODEL OF THE MOLECULAR EVOLUTION OF RIP GENES

143

Oryg03.A8 Oryg03.A11 Oryg03.A6 Oryg03.A9 Oryg03.A12 Oryg03.A10 Oryg03.A13 Oryg03.Aps2 Oryg03.A7 Oryg10.A4 Oryg02.A1 Oryg04.A1 Oryg10.A1 Oryg10.A2 Oryg10.A3 Oryg03.A1 Oryg03.A3 Oryg03.Aps1 Oryg03.A5 Oryg03.A2 Oryg03.A4 Oryg07.A3 Oryg01.A1 Oryg01.A2 Oryg07.Aps1 Oryg07.A4 Oryg07.A5 Oryg11.A1 Oryg11.A2 Oryg04.A2 Oryg12.A1 Oryg08.A4 Oryg08.A3 Oryg08.A1 Oryg08.A2 Oryg02.A2 Oryg02.A3 Oryg12.A3 Oryg12.A4 Oryg11.AC3 Oryg11.AC1 Oryg11.AC2 Oryg05.A1 Oryg09.A1 Oryg01.AP1 Oryg04.A3 Oryg11.A4 Oryg11.A5 Oryg12.A2 Oryg07.A2 Oryg11.A3 Oryg12.A5 Oryg12.A6 Figure 9.4 Phylogenetic tree of the rice RIP gene family. Genes are indicated by their relative position on the chromosomes. Dendrogram was rendered using Mega5.7

144

RIBOSOME-INACTIVATING PROTEINS

found in barley, the overall sequence similarity between both proteins is rather low, especially in the C-terminal domain. Therefore, these newly identified chimers are considered members of a novel type of RIPs, further referred to as the type [AX] RIP. Phylogenetic analyses leave no doubt about the common origin of the “primary type 1 RIPs” and the [A] domain of the [AX] chimers. Moreover, both the phylogeny and taxonomic distribution indicate that what we believed to represent “primary type 1 RIPs” is in fact a second lineage of secondary type 1 RIP (further referred to as type [AΔX]) derived from type [AX] RIP through a domain deletion event (Figure  9.2). Besides in dicots, type [AΔX] RIPs and/or corresponding genes were also identified in monocots. Some of these monocot type [AΔX] RIP have been isolated and (partly) characterized like asparin 1 and 2 from Asparagus officinalis,8 the musarmins from Muscari armeniacum9 and charybdin from Charybdis maritima.10 At present, there are no data that challenge the proposed origin of an ancestral type 2 RIP gene through fusion of an A domain with a duplicated (prokaryotic) ricin-B domain. Issue 3: the ancestral type 2 RIP gave rise to (i) a monophyletic line covering all modern type 2 RIPs, and (ii) different lines of “secondary” type 1 RIPs and ricin-B type lectins (through domain deletion events). Both conclusions remain valid. Moreover, sequence-based evidence was obtained for several independent domain deletion events (Figure 9.1). Issue 4: at least three other domain fusions took place in the recent past within the family Poaceae, whereby type AC1 (type 3), type AC2, and type AD chimeric forms were generated. The presumed origin within the family Poaceae of three other chimeric forms through recent independent domain fusions needs to be revised in view of the identification of the [AX] RIP lineage. Though the Poaceae [AC] and [AD] types do not share a high sequence identity with the [AX] RIP, the residual similarity might be indicative for a common origin. Accordingly, the Poaceae [AC] and [AD] types might have evolved directly from an early Poaceae [AX] form. Evolutionary events comparable to those observed for the dicot [AX] genes can explain the occurrence in some grasses of proteins with a C or D domain and, what is more important, might have given rise to still other lineages of secondary type 1 RIPs. An additional but important novel insight concerns the identification in some grass species of chimeric RIPs with a C-terminal peptidase domain. In Oryza sativa the RIP domain is fused to a peptidase M41 domain (RIP [APM41]) and in Brachypodium distachyon and Triticum aestivum to a peptidase C19 domain (RIP [APC19]). Both the rice RIP [APM41] and wheat RIP [APC19] are expressed, indicating that these grass species do synthesize (cytoplasmic) proteins with a (putative) dual N-glycosidase/protease activity. At present, there is no experimental evidence in support of the possible protease activity of the expressed proteins. However, since the canonical metal binding and catalytic HEXXH motif of the peptidase M41 family is conserved in rice RIP [APM41], and all four active site residues of the peptidase C19 domain are strictly conserved in wheat RIP [APC19], there is a reasonable chance that both RIP-protease chimers possess proteolytic activity. Evidently, the identification of these two novel previously unconceived domain architectures widens the range of possible biological activities/functions of plant RIPs and increases the complexity of the evolution of the plant RIP family.

An Updated Model of the Molecular Evolution of the Plant RIP Gene Family

The novel insights in the taxonomic distribution and overall phylogeny render our recently proposed model5 of the molecular evolution of the plant RIP gene family outdated. Part of the original scheme remains valid but three major novelties need to be incorporated in the new model: (i) the occurrence

145

UPDATED MODEL OF THE MOLECULAR EVOLUTION OF RIP GENES

of the [AX] RIP lineage, (ii) the presence in some grass species of two different RIP-protease chimers, and (iii) the revised origin of the [AC] and [AD] lineages within the family Poaceae. A detailed scheme of the updated model is presented in Figure 9.5. At present there are no indications to reject the idea that the RIP domain itself was developed in plants before the Gnetophyta and Magnoliophyta lineages diverged from a common ancestor approximately 300 Myr ago,11 and that – also before the Gnetophyta and Magnoliophyta lineages separated – this RIP domain was fused to a sugar-binding domain that might have been acquired by lateral transfer from a bacterium. The resulting ancestral type [AB] RIP gave rise (through vertical inheritance) to the modern type [AB] RIPs and by B domain deletion/gene truncation events to multiple lines of (secondary) type [AΔB]. In addition, deletion of the A domain resulted in the generation of [BΔA] lectin genes. Genome analyses demonstrated that type [AB] RIP genes have been purged from many genomes, and reveal that this process of gene loss is still ongoing. In contrast, other species developed a whole set of type [AB] RIP genes by genome-wide amplifications and gene duplications. A comparable fusion generated an ancestral chimeric type [AX] RIP gene that evolved in a similar way as the ancestral type [AB] RIP gene and gave rise to the modern type [AX] RIPs (through vertical inheritance), to multiple lines of (secondary) type [AΔX] RIPs (by deletion/gene truncation events), and by deletion of the A domain also to [XΔA] genes. Genome analyses

Poaceae