Promise of Nutritional Lithium for Mental Health : Annual Review of Pharmacology and Toxicology: 2001 (Annual Review of Pharmacology & Toxicology) 0824304411, 9780824304416

Citation preview

ANU ROOObIb37 76

ake Cy 33 ~as

: a Wwae Loe

Be?

:

4UN

ets

EU

PeVOsey

hac

path

eke Br Eo) ER 14. Dy Burke

eoae

RS Pa

Md Carlsson) (BGetson A

RLS

Can)

Chin ee PCs 0 Jed a Contos SG Dau

2B

Kine Ee

x Dashfoe ae SC Relara

Eebecava ee

Slee

de Lesko. Oe Ra ee, Lin

oe

ee ale Linden |

WC. deGroat: .ae

ON Dore

a ee

DE Denhardt Ci Maren Zi Boull eas S RecaD VOR Ruane tt e gees ea

Li Loy

a Rn

;

Digitized by the Internet Archive in 2022 with funding from Kahle/Austin Foundation

https://archive.org/details/annualreviewofph0041 unse_z2l8

ATOM ROOOb343778

Wallace

. Library

DUE DATE (stamped in blue)

RETURN

DATE (stamped in b

| ANNUAL REVIEW OF PHARMACOLOGY AND TOXICOLOGY

EDITORIAL COMMITTEE (2001) TERRENCE F. BLASCHKE ARTHUR K. CHO JAMES K. COWARD RITA M. HUFF PAUL A. INSEL CurTIS D. KLAASSEN

JOHN S. LAZO HORACE H. LOH

ANTHONY Y. H. Lu

RESPONSIBLE FOR THE ORGANIZATION OF VOLUME 41 (EDITORIAL COMMITTEE, 1999) TERRENCE F. BLASCHKE ARTHUR K. CHO JAMES K. COWARD RAYMOND J. DINGLEDINE

ING K. Ho RITA M. HUFF PAUL A. INSEL JOHN S. LAZO HORACE H. LOH ANTHONY Y. H. Lu KENNETH S. RAMOS

LESLIE Z. BENET (GUEST)

Production Editor: CLAIRE INSEL Bibliographic Quality Control: MARY A. GLASS

Color Graphics Coordinator: EME O. AKPABIO Online Quality Control: NOEL J. THOMAS Subject Indexer: SUZANNE COPENHAGEN

ANNUAL REVIEW OF PHARMACOLOGY AND TOXICOLOGY VOLUME 41, 2001

ARTHUR K. CHO, Editor University of California School of Medicine, Los Angeles

TERRENCE

F. BLASCHKE, Associate Editor

Stanford University Medical Center, Stanford

PAUL A. INSEL, Associate Editor University of California School of Medicine, San Diego

HORACE

H. LOH, Associate Editor

University of Minnesota Medical School, Minneapolis

www.AnnualReviews.org

[email protected]

650-493-4400

ANNUAL REVIEWS 4139 El Camino Way « P.O. BOX 10139 * Palo Alto, California 94303-0139

>| ANNUAL REVIEWS Palo Alto, California, USA

COPYRIGHT © 2001 BY ANNUAL REVIEWS, PALO ALTO, CALIFORNIA, USA. ALL RIGHTS RESERVED. The appearance of the code at the bottom of the first page of an article in this serial indicates the copyright owner’s consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition that the copier pay the stated per-copy fee of $14.00 per article through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, MA 01923) for copying beyond that permitted by Section 107 or 108 of the US Copyright Law. The per-copy fee of $14.00 per article also applies to the copying, under the stated conditions, of articles published in any Annual Review serial before January 1, 1978. Individual readers, and nonprofit libraries acting for them, are permitted to make a single copy of an article without charge for use in research or teaching. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. For such uses, written permission is required. Write to Permissions Dept., Annual Reviews, 4139 El Camino Way, P.O. Box 10139, Palo Alto, CA 94303-0139 USA. International Standard Serial Number:

International Standard Book Number:

0362-1642

O0-8243-044]-1

Annual Review and publication titles are registered trademarks of Annual Reviews.

@& The paper used in this publication meets the minimum requirements of American National Standards for Information Sciences—Permanence of Paper for Printed Library

Materials, ANSI Z39.48-1992. Annual Reviews and the Editors of its publications assume no responsibility for the statements expressed by the contributors to this Annual Review.

TYPESET BY TECHBOOKS, FAIRFAX, VA PRINTED AND BOUND IN THE UNITED STATES OF AMERICA

AR nua Review of Pharmacology and Toxicology Volume 41, 2001

CONTENTS FRONTISPIECE, John Doull TOXICOLOGY

COMES

ANESTHETICS

AND

OF AGE, John Doull

ION CHANNELS:

MOLECULAR

MODELS

AND

SITES

OF ACTION, Tomohiro Yamakura, Edward Bertaccini, James R. Trudell, and R. Adron Harris TUMOR

CELL DEATH

INDUCED

BY TOPOISOMERASE-TARGETING

DRUGS, Tsai-Kun Li and Leroy F. Liu THE CLINICAL

PHARMACOLOGY

OF L-ARGININE,

Rainer H. Boger

and Stefanie M. Bode-Boger PHARMACOGENOMICS:

UNLOCKING

THE HUMAN

GENOME

FOR

101

BETTER DRUG THERAPY, Howard L. McLeod and William E. Evans

PHENOBARBITAL RESPONSE ELEMENTS OF CYTOCHROME P450 GENES 123

AND NUCLEAR RECEPTORS, T. Sueyoshi and M. Negishi REGULATION

AND

ROLE

OF ADENYLYL

CYCLASE

ISOFORMS,

Jacques

145

Hanoune and Nicole Defer THE BASIC

AND

VASOPRESSIN

CLINICAL

PHARMACOLOGY

RECEPTOR

ANTAGONISTS,

OF NONPEPTIDE M. Thibonnier, P. Coles,

15

A. Thibonnier, and M. Shoham

NOVEL EFFECTS OF NITRIC OXIDE, Karen L. Davis, Emil Martin, Illarion V. Turko, and Ferid Murad INTERACTIONS

BETWEEN

MONOAMINES,

GLUTAMATE,

AND

GABA

203

IN

SCHIZOPHRENIA: NEW EVIDENCE, Arvid Carlsson, Nicholas Waters, Susanna Holm-Waters, Joakim Tedroff, Marie Nilsson, and

Pie)

Maria L. Carlsson PROPERTIES

AND

BIOLOGICAL

ACTIVITIES

OF THIOREDOXINS,

261

Garth Powis and William R. Montfort REGULATION,

FUNCTION,

AND

TISSUE-SPECIFIC

EXPRESSION

OF

CYTOCHROME P450 CYPIB1, Graeme I. Murray, William T. Melvin, William F. Greenlee, and M. Danny Burke

PHYSIOLOGICAL FUNCTIONS OF CYCLIC ADP-RIBOSE AND NAADP AS CALCIUM MESSENGERS, Hon Cheung Lee

oo,

vi

CONTENTS

USE OF BIOMARKERS AND

DEVELOPMENT

AND

SURROGATE

REGULATORY

ENDPOINTS

IN DRUG

MAKING:

DECISION

CRITERIA,

347

VALIDATION, STRATEGIES, LJ Lesko and AJ Atkinson, Jr.

CELLULAR RESPONSES TO DNA DAMAGE, Chris J. Norbury and 367

lan D. Hickson ANTISENSE

OLIGONUCLEOTIDES:

PROMISE

AND

REALITY,

403

Trina Lebedeva and C. A. Stein CANCER

CHEMOPREVENTION

USING

NATURAL

VITAMIN

D AND

SYNTHETIC ANALOGS, Kathryn Z. Guyton, Thomas W. Kensler, and 421

Gary H. Posner METABOLISM

OF FLUORINE-CONTAINING

DRUGS,

B. Kevin Park,

Neil R. Kitteringham, and Paul M. O'Neill

Ca*+/CaM-DEPENDENT KINASES: FROM ACTIVATION TO FUNCTION, Sara S. Hook and Anthony R. Means

471

LYSOPHOSPHOLIPID RECEPTORS, Nobuyuki Fukushima, Isao Ishii, James J. 507

A. Contos, Joshua A. Weiner, and Jerold Chun INTERINDIVIDUAL

VARIABILITY

IN INHIBITION

AND

INDUCTION

OF

CYTOCHROME P450 ENZYMES, Jiunn H. Lin and Anthony Y. H. Lu NEUROTROPHIC

AND NEUROPROTECTIVE

ACTIONS

OF ESTROGENS

Ia AND

THEIR THERAPEUTIC IMPLICATIONS, Susan J. Lee and Bruce S. McEwen

569

GENETIC VARIATIONS AND POLYMORPHISMS OF G PROTEIN—COUPLED RECEPTORS:

FUNCTIONAL

AND

THERAPEUTIC

IMPLICATIONS,

Brinda K. Rana, Tetsuo Shiina, and Paul A. Insel

593

DRUG TREATMENT EFFECTS ON DISEASE PROGRESSION, P. L. S. Chan and N. H. G. Holford

625

PROSTANOID RECEPTORS: SUBTYPES AND SIGNALING, Richard M. Breyer, Carey K. Bagdassarian, ScottA.Myers, and Matthew D. Breyer

661

PHARMACOLOGY

OF THE LOWER URINARY TRACT, William C. de Groat

and Naoki Yoshimura

ROLE OF OSTEOPONTIN IN CELLULAR SIGNALING AND TOXICANT INJURY, David T. Denhardt, Cecilia M. Giachelli, and Susan R. Rittling

691 732

COMPARTMENTATION OF G PROTEIN—COUPLED SIGNALING PATHWAYS

IN CARDIAC MYOCYTES, Susan F. Steinberg and Laurence L. Brunton

WW

MOLECULAR APPROACH TO ADENOSINE RECEPTORS: RECEPTOR-MEDIATED MECHANISMS OF TISSUE PROTECTION, J. Linden

115

MOLECULAR TARGETS OF LITHIUM ACTION, Christopher J. Phiel and Peter S. Klein

789

CONTENTS

MOLECULAR

BASIS

OF ETHNIC

DIFFERENCES

IN DRUG

DISPOSITION

AND RESPONSE, Hong-Guang Xie, Richard B. Kim, Alastair J. J. Wood, and C. Michael Stein ENDOTHELIN

SYSTEM:

THE DOUBLE-EDGED

SWORD

IN HEALTH

DISEASE, Rafal M. Kedzierski and Masashi Yanagisawa NEUROKININ

RECEPTOR

ANTIDEPRESSANTS,

ANTAGONISTS

Vii

815

AND

85]

AS POTENTIAL

Steven C. Stout, Michael J. Owens, and

Charles B. Nemeroff

877

INDEXES

Subject Index Cumulative Index of Contributing Authors, Volumes 37-41

Cumulative Index of Chapter Titles, Volumes 37—41

907 923 926

RELATED ARTICLES From the Annual Review of Biochemistry, Volume 69, 2000 GTPase-Activating Proteins for Heterotrimeric G Proteins: Regulators of G Protein Signaling (RGS) and RGS-Like Proteins, Elliott M. Ross and Thomas M. Wilkie

Cyclooxygenases:

Structural,

Cellular, and Molecular Biology, William L.

Smith, David L. DeWitt, and R. Michael Garavito

Apoptosis Signaling, Andreas Strasser, Liam O’Connor, and Vishva M. Dixit From the Annual Review of Neuroscience, Volume 23, 2000 Apoptosis in Neural Development and Disease, Deepak Nijhawan, Narimon Honarpour, and Xiaodong Wang

Gain of Function Mutants: Ion Channels and G Protein-Coupled Receptors, Henry A. Lester and Andreas Karschin

Glutamine Repeats and Neurodegeneration, Huda Y. Zoghbi and Harry T. Orr Dopaminergic Modulation of Neuronal Excitability in the Striatum and Nucleus Accumbens, Saleem M. Nicola, D. James Surmeier, and Robert C. Malenka

From the Annual Review of Physiology, Volume 63, 2001 Cellular Mechanisms of Oxygen Sensing, Jose Lopez-Barneo, Ricardo Pardal, and Patricia Ortega-Saenz Role of Estrogen Receptor Beta in Estrogen Action, Katarina Pettersson and Jan-Ake Gustafsson

Viii

ANNUAL REVIEWS is a nonprofit scientific publisher established to promote the advancement of the sciences. Beginning in 1932 with the Annual Review of Biochemistry, the Company has pursued as its principal function the publication of high-quality, reasonably priced Annual Review volumes. The volumes are organized by Editors and Editorial Committees who invite qualified authors to contribute critical articles reviewing significant developments within each major discipline. The Editor-in-Chief invites those interested in serving as future Editorial Committee members to communicate directly with him. Annual Reviews is administered by a Board of Directors, whose members serve without

compensation. 2001 Board of Directors, Annual Reviews Richard N. Zare, Chairman of Annual Reviews

Marguerite Blake Wilbur, Professor of Chemistry, Stanford University John I. Brauman, J. G. Jackson—C. J. Wood Professor of Chemistry, Stanford University Peter F. Carpenter, Founder, Mission and Values Institute

W. Maxwell Cowan, Vice President and Chief Scientific Officer, Howard Hughes Medical Institute, Bethesda Sandra M. Faber, Professor of Astronomy and Astronomer at Lick Observatory, University of California at Santa Cruz Eugene Garfield, Publisher, The Scientist Samuel Gubins, President and Editor-in-Chief, Annual Reviews

Daniel E. Koshland, Jr., Professor of Biochemistry, University of California at Berkeley Joshua Lederberg, University Professor, The Rockefeller University Gardner Lindzey, Director Emeritus, Center for Advanced Study in the Behavioral

Sciences, Stanford University Sharon R. Long, Professor of Biological Sciences, Stanford University J. Boyce Nute, President and CEO, Mayfield Publishing Co. Michael Peskin, Professor of Theoretical Physics, Stanford Linear Accelerator Ctr. Harriet A. Zuckerman, Vice President, The Andrew W. Mellon Foundation

Management of Annual Reviews Samuel Gubins, President and Editor-in-Chief

Richard L. Burke, Director of Production Paul J. Calvi, Jr., Director of Information Technology

Steven J. Castro, Chief Financial Officer

John W. Harpster, Director of Sales and Marketing

Annual Reviews of Anthropology

Astronomy and Astrophysics Biochemistry Biomedical Engineering Biophysics and Biomolecular Structure Cell and Developmental Biology Earth and Planetary Sciences Ecology and Systematics Energy and the Environment Entomology

Fluid Mechanics Genetics Genomics and Human Genetics Immunology Materials Science Medicine

Microbiology Neuroscience Nuclear and Particle Science Nutrition Pharmacology and Toxicology

Physical Chemistry

Physiology Phytopathology Plant Physiology and Plant

Molecular Biology Political Science Psychology Public Health Sociology SPECIAL PUBLICATIONS Excitement and Fascination of Science, Vols. 1, 2, 3, and 4

\

a

Annu. Rey. Pharmacol. Toxicol. 2001. 41;1—21 Copyright © 2001 by Annual Reviews. All rights reserved

ToxICOLOGY COMES OF AGE John Doull Department of Pharmacology, Toxicology, and Therapeutics, University of Kansas Medical Center, Kansas City, Kansas 66160; e-mail: [email protected]

Key Words

history, Tox Lab, SOT

M@ Abstract This paper contains recollections of some of the people and events that influenced the development of toxicology as an academic discipline. It also describes my experiences in pharmacology at the University of Chicago and the University of Kansas Medical Center and concludes with speculation concerning the future of toxicology.

Moderation in all things/Ne quid nimis. —Terence in Andria

INTRODUCTION Describing the origins of toxicology is somewhat controversial because there are those who argue that toxicology is one of the oldest of all sciences because its roots go back to when primitive man tested the safety of his food by giving it to his dog. Toxic substances have certainly been used for both offense and defense ever since our advent on this earth, but the earliest written records were probably the Ebers Papyrus and Hippocrates’ description of hemlock as the state poison of the Greeks. Both the Romans and the Arabic physicians contributed to the science of toxicology, and our knowledge of poisoning and toxicology mushroomed during the Middle Ages and the Renaissance. There are others, however, who link the

beginnings of toxicology, or at least modern toxicology, to Paracelsus, who established toxicology as a quantitative science, or to Orfila, who introduced qualitative science into toxicology. It is much easier to define toxicology’s coming of age or the transition of toxicology into an academic discipline with journals, textbooks, departments, curricula, societies, annual meetings, certification, etc. This is be-

cause the origins of these developments are well documented and there are still some survivors with memories from the 1950s and 1960s, when many of these events took place. Most of the key players in the development of toxicology as an academic discipline were pharmacologists, and two of these, Dr. Eugene Maxmillian Karl Geiling and Dr. Kenneth Patrick DuBois, were from the University of Chicago. Those of us who were graduate students in pharmacology at Chicago at that time, 0362-1642/01/0421-0001$14.00

1

2

DOULL

shared in the excitement of these developments in toxicology. This was particularly true for those students who also worked behind the so-called “bamboo curtain” in the University of Chicago Toxicity Laboratory. ‘The following pages contain my recollections of some of the people and events that have influenced the development of toxicology as an academic discipline. They also include a description of my experiences in the pharmacology departments at the University of Chicago and the University of Kansas Medical Center, and they conclude with some speculation about the future of our discipline.

CHICAGO AND THE TOX LAB The University of Chicago Toxicity Laboratory, or the “Tox Lab” as it was commonly called, was created in 1941 to evaluate potential chemical warfare agents that were synthesized by chemists participating in the National Defense Research Council Office of Scientific Research and Development program. The University of Chicago was selected as the site for this laboratory in part because it had a very tall smokestack that was no longer in use and could be used to ventilate the laboratory (1). Dr. EMK Geiling served as the official investigator, and Dr. Franklin C McLean was appointed as the first director. In 1943 McLean resigned to accept a commission in the US Army Chemical Warfare Service, and Dr. Keith Cannan from New York University became the new director. Dr. William Bloom from the University of Chicago anatomy department was placed in charge of the programs to test agents with skin-blistering properties and to develop protective ointments. By this time the staff had increased to over 50 professionals. In addition to the program to evaluate potential chemical warfare agents there were programs under way to develop new nitrogen mustards and to evaluate these and other vesicants in recruits from the Great Lakes Naval Training Station. There were also programs to study the retention and distribution of various types of aerosols in rodents and primates and to study the toxicity of the nerve gases and related insecticides and other pesticides, such as alpha naphyl thiourea (ANTU), sodium fluoracetate [1080], and castrix. In 1945 the Tox Lab became a part of the US Army Chemical Warfare Service, and Dr. William Doyle and Dr. John Hutchens served as interim directors

until Dr. George Mangun was recruited for this position in 1946. Between 1947 and 1950 the Tox Lab operated under a contract with the Atomic Energy Commission to study the toxicity of radioactive metals and the medical effects of ionizing radiation. The name of the lab was changed in 1951 to the University of Chicago US Air Force Radiation Laboratory to reflect a new research program and source of funding, and Dr. Julius Coon became the director. Dr. Kenneth DuBois became the director in 1953 and he served until his death in 1973. The laboratory closed shortly thereafter. During the three decades of its existence, the Tox Lab was a research home for clinicians such as Drs. Erwin Levin, Leon Jacobsen, Charles Spurr, John Rust, Robert Block, and William Adams; anatomists

William Bloom and William Doyle; pathologists Clarence Lushbaugh, John Storer,

TOXICOLOGY COMES OF AGE

3

Frank Fitch, Janet Rowley, and Draga Vesselinovitch; physiologists Dan Oldfield, Walter Stumpf, Herb Landahl, John Hutchens, and Franklin McLean; biochemists George Mangun, Keith Cannan, Robert Feinstein, and Jean Sice; and pharmacologists John Thompson, Dick Byerrum, Julius Coon, Ken DuBois, and John

Doull. During this period the laboratory also employed graduate and postdoctoral students in all of these disciplines; over four hundred publications attest to the productivity of this group. I came to the Tox Lab in the fall of 1946, after receiving a BS in chemistry from Montana State College in 1944 and then spending two years in the navy taking care of radar and other electronics on the battleship New Jersey in the South Pacific. Dr. BL Johnson, who was one of my professors at Montana State, had arranged an interview with Dr. George Mangun, a former Montana State graduate, who was then the director of the University of Chicago Toxicity Laboratory and a professor of biochemistry. Although I was admitted as a graduate student in biochemistry, Dr. Mangun suggested that I switch to pharmacology and work with Dr. DuBois, who became my graduate advisor, coileague, and good friend. My PhD thesis described the cardiotoxic and other effects of bufagin that we obtained from the parotid gland of the giant toad Bufo Marinus and labeled by feeding the toads '4C-labeled algae. Most of my research during this period, however, involved the acute and chronic toxicity of the organophosphate (OP) insecticides. During the war di-isopropy] fluorophosphate had been investigated as a potential chemical warfare gas. Although it was discarded as a war gas, it was shown to inhibit cholinesterase and was therefore evaluated for use in glaucoma and in myasthenia gravis. Thus, when the first organophosphate HETP (hexaethy] tetra phosphate) emerged from Dr. Gerhard Schrader’s laboratory at Farbenfabriken, Germany in the early 1940s, DuBois and his associates recognized its cholinergic symptoms and showed that atropine would be an effective antidote. They also carried out similar early studies with tetraethyl pyrophosphate (TEPP), parathion (E-605), and later on with the nerve gases (tabun, sarin, and soman) when they

became available. These studies stimulated a life-long interest for DuBois in the toxicity of the OP insecticides, and his students and associates shared his interest.

Sheldon Murphy and DuBois, for example, elucidated the biochemical basis for potentiation of the effects of malathion and other OP insecticides, and Robert Neal

and DuBois identified several enzymes that catalyzed the detoxification of some of the OP insecticides. During my graduate period with Ken DuBois we characterized both the acute and chronic toxicity of several OP insecticides and a few cholinergic carbamates in rodents and beagles. Most of these agents were developed for agricultural use, and we worked closely with Drs. Dan MacDougall and Dallas Nelson from Chemagro (now Bayer, Inc.) in Kansas City in planning and carrying out the studies and eventually in defending them before the Food and Drug Administration (FDA). These meetings were usually held in the FDA commissioner’s office, with Drs. Arnold Lehman, Garth Fitzhugh, Bert Vos, and Arthur Nelson representing the FDA, and DuBois, Doull, and McDougall representing Chemagro. In contrast to the complex and lengthy procedure currently required to obtain pesticide

DOULL

tolerances, these meetings were short, informal, and focused on the science (tox-

icology and pathology) rather than on any of the legal or political considerations that often seem to be of primary importance today. It was in this office that the often-quoted admonition was displayed on the wall, “You too can learn toxicology in two easy lessons, each ten years long.” After receiving my pharmacology PhD in 1951 I enrolled in the University of Chicago Medical School but continued with the Tox Lab as a research associate working mainly on the toxicity of pesticides and radiation. When the US Air Force took over the Tox Lab in 1951, Vivian Plzak, Mildred Root, and I were requested to establish a screening program for radio protective agents. This program involved obtaining acute IP LD/SOs on several thousand agents in mice. The resulting large data base of acute toxicity data in male mice has subsequently proven to be of more lasting value than the few radio-protectors we found. After finishing medical school in 1953 I became the assistant director of the Tox Lab in 1954 and began advancing through the academic ranks as an assistant professor of pharmacology. In looking back, these were indeed golden years. The university and the deans of biological sciences supported our program, grants were relatively easy to obtain, toxicology was becoming established as a discipline, and both our department and the staff and students at the Tox Lab were involved in the conversion process. One of the most important accomplishments of the University of Chicago Toxicity Laboratory is the number of PhD and postdoctoral students who were trained during its 30 years. These include Drs. John Ballin, Robert Bagdon, Jules Brodeur, Gary Carlson, Kenneth Cochran, Bernard Conley, John Doull, Marion Ehrich, Bernard Heitbrink, Roy Herrman, Eugene Kimura, Florence Kinoshita, Hugo Moeller, Albert Moraczewski, Sheldon Murphy, Robert Neal, John Noble, Arthur Okinaka, Donald Petersen, Alfred Rider, Tetsuo Satoh, Mei Su, Maurice Sullivan, Robert Tardiff, Don Thursh,

John Thompson, Edwin Uyeki,

James Wilson, Dinah

Wu, Kei Yam, and Gerald Zins, plus numerous other masters and nondegree students. Both student and staff members of the Tox Lab presented toxicology lectures to medical students and graduate students in other departments and postgraduate programs. Dr. Geiling treated the staff and students in the Tox Lab and those in the pharmacology department like his family, and holiday dinners at his home were mandatory occasions. He took a personal interest in each student, technician, and faculty member, and his Christmas gift to each was always a book he chose carefully » to relate to the specific interests of the recipient. Dr. Geiling never married, and although he tolerated marriage in his faculty and students, he regarded it generally as a detrimental or counterproductive influence. (Neither Ken DuBois nor I got married until after Dr. Geiling retired.) Dr. Geiling made all decisions regarding faculty promotions, and each year when he gave his recommendations for promotion to the dean, he included his own letter of resignation in the event that the dean failed to follow any of his recommendations. It was also traditional for Dr. Geiling to deliver serious good or bad news after a one-on-one breakfast at the University Club. Because one never knew whether the news was good or bad, it was

TOXICOLOGY COMES OF AGE

5

impossible to eat until it was too late. We discussed my going to medical school at one such breakfast, and when I indicated that I would like to think about it, he sug-

gested that I do so quickly because I was already enrolled in gross anatomy starting the following week. When one compares this patriarchal approach to chairing a department to the more opportunistic approach exhibited today by most faculty, chairs, and even administrators, it is not surprising that institutional commitment is becoming a rare and vanishing trait in most universities. Dr. Geiling’s approach was more

in the nature of a benevolent advisor, and like most of his faculty I

anticipated that Chicago would be my permanent academic home. Dr. Geiling retired in 1957 and was replaced by Dr. Lloyd Roth. During the next decade, the environment at the University of Chicago and elsewhere in this country changed as applied research lost favor. The support of Dr. Leon Jacobsen, the Dean of Biological Sciences, shifted from studies that focused on whole animal responses to those that focused on DNA. By the end of 1967, Dr. Gail Dack was planning to move Chicago’s Meat Institute to the University of Wisconsin at Madison, Dr. DuBois was considering closing the Tox Lab and returning to the McArdle Lab at Wisconsin, and I had accepted Dr. Edward Walaszek’s invitation

to become a professor of pharmacology and toxicology at the University of Kansas. Years later when I returned to the University of Chicago to accept a Distinguished Medical Alumnus Award, Dr. Jacobson told me that in retrospect he felt he probably made the wrong decision for Chicago regarding the merits of whole animal studies.

JOURNALS Historically, the first step in the transition of toxicology into an academic discipline was the establishment of the Journal of Toxicology and Applied Pharmacology in 1959. The editors were Drs. Fred Coulston and Arnold J. Lehman, and Dr. Harry

Hays was the managing editor. An editorial in the first issue stated that the journal was needed because “toxicology is emerging as a scientific discipline and that there is a need for centralization of toxicology data.” The reluctance of the The Journal of Pharmacology and Experimental Therapeutics to publish toxicology studies on products or chernicals was not mentioned, although it was widely recognized and was certainly one of the reasons for creating the new toxicology journal. Dr. DuBois took over as the managing editor of the journal in 1960, and Dr. Coulston became the managing editor in 1964, with Drs. Horace Gerarde and

Arnold Lehman as associate editors. Dr. Boyd Shafer became the editor in 1965, replaced by Dr. Earl Dearborn in 1967. Dr. Gabriel Plaa, who served as the associate editor for Dearborn became the editor in 1973. In 1980 Dr. Robert Neal became the editor and was succeeded by Dr. Wallace Hayes in 1981. Dr. Glenn Sipes became the editor in 1986, and Dr. Edward Bresnick, the current editor, took

over in 1993. Although the Journal of Toxicology and Applied Pharmacology became the official journal of the Society of Toxicology in 1963, it was not owned by the

6

DOULL

society. After long deliberation, the society established a second journal, Fundamental and Applied Toxicology, in 1981, with Drs. William Carlton and Phil Watanabe as the editors. Dr. Bern Schwetz became the editor of this new journal in 1986, and from 1992 to 1998, the editor was Dr. Henry d’A Heck. The name of

the journal was changed to Toxicological Sciences in 1998; Dr. Curtis Klaassen became the new editor and a new editorial board was appointed. They broadened the scope of the journal to include reviews, editorials, and coverage of contemporary issues in toxicology. More recently the journal has added a section titled Profiles in Toxicology, edited by Dr. Hans Peter Witchi, which is designed to call attention to historical events and contributors to toxicology. Today there are many journals devoted entirely or in part to toxicology, but the two journals described above were historically important factors in the transition of toxicology into an academic discipline. Furthermore, it is likely that they will continue to significantly influence the direction and development of toxicology as a science and as an academic discipline.

TEXTBOOKS During the 1950s many departments of pharmacology began to introduce courses in toxicology into their graduate curriculum and toxicologic materials into the medical school courses. Because there was no textbook for this area of toxicology, DuBois and Geiling wrote and published the first textbook of toxicology in this

country in 1959. During the late-1960s DuBois was working on a second edition of the book, but died before it was completed. A similar textbook was published by Ted Loomis in 1968; subsequent editions appeared in 1974 and 1978. The fourth edition of this book was published in 1996 as Loomis’s Essentials of Toxicology, with Wally Hayes as a coeditor. In 1967 Lou Casarrett was searching for a toxicology textbook that would include not only the classes of toxic agents (metals, solvents, pesticides, etc) but also the organ system involved (kidney, liver, etc). Because neither the DuBois or Loomis texts provided this approach, Lou asked me to help him put such a textbook together. I agreed and we discussed this further during a dinner meeting after a National Institutes of Health Toxicology Study Section meeting in Bethesda. We recruited study section members who endorsed the plan to write chapters, and Lou then approached MacMillan to be our publisher. During a subsequent summer vacation in Hawaii with my family, Lou and I selected additional contributors and planned the organization of the book. Tragically, as the first chapters were being received, Lou developed brain cancer. With the perseverance of his wife Peggy, who read the chapters to Lou and handled the correspondence with the authors and with the enthusiastic and capable support of the contributors, friends, and colleagues, we finished and published Toxicology, the Basic Science of Poisons in 1975. Drs. Curt Klaassen and Mary Amdur agreed to join me as editors for the second edition, and Mary Amdur suggested that we change the name of the book to Casarrett and Doull’s Toxicology to provide a fitting memorial to Lou’s dedication

TOXICOLOGY COMES OF AGE

uy

to toxicologic education. The second edition was published in 1980, and third and fourth editions were published in 1986 and 1991, with the listing of the editors

arranged to reflect which of us had the major responsibility for getting the book to press. Mary Amdur and I became emeritus editors for the fifth edition, which was published in 1996 with Curt Klaassen as the editor. Our textbook has been widely used in graduate courses of toxicology and is the basis for curricular design in such programs both in this country and abroad. It is widely used as a reference in other disciplines and is the choice of most students preparing for the certification exams in toxicology. Although Lou and I had high expectations for our book, neither of us could have anticipated that Casaret and Doull would become a major factor in the development of toxicology as an academic discipline and that it would persist as a virtual landmark in academic toxicology for over two decades and five editions.

SOCIETIES The formation of the Society of Toxicology (SOT) was the third major event in the history of the transformation of toxicology into an academic discipline. It started in early 1961 with a discussion involving Drs. Fred Coulston, Victor Drill, William

Deichman, Harry Hays, Harold Hodge, Arnold Lehman, and Boyd Schafer, with Drs. Ken DuBois and Paul Larson connected by phone. These nine individuals became the founders of the SOT. Although they were concerned about deleterious effects on the Society of Pharmacology, the group voted to form the SOT and elected Lehman as honorary president, Hodge as president, and DuBois as vice president. They planned for meetings to be held in conjunction with meetings of Federation of American Societies for Experimental Biology, American Industrial Hygiene Association, American Society for Pharmacology & Experimental Therapeutics (fall meeting), and the Gordon Conference to elicit support. All members who joined during the first year became charter members, and there were 183 of us by the time of the first annual meeting. The first annual meeting took place in Atlantic City in 1962, during which Drs. Torald Sollman, Wolfgang von Oettingen, and EMK Geiling became the first honorary members. Dr. Geiling was a strong supporter of the new society and he provided advice based on his association with Dr. JJ Abel when Abel was attempting to separate pharmacology from physiology and biochemistry. Geiling first recommended that we focus on the unique aspects of the new discipline to sharply separate it from the old, and second to clearly identify the societal benefits the new discipline would provide. He noted that describing pharmacology as the science of drugs separated it from physiology and biochemistry, and that including therapeutics as part of the discipline identified the rationale for the societal support of pharmacology. He suggested defining toxicology as the science of poisons to separate it from pharmacology and advised including safety evaluation as the justification for support by the public. When we define toxicology simply as the adverse effects

8

DOULL

of chemicals on living systems without including the use of that information to evaluate safety or predict risk we describe what we do but not why we do it. If our discipline focuses on this limited mission, we risk eroding public support for toxicology, the regulatory process, and science in general. In a previous prefatory chapter Dr. Avram Goldstein voiced similar concerns about the growing failure of pharmacology to focus on the unique attributes of the discipline rather than on the techniques of molecular biology (2), which suggests that toxicology and pharmacology face some common problems. During its early history the SOT was closely associated with both pharmacology and industrial hygiene. Dr. William McCormack, the executive secretary for the American Industrial Hygiene Association, also served as the executive secretary for SOT until 1985 when we moved the office from Akron, Ohio to Washington, DC and hired Joan Cassidy as our new executive secretary. In 1986 we celebrated our silver anniversary in New Orleans with seven of the original nine founders in attendance (DuBois and Lehman had died). My first committee appointment with SOT was made on October 4, 1961 as a member of the technical committee, and this led to many other SOT committee appointments. My most prestigious, and also most enjoyable, task for SOT was serving as president from 1986 to 1987, and I later served as president of our Toxicology Education Foundation from 1998 to 1999. In the nearly four decades of SOT’s existence, it has exhibited remarkable progress, its annual meetings have become major scientific events, it has shaped and fostered the development of toxicology as a science, and it has been a key element in the transition of toxicology into an academic discipline.

CERTIFICATION During the 1960s and 1970s there were a few episodes in which the FDA received erroneous or even fraudulent reports on toxicity studies, and the agency responded by issuing guidelines for “Good Laboratory Practice.” Because of concern over the possibility that the agency might also define the qualifications for those who carried out such studies, the SOT council asked Dr. Fred Oehme to convene a group to explore certification for testing laboratories and toxicologists. In response to their recommendations and recognizing that such groups needed to be totally. independent, the SOT council provided start-up funding, solicited nominees from SOT as well as other existing boards (forensic toxicology, veterinary toxicology and pathology, and occupational medicine), appointed a start-up board, and then cut the ties. The laboratory certification program was initially quite active, but demand for this gradually decreased and the program was essentially phased out by the end of the decade. The toxicologist certification program was incorporated in 1979 as the American Board of Toxicology (ABT); the initial board members were Drs. Bert Dinman, John Doull, Bob Forney, Seymour Friess, Bert Koestner, Marv Kuschner,

TOXICOLOGY COMES OF AGE

9

Gene Paynter, Bill Rinehart, and Ray Suskind. Friess was elected president with Doull as vice president, Paynter as secretary, and Koestner as treasurer. The board

decided that start-up board members would not be grandfathered into the organization and that board members would be required to wait for three years after completion of their term before applying for certification. Committees were appointed to evaluate the qualifications of the candidates and to prepare examinations that could be used by candidates for self-assessment and for certification. The eligibility committee required proof of both educational background and current involvement in the practice of toxicology, and their recommendations were reviewed by the full board. The exam committee relied heavily on a toxicology question data bank containing questions from Chicago and Kansas, and these were classified into three areas (general toxicology, special toxicology, and applied toxicology). Selected questions in each category were modified if necessary, additional questions were developed, and the sample and candidate exams were generated using a random numbering selection system. Finished exams were reviewed by the full board, and on August 4, 1980, the first certification examination was given simultaneously in Washington, DC; Indianapolis, Indiana; San Francisco; and London.

There were

460 candidates taking the exam; 217 of these were certified in all 3 parts (general,

special, and applied toxicology). Candidates who failed only one part of the exam were permitted to correct the deficiency when the exam was given the following year, but candidates failing two parts were required to retake all three parts of the examination. A total of 1751 candidates have become diplomates of ABT during the past two decades. To provide for an orderly replacement of the start-up board by diplomates, it was agreed that three new members would be elected each year. In 1981, Drs. James Beall, John Moore, and Fred Oehme were elected, and Dr. Charles Reinhardt re-

placed Kuschner, who resigned. Doull became the president in 1982, with Charley Reinhardt

as vice president and Beall as secretary and Drs. Mike

McKenna,

Paul Newberne, Robert Scala, and TJ Terhaar were elected to replace Dinman, Forney, Friess, and Paynter. In 1983, Charley Reinhardt became the president, Scala became the vice president, and Moore became the secretary. Sue Moore was appointed as administrative assistant to handle the increasing correspondence and administrative duties of ABT, and Drs. Jim Bus, Robert Drew, Bernard Goldstein, and Mark Hite were elected to replace the four remaining members of the start-up board (Doull, Koestner, Suskind, and Bill Rinehart). The following year

the board, under the direction of the new president, Dr. Robert Scala, appointed Sue Moore to be the Executive Director of ABT, and she has continued to serve

in this position to the present. The Academy of Toxicological Sciences was established at about the same time as ABT. Its purpose is to honor and certify toxicologists who have achieved peer recognition for their expertise and sound scientific judgment. The criteria for certification in the Academy of Toxicological Sciences include formal training, proven ability, and experience; selection is by peer review involving the entire board of directors, and certified experts are awarded the title of Fellow. There are

10

DOULL currently more than 176 fellows, although 24 of these are now emeriti. Dr. Donald

Gardner is the current president with Drs. John Thomas and David Brusick as vice president and secretary. In 1997, 1998, and'1999 the Academy awarded two

student travel awards to the American College of Toxicology meeting and in 1999, a $1000 travel grant to the British Toxicology Society. The Academy will also sponsor lectures at the European Toxicology Society meeting in 2000 and at the 9th International Congress of Toxicology in 2001. Although there have been several attempts over the years to consolidate the ABT and the Academy of Toxicological Sciences, these efforts were not able to resolve the issue of the examination requirement. However, the proposed formation of a new International Assembly for the Recognition of Toxicologists, which would be open to all organizations that recognize toxicologic expertise, may ultimately diffuse or resolve the issue.

THE KANSAS YEARS My family (wife Vera, daughter Ellen and twin sons John and James) and I arrived in Kansas in the spring of 1967. Leaving Chicago, which was still suffering the aftermath of a record snowfall, and arriving with the flowers blooming seemed like a good omen. We were surprised and delighted to find that the Kansas City landscape was not flat but gently rolling. I was also surprised to learn that Kansas has a strong historical association with toxicology. In 1884 a chemistry professor, Dr. Edgar Henry Summerfield Bailey, was teaching toxicology to medical students in Kansas while JJ Abel was still a graduate student in physiology at Johns Hopkins (3). Although Bailey’s course was called pharmacology, the catalog described the course as “including a discussion of the source, properties, methods of detection, post mortem appearances, fatal doses and methods of treatment for inorganic and organic poisons.” Bailey was also ahead of his time as an environmental activist. He analyzed many samples of food, water, and various remedies of that time to detect adulteration. Thus, if we follow the example of naming Abel as the father of pharmacology because of his early contributions to pharmacology, we might consider a similar argument for EHS Bailey as the father of academic toxicology and Lawrence, Kansas as its birthplace. Shortly after my arrival in Kansas, Dr. Daniel Azarnoff and I applied to NIH for a Center in Clinical Pharmacology and Toxicology. When the grant was awarded in 1968, we recruited Dr. Aryeh Hurwitz, who had gone to NIH after getting his MD at Washington University in St. Louis, and Dr. Curtis Klaassen, who had just finished his PhD with Dr. Gabriel Plaa at the University of Iowa. Because we were short of space, we negotiated a provision whereby the center grant could pay rent to the Endowment Association, who then agreed to erect a building to house our new center. During the next decade we focused on developing a pharm/tox research program and on training graduate and medical students and residents. During the second five-year renewal of the center, Dr. Azarnoff left to become

TOXICOLOGY COMES OF AGE

11

the research director of GD Searle and no further renewals were submitted. The clinical pharmacology program has continued, however, as an active research and training division in the Department of Medicine, with Dr. Aryeh Hurwitz as the clinical director. The toxicology program at Kansas gained momentum in 1979 when Dr. Curtis Klaassen obtained a National Institute of Environmental Health Sciences toxicology training grant (which has now been active for over 20 years). In 1981 Dr. Karl Rozman, who graduated from Leopold Franzen’s University in Innsbruck, Austria, was recruited in part because of Klaassen’s sabbatical with Dr. Helmut Greim at the GSF Institut ftir Toxikologie in Munich, Germany. In 1982 Klaassen received the Burroughs Wellcome Toxicology Scholar Award, and as a result salary funds became available to hire Dr. Andrew Parkinson, who had trained with

Drs. Stephen Safe and Allen Conney.

During this period Dr. Edwin Uyeki who

trained at Chicago with DuBois, Dr. Tom Pazdernik who trained at Kansas with

Dr. Ed Smissman in medicinal chemistry, and Dr. Stata Norton who had previously managed toxicology programs for the Burroughs Wellcome company changed the focus of their research from pharmacology io toxicology. Our department received permission to grant a doctorate in toxicology rather than in pharmacology, and the department changed its name from Pharmacology to Pharmacology, Toxicology, and Therapeutics. This was actually a transcription of a previous name adopted by the department in 1909 but abandoned in 1926. Largely as a result of these advances, the State of Kansas provided funds in 1986 for an Environmental and Occupational Health Center, and I became the director

with Dr. Klaassen as the associate director. Funds were provided for equipment and salaries. We used the equipment money to establish a common instrument room, and we hired Drs. Gregory Reed and David Beer with the salary funds. Reed had worked with Dr. Larry Marnett at Wayne State University and with Dr. Tom Ealing at NIEHS, and Beer worked with Dr. AM Malkinson at the University of Colorado and had postdoctoral training in the laboratory of Dr. Henry Pitot at Wisconsin. In 1989 I developed kidney cancer and resigned as the center director. Dr. Curt Klaassen served as interim director until 1991 when Dr. H William Barkman was recruited from the pulmonary medicine division at the Tulane University Medical School. During his tenure Dr. Barkman has strengthened the links between the center and the clinical departments, between the Schooi of Veterinary Medicine and other departments at Kansas State University and similar programs in lowa. He has also organized several joint regional conferences on Agro Medicine. Dr. Thomas Pierce has been recruited to provide industrial hygiene expertise for the center and to set up a laboratory to measure environmental levels of metals and other xenobiotics. Dr. Salvatore Enna became the chairman of the department in 1992, and although he has recruited toxicologists as faculty, these additions have not replaced the loss of David Beer, the retirement of Ed Uyeki, and the emeritus status of Stata Norton and John Doull. Thus, toxicology will probably retain a strong

12

DOULL

presence in this department in the coming years, but it does not seem poised to regain the prominence it exhibited during the past two decades. Another program developed during my early years at Kansas was the computerized teaching system known as Computer Assisted Teaching System (CATS). Together with Drs. Stata Norton, Stan Nelson, and Ed Walaszek, we developed an interactive computer-based program for our medical school pharmacology course. We also developed an extensive question bank in pharmacology and toxicology so that we could generate examinations almost on demand to respond to student requests, and we developed an exam scoring system and grade book program to manage the entire system. Because other medical schools were interested in sharing in the development and use of this system, we created a consortium of CATS users in 1974 and began holding annual meetings to encourage growth of © the program. During the early 1980s, over 60 medical and dental schools plus some graduate schools used all or part of the 166 interactive teaching programs that were available in CATS. Through the cooperation of our partners in CATS, there were over 26,000 questions in the exam database, although only half of these were included in the preferred file. With the development of similar commercial programs in pharmacology and other areas during the early 1980s, interest in our programs declined, although the grading program is still used by the ABT and pharmacology courses at the University of Kansas Medical Center. During the past few years, interest in these teaching and grading programs has been rekindled by Dr. Lazlo Kerecsen, a former staff member who is teaching pharmacology at the Glendale Campus of the Midwestern

University in Phoenix. Dr. Kerecsen,

in cooperation with Dr. Tom Pazdernik, has updated the existing course material, added new programs and converted all of the computer programs into a more powerful and contemporary language. While I was at Chicago we developed a database that could be used in the emergency room to diagnose and treat poisoning. Like most systems of that time, it consisted of a card file, wall charts on snakes and plants, and a few books on clinical toxicology. When I was asked to establish a poison control center at the University of Kansas Medical Center, I duplicated this system. However, when Dr. Barry Rumack introduced a computer-based system for generating poison control information at Denver General Hospital, we were anxious to implement the system in our medical center. To facilitate this, I joined the newly created Academy of Clinical Toxicology and served on its board during the period when building the computer database was an Academy-related activity and continued on as a consultant when the database was taken over by Micromedex. During this period we also instituted a series of quarterly lectures in toxicology for each new group of incoming interns and residents in our emergency room and we provided them with consultant support as needed. We also identified regional experts who could be consulted for poisoning by plants, snakes, spiders, and other local problems. An ancillary task was to oversee the University of Kansas Medical Center Safety Office and ensure compliance with the requirements of the Joint Commission regarding electrical, chemical, and radiation safety. With the

TOXICOLOGY COMES OF AGE

13

help of Ruth Schukman-Dakotas, who now serves as the director of the Medical

Center Safety Office, we increased the staff in industrial hygiene and radiation safety and expanded the services provided to the hospital and basic science areas. My first association with the Environmental Protection Agency (EPA) came in 1976 as a member of the Federal Insecticide Fungicide and Rodenticide Act Science Advisory Panel. Unlike the Science Advisory Board, which reported to the administrator, our panel was linked to Congress, and the EPA administrator was

required to respond formally to recommendations of our panel. With members such as Drs. Edward Smukler, Robert Metcalf, Robert Neal, and Christopher Wilkinson,

this panel was a strong advocate for basing decisions on good science rather than on policy or politics. My subsequent service on the Environmental Health Committee of the Science Advisory Board, the Dioxin Reassessment

Review

Committee,

and other EPA committees has been rewarding, but the growing resurgence of policy rather than good science as the basis for many decisions in recent years is disappointing. The three professional groups I have most enjoyed working with are the Expert Panel of the Flavor Extract Manufacturing Association (FEMA), the National Academy of Sciences/National Research Council’s Committee on Toxicology (COT), and the Threshold Limit Value (TLV) committee of the American Confer-

ence of Governmental Industrial Hygienists. The reason for this is that these three groups and the Food and Drug Administration are among the leading proponents of the first and most venerable principle of toxicology, which is that “the dose makes the poison.” The FEMA expert panel was established following passage of the 1958 Food Additives Amendment in which the concepts of preclearance and safety in use were introduced. This led to the concept of GRAS, which stands for generally recognized as safe by experts qualified by scientific training and experience to evaluate a substance’s safety under specified conditions of use. Beginning in 1960, this approach was used by FDA and by a FEMA-sponsored expert panel appointed by Dr. Bernard Oser to classify the safety of food additives. Dr. Oser served as the nonvoting chair of this panel, which included Drs. David Fassett, Horace Gerard, Maurice Seevers, Howard Spenser, Jakob Stekol, and Lauren Woods.

Over the

next few years, Drs. Frank Blood, Frank Strong, Anthony Ambrose, and R Tecwyn

Williams were appointed as new or replacement members. I joined the panel in 1977 along with Paul Newberne and Carrol Weil. The current members of the FEMA expert panel are Drs. Victor Feron, Jay Goodman, Larry Marnett, Ian Munro, Phillip Portghese, Robert Smith, William Waddell, and Bernard Wagner. During the four decades of the panel’s existence, several thousand flavors have been evaluated, over 2000 have been GRASed, and a few have been deGRASed.

The criteria used by the panel include (a) exposure to the flavor in specific foods, the total amount in the diet and the total poundage; (b) natural occurrence in food;

(c) chemical identity; (d) metabolic and pharmacokinetic characteristics; and (e) animal toxicity (4). In recent years, the panel has become increasingly involved in global certification issues, in developing protocols for assessing the safety of

14

DOULL

natural flavors and other mixtures, and in the utilization and role of cutting-edge toxicologic techniques and procedures in safety evaluation. The COT is one of the oldest of the National Academy of Science/National Research Council committees. It has provided military and civilian agencies with scientific information and expert advice on issues involving toxicology and the health effects of hazardous substances since 1947. To celebrate the “First 50 Years,” the COT recently held a colloquium in which previous members described contributions of the COT and offered predictions for the future (5). Many of the issues

considered by the COT over the years have been extremely controversial. Thus, the ability of the many COT committees to deal with these issues in a balanced yet authoritative way while maintaining credibility with the scientific community and various other stakeholders provides a validation of the COT model. The COT, like other National Research Council committees, has learned that the only effective way to manage bias is not to eliminate it but to balance it, because real experts are rarely unbiased. A major responsibility of the COT during its long history has been to provide emergency response advice and guidance to the Armed Forces. The COT director and staff usually provided this service, although COT members were occasionally involved. The executive directors of the COT have included Drs. Harry Hays, Ralph Wands, Gordon Newell, Gary Kielson, Francis Marzulli, and Richard Thomas; the current director is Dr. Kulbir Bakshi. Previous chairs of the COT were Drs. HH Shrenk, Lawrence Fairhall, Harold Hodge, Norton Nelson,

Arnold Lehman,

William Sutton,

Herb Stockinger,

Bertram

Dinman,

Joseph Borzelleca, Roger McClellan, John Doull, Rogene Henderson; the current chair is Dr. Bailus Walker. In addition to the COT committees, I have also served on the Food Protection Committee, the Committee on Mixtures, the Safe Drinking Water Committee, the

Committee on Hazardous Air Pollutants, which produced the book Science and Judgment and others, and am currently serving on the Subcommittee on Acute Exposure Guideline Levels. These experiences have convinced me that there is no substitute for using scientists to resolve science issues and that the academy committees are the closest approach we have to a supreme court for science in this country.

The origin of the TLVs as occupational exposure limits is often attributed to Professor Warren Cook, although Dr. William Fredrick served as the chair of the first American Conference of Governmental Industrial Hygienists TLV committee, . which published an exposure limit list in 1942 (6). In 1945, Cook published a list of maximum allowable concentrations for 132 substances, and the TLV committee used this list plus other sources to develop the first official TLV list in 1946. The concept of excursion limits with short-term exposure limits and ceiling values was introduced in 1961 along with the SKIN and cancer designations, and the first volume of TLV documentations appeared in 1962 when Dr. Herbert Stokinger was the chair. TLVs for physical agents were introduced in 1969, and the biological limit values appeared in 1974. Particle size selective threshold limit values were introduced in 1982, although the distinction between “respirable” and “total” dusts was

TOXICOLOGY COMES OF AGE

15

discussed in the earlier lists. I was appointed as chair of the American Conference of Governmental Industrial Hygienists Chemical Substances TLV Committee in 1990 and with the help and support of Dr. William Wagner as mentor, colleague, and friend, I served through 1998. During that period we focused on improving the documentation for the existing TLVs and developing closer ties with Dr. Helmut Greim and the German maximale Arbeitsplatz Konzentration (MAK) commission and with other groups involved in setting occupational exposure levels. To increase the epidemiologic strength of the TLV committee, we added Drs. Philip Guzelian, William Waddell, and Ian Greaves to the committee and we recruited Drs. Bob

Scala, Karl Rozman, and Roger Smith to strengthen the toxicology input. In addition to stimulating the interest of the committee in kinetics and Structure Activity Relationships, Dr. Rozman served as liaison to the German MAK commission. The major hurdle to the harmonization of TLVs with MAKs or other occupational exposure limits in Europe is the result of differences in our approach to genotoxic agents and carcinogens. With the dedicated but patient efforts of Dr. Helmut Greim and the members of the MAK commission and the TLV committee, we have made

significant progress in resolving these issues. We have also moved closer to the idea of sharing resources in the preparation of common documentation for occupational exposure limits. Because all such documentation is based on the same toxicology and epidemiology databases, different exposure-setting groups or countries could use the same basic documentation and customize the rationale for their individual occupational exposure levels as needed.

THE FUTURE Before speculating on the future of toxicology, I would like to briefly consider its current condition. Toxicology, like medicine, is both a science and an art. The science of toxicology consists of the observational or data gathering phase, and the art of toxicology is the predictive or application phase. In most cases, the science and the art of toxicology are linked because we use the facts provided by the science of toxicology as the basis for the prediction or hypothesis regarding the potential adverse effect of exposure to an agent in a situation where we have limited information. For example, the argument that chloroform is a carcinogen is a fact because it has been shown to cause cancer in rodents, but the prediction that it will also cause cancer in humans is a hypothesis. We now have several hundred agents that are known to cause cancer in rodents, and although some of these would

probably do so in humans, there are actually only a few dozen known human carcinogens. By distinguishing between the science and the art of toxicology, we recognize that the validity of our prediction or hypothesis depends not only on the quality of the science in the toxicology data base but also on the relevance of that information to the species or situation for which we are predicting. When we fail to clearly distinguish between the science and the art of toxicology, we tend to confuse our facts with our hypotheses and argue that they have equal validity,

16

DOULL

which they clearly do not. Separating the science from the art of toxicology can often be helpful in determining whether flaws in our risk assessments are the result of problems in the data gathering process or occur because of how we analyze or interpret the data from the toxicology and epidemiology data bases. During recent years we have witnessed a virtual explosion in the science of toxicology. This has resulted primarily from the contributions of molecular biology and transgenic technology (we now have over 1000 types of knockout mice) to understanding toxicologic mechanisms of action. There have also been major advances in neuro- and immuno-toxicology, in the use of alternative species as predictors, and in other areas related to hazard identification. Advances in the art of

toxicology have been less dramatic and have focused primarily on new approaches for exposure assessment and on dose response. Thus, for the two major problems in risk assessment (species-to-species and high to low dose extrapolations) it is likely that we will make more progress in resolving the species-to-species extrapolation issues because these can and are being resolved by advances in the science of toxicology. Although toxicologists attribute the axiom “the dose makes the poison” to Paracelsus, this concept is much older. Publius Terentius Afer (also called Terence) (c. 190-159) suggested “moderation in all things” in Andria (The Lady of Andros) (7:96), and there is a similar biblical admonition in Proverbs 25:27. Based on this principle, toxicologists conclude that there are no safe chemicals because all chemicals will be toxic under some conditions of exposure. Conversely, we also conclude that there is no chemical that cannot be used safely by simply reducing the exposure. Distinguishing between a safe and a toxic dose for any chemical or agent is determined by our ability to establish a threshold for its adverse effects. This is more complex for agents that have low-dose beneficial effects, such as drugs, or are essential nutrients because the dose responses for the beneficial and adverse effects may overlap. This phenomenon of low-dose stimulation by toxic agents is called hormesis by Calabrese, who has linked it to the Arndt-Schultz law (8). Most drugs and other chemicals exhibit more than one type of effect and thus, low-dose effects may appear to be antagonistic to high-dose effects. However, even when the lowand high-dose effects involve the same response, it is likely that these effects have different mechanisms of action. If we use Sagan’s definition of hormesis as a paradoxical or unanticipated effect of a toxic agent at low doses (9), then hormesis would be the rule rather than the exception not only in pharmacology but also in toxicology and most other biological sciences.

However,

the most

important contribution of this concept derives from the regulatory implications of hormesis. Because the U-shaped dose-response curves that are characteristic for hormetic agents all have thresholds, the linear multistage or no-threshold approach should not be used for regulating hormetic agents. Furthermore, banning or any approach that relies on a zero tolerance is not appropriate for agents with beneficial effects at low doses. Tolerances for such agents should be established by setting exposure levels that preserve the beneficial effects and protect against the adverse

TOXICOLOGY COMES OF AGE

17

effects (10). In his dedication to Toxicology and Risk Assessment, George Koelle argues that we should abandon the concept of zero tolerance for any regulatory purpose because it lacks a scientific basis (11). The existence of hormesis as a

toxicologic or biologic phenomenon also raises serious questions about the validity of using the maximum-tolerated-dose approach in testing for carcinogenesis and about any similar attempt to use a high-dose effect to predict a different low-dose effect. Faced with the need to develop regulations for exposure to carcinogens in the early 1970s and lacking evidence for carcinogenic thresholds, the EPA developed the linear multistage, or no-threshold, approach for regulating carcinogens. This approach extrapolated the results of high-dose studies in animals to zero, and the slope of this line was used to estimate the potential human health risk at very low exposures. This extrapolation had two errors; First, it assumed that time is not a variable of exposure at low doses and second, it used zero rather than one molecule

as the intercept for the extrapolation. Biological effects result from the interaction of chemical and host molecules and do not occur with less than whole molecules. Thus, one molecule is an absolute threshold dose for every chemical, including

both genotoxic and nongenotoxic carcinogens. It is also more informative when expressing such data graphically to define thresholds in terms of the number of molecules of the agent rather than as dose per body weight or with surface area metrics (12).

The Presidential/Congressional Commission on Risk Assessment and Risk Management was concerned by the discrepancies in the way our society currently regulates carcinogens and noncarcinogens, and recommended changes to improve the regulatory process (13). These and other changes are being included in the development of new risk-assessment protocols (14-16), and the new EPA

cancer risk guidelines now include a threshold or reference dose option for all carcinogens (17). These changes provide support for previous recommendations that it is time to abandon the linearized multistage approach and use the reference dose or threshold approach for all toxic effects (18-20). This would simplify risk assessment and improve the balance of regulation but would not eliminate all of the high to low dose extrapolation problems. Rozman has suggested a new paradigm for risk assessment that would address these problems and would also enhance our concepts of toxicology (12, 21-23). This paradigm is based on the recognition that both dose and time are independent variables in exposure. Dose or concentration is a simple variable (number of molecules), whereas time is a complex variable with both kinetic and dynamic time scales in addition to frequency/duration. We can use both variables to define the minimal exposure conditions (worst case) of dose and time that will produce a specific adverse effect with continuous exposure. Plotting the logs of these time and dose values will produce a line with a slope of —1 (Figure 1), which can be

described by Haber’s Law (ct = K"). The ends of this line are determined by the

dose and time thresholds for the specific adverse effects. The time threshold occurs when further increases in dose do not result in any further shortening of the time to

18

DOULL

Time Log

Log Concentration Figure 1

Concentration time relationships using Haber’s Law (24)

response, and the dose threshold occurs when increases in time no longer reduce the dose needed for an effect. Practical limits for the dose and time thresholds are provided by the life span (limits dose threshold) and by the amount of material that can be administered to the host (limits time threshold). Because any combination

of dose and time that produces values below this Haber’s line would not produce the specified adverse effect, this line defines safety. Fractionation of either the dose or time will produce lines with different slopes and K values (ct* = K, ct’ = K),

as will fractionation of both dose and time values (c*t”Y = K>*?). To establish a margin of exposure for any real-world situation, we can simply compare the

TOXICOLOGY COMES OF AGE

19

K values for the real world or experimental situation and the Haber’s or worse case

situation (margin of exposure = KP*?/K"), The advantages of this approach over the methods currently used in risk assessment are that it is based on actual data rather than on defaults, assumptions, models, or other hypothetical guidelines and that it allows us to base our decisions on safety rather than risk. Furthermore, this paradigm, like the principle of dose response, is independent of any mechanistic considerations. Although this approach provides a new paradigm for risk assessment, the concept inherent in Haber’s Law has been used for many years by the National Academy of Science/National Research Council Committee on Toxicology and by groups both in this country and abroad to evaluate and predict hazards. The line in Figure | represents a single effect in a single species, but similar lines could be obtained for other effects (either adverse or beneficial) and other species. The first step in using this approach to evaluate toxicity is to determine which of the candidate’s toxic effects is most critical, and the second is to determine whether the critical toxic effect is mediated primarily by the toxicokinetics or the toxicodynamics of the agent. Toxicity will only be produced by continuous or fractionated exposure when intake exceeds elimination (via distribution, metabolism, or excretion) or when injury exceeds recovery (via

recovery, repair, or adaptation). For chemicals with a very long half life, such as mirex or dioxins, kinetics is likely to be the mediating or rate-limiting step, whereas dynamics is likely to be rate limiting in chemicals that produce persistent damage, such as nitrosomines (24). This concept has implications for both the science and the art of toxicology and could be important for other disciplines in which time is an independent variable. It seems clear that the future of the science of toxicology is promising and will be exciting. However, the future of the art of toxicology may depend on the recognition by toxicologists that it is fundamental part of our discipline and is worthy of greater attention. In the final remarks of his prefatory chapter, Bob Neal (25) admonished toxicologists to “take a more active role in refuting public announcements of chemical risk by other scientists and by the print and electronic media that are not based on a scientific objective evaluation of the available data.” In a previous similar commentary, Bob and I suggested that toxicologists also need to be more responsible in presenting the results of their own studies and evaluations (26). We tend to focus on the trees of our individual findings

or opinions rather than on the forest of public health. Writing in a similar vein, Scala suggested that scientists must become democratically accountable to the broader needs of society in order to sustain societal support (27). The common thread of these messages suggests that the mandate of toxicology is not to use “what if” predictions to produce media headlines or to stimulate funding to investigate phantom risks, but it is rather to improve public health. To do this we must tell both sides of the story. There is a statue of Albert Einstein in the front yard of the National Academy of Sciences, and carved into its base are some of his quotes and observations. One of these that I think is particularly appropriate for toxicologists, risk assessors, and all scientists is “The right to search for truth

20

DOULL

implies also a duty; one must not conceal any part of what one has recognized to be true.”

Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED il. Hutchens JO. 1948. The Tox Lab. Sci. Mon. LXVI:107-12

. Goldstein A. 1997. A rewarding research pathway. Annu. Rev. Pharmacol. Toxicol. 37:1-28 . Nelson SR, Walaszek EJ. 1976. A history

of pharmacology at Kansas University. J. Kansas Med. Soc. LXXV11:339-43 . Woods LA, Doull J. 1991. GRAS

evalua-

tion of flavoring substances by the expert panel of FEMA. Regul. Toxicol. Pharmacol. 14:48-58 . Natl. Acad. Sci./Natl. Res. Counc./Comm.

Toxicol. 1999. Four decades of scientific service: a celebratory colloqium. Jnhal. Toxicol. 11:455-636 . Cook WA. 1985. History of ACGIH TLV’s. In Ann. Am. Conf. Gov. Ind. Hyg., 12:3-9. Cincinnati, OH: ACGIH. 389 pp. . Bartlett J. 1981. Familiar Quotations 15th and 125th Anniversary Edition, ed. EM Beck. Boston/Toronto: Little Brown . Calabrese EJ, Baldwin LA. 1999, Chemical hormesis: its historical foundation as

a biological hypothesis. Toxicol. Pathol. 27:195-216

. Sagan L. 1993. A brief history and critique of the low dose effects paradigm. Belle Newsl. 2(2):1-7 10. Rozman KK, Doull J. 1999. Hormesis, regulation, toxicity and risk assessment. Belle Newsl. 8(1):2-6 . Koelle GB. 2000. Dedication: the zero tol-

erance concept. In Toxicology and Risk Assessment, ed H. Salem, EJ Olajos, pp. viix. Philadelphia: Taylor & Francis Wee Rozman KK, Kerecsen L, Viluksela MK, Osterle D, Deml E, et al. 1996. A toxicolo-

gist’s view of cancer risk assessment. Drug Metab. Rev. 28:29-52

1S Risk Commission. 1997. Risk Asseement and Risk Management in Regulatory Decision Making in U.S. Comm. Risk Assess. Risk Manage. Final Rep., Vol. 2. Washington, DC: GPO #055-000-00568-1 14. Crump K. 1996. The linearized multistage model and the future of quantitative risk assessment. Hum. Exp. Toxicol. 15:787-98 15: Gaylor DW, Kodell RL, Chen JJ, Krewski D. 1998. A unified approach to risk assessment for cancer and noncancer endpoints based on benchmark doses and uncertainty/safety factors. Regul. Toxicol. Pharmacol. 29:151—S7 16. Gaylor DW, Kodell RL, Chen JJ, Krewski D. 1989. A unified approach to risk characterization. Inhal. Toxicol. 11:575—78 1h Environ. Prot. Agency. 1996. Proposed guidelines for carcinogenic risk assessment. Fed. Regist. 61(79):17960-8011 . Lu FC, Sielkin RL. Assessment of safety/ risk of chemicals: inception and evolution of the ADI and dose response modeling procedures. Toxicol. Lett. 59:5—40 Doull J. 1994. Back to basics. Int. J. Toxicol. 16:185-91 Purchase IFH, Auton TR. 1995. Thresholds in chemical carcinogenesis. Regul. Toxicol. Pharmacol. 22:199-205 . Rozman KK. 1999. Delayed acute toxicity of 1,2,3,4,6,7,8-heptachloro dibenzo-pdioxin (HpCDD) after oral administration obeys Haber’s rule of inhalation toxicity. Toxicol. Sci. 49:102-9 22, Rozman KK, Doull J. 2000. Dose and time as variables of toxicity. Toxicology 144:169-78 as}. Rozman KK. 2000. The role of time in toxicity or Haber’s ct product. Toxicology 149:35-42

TOXICOLOGY COMES OF AGE

21

24. Doull J, Rozman KK. 2000. Using Haber’s

26. Neal RA, Doull J. 1995. Commentary:

law to define the margin of exposure. Toxicology 149:1—2 25. Neal RA. 1996. A career in toxicology. Annu. Rey. Pharmacol. Toxicol. 36:35-46

discipline of toxicology. Fund. Appl. Toxicol. 24:151-53 27. ScalaRA. 1998. Reflections. Int. J. Toxicol. 18:1-6

the

a

@6

7

a 4

Ave

-

Annu. Rey. Pharmacol. Toxicol. 2001. 41:23-51

ANESTHETICS AND ION CHANNELS: Molecular Models and Sites of Action* Tomohiro Yamakura!, Edward Bertaccini?’, James R Trudell’, and R Adron Harris! 'Waggoner Center for Alcohol and Addiction Research and Institute for Cellular and Molecular Biology, University of Texas, Austin, Texas 78712; e-mail: harris @mail.utexas.edu ? Department of Anesthesiology, Niigata University School of Medicine, Niigata 951-8510, Japan; e-mail: yamakura@ med.niigata-u.ac.jp 3Department of Anesthesia, Stanford University School of Medicine, Stanford, California 94305; e-mail: edwardb@ stanford.edu, trudell@ leland.stanford.edu

Key Words _ general anesthetics, ligand-gated ion channels, recombinant receptors, electrophysiology, molecular modeling @ Abstract The mechanisms of general anesthesia in the central nervous system are finally yielding to molecular examination. As a result of research during the past several decades, a group of ligand-gated ion channels have emerged as plausible targets for general anesthetics. Molecular biology techniques have greatly accelerated attempts to classify ligand-gated ion channel sensitivity to general anesthetics, and have identified the sites of receptor subunits critical for anesthetic modulation using chimeric and mutated receptors. The experimental data have facilitated the construction of tenable molecular models for anesthetic binding sites, which in turn allows structural predictions to be tested. In vivo significance of a putative anesthetic target can now be examined by targeted gene manipulations in mice. In this review, we summarize from a molecular perspective recent advances in our understanding of mechanisms of action of general anesthetics on ligand-gated ion channels.

INTRODUCTION General anesthetics are some of the most widely used and important therapeutic agents. However, despite over a century of research, the molecular mechanisms of general anesthesia in the central nervous system remain elusive. As a result of research during the past several decades, there has been a transition from initial notions of nonspecific actions of general anesthetics on membrane lipids to the *The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

23

24

YAMAKURA ET AL

belief that their primary sites of action are neuronal proteins. Although many neuronal proteins can be affected by general anesthetics, a consensus has emerged that a group of ligand-gated ion channels, which are important for neuronal function, are particularly sensitive to general anesthetics (4). This review is restricted to discussion of ligand-gated ion channels! Ligand-gated ion channels include y-aminobutyric acid type A (GABA,), glycine, nicotinic acetylcholine (nACh), and 5-hydroxytryptamine, (5-HT;) receptors, along with a-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA)-, kainate-, and N-methyl-D-aspartate (NMDA)-selective glutamate receptors. GABAg, glycine, nACh, and 5-HT; receptors form part of an evolutionarily related ligand-gated ion channel gene superfamily (5), whereas glutamate receptors are thought to belong to a distinct ion channel class. Members of a ligand-gated ion channel superfamily have a basic transmembrane topology, with a large N-terminal extracellular domain, four putative transmembrane segments (TM1—TM4),

a heterogeneous intracellular loop between TM3

and TM4, and a

short extracellular C-terminal domain. Residues within the extracellular domain form the agonist-binding domain, whereas amino acid residues within TM2 line the ion channel pore. Native receptors are composed of pentameric arrangements of individual receptor subunits (3). On the other hand, glutamate receptors have three transmembrane segments (M1, M3, and M4) plus a cytoplasm-facing re-entrant membrane loop (M2) that lines the ion channel pore. Thus, the N-terminal domain

is located extracellularly, and the C-terminal domain, intracellularly. Subunit stoichiometry of native glutamate receptors is controversial between tetrameric and pentameric structures (6). General anesthesia is a behavioral state that requires critical degrees of immobility, amnesia, hypnosis/unconsciousness, analgesia, and muscle relaxation. Thus, key questions include the following:

1. Sensitivity: Which ligand-gated ion channels are sufficiently sensitive to clinically relevant concentrations of general anesthetics? How important is subunit composition for anesthetic sensitivity? 2. Mechanism: What is the molecular mechanism by which general anesthetics affect the function of ligand-gated ion channels?

3. In vivo importance: Which ligand-gated ion channels determine specific behavioral actions of general anesthetics? Is the responsible target common to a wide variety of general anesthetics or specific for certain agents?

' Abbreviations: ACh, acetylcholine; AMPA, a-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid; F6, 1,2-dichlorohexafluorocyclobutane; GABA, y-aminobutyric acid; S-HT, 5-hydroxytryptamine; MAC, minimum alveolar concentration: MPPB, 1-methyl-5phenyl-5-propyl barbiturate; nACh, nicotinic acetylcholine; NMDA, N-methyl-D-aspartate; TM, transmembrane segment.

MOLECULAR MECHANISMS OF ANESTHESIA

25

MOLECULAR ACTION OF ANESTHETICS ON LIGAND-GATED ION CHANNELS Relevant Anesthetic Concentrations For a target to have relevance to anesthesia, it must be sensitive to clinical concen-

trations of general anesthetics. Minimum alveolar concentration (MAC) conven-

tionally refers to the concentration of inhaled anesthetic that produces immobility, the lack of movement in response to a noxious stimulus, in 50% of subjects studied (7). The use of immobility as an anesthetic endpoint is helpful in that, for most general anesthetics, anesthetic concentrations two- to fourfold above the MAC cause deleterious side effects (1). Thus, anesthetic concentrations severalfold greater than

the MAC define the upper boundary of the concentration range that is clinically relevant. The Hill coefficient value of concentration-effect relationships for inhaled anesthetic MAC is generally large—6 to 20 (8). Some would argue that this is related to anesthetic mechanisms (9), but others argue that the steepness results from small population variations (8). The important point we have to keep in mind, however, is the categorical (move/no move) nature of the MAC measurement. Specifically, the slope of dose-response curves becomes steep when the data are categorically processed. The slope of categorical responses for each individual does not depend on the Hill value of the underlying dose-response relationship, but depends on the interval of concentrations examined, and it could be infinitely steep (Figure 1). This steep slope for the individual categorical response would be a basis of the steep slope of a population dose-response relationship (for MAC determination). Thus, the steep slope of MAC response may not necessarily be related to mechanisms of anesthetic actions on targets. The issue ofclinically relevant concentrations for intravenous anesthetics is considerably more complicated than that for volatile anesthetics because of limited pharmacokinetic data adequately collected for a defined endpoint of anesthesia and because of the difficulty in ascertaining steady-state free aqueous anesthetic concentrations in the brain (1). Relevant anesthetic concentrations of propofol and barbiturates have been carefully estimated (1). As for other intravenous anesthet-

ics, plausible concentrations for producing immobility estimated from available data are shown by Krasowski and Harrison (3).

General Anesthetic Actions on Recombinant

Ligand-Gated Ion Channels The advent of cloning and expression techniques has greatly accelerated and facilitated attempts to classify ligand-gated ion channel sensitivity to general anesthetics. Molecular biology techniques have the advantage of identifying the sites of receptor subunits critical for anesthetic modulation. Because sensitivity to general anesthetics varies considerably, sometimes even among closely related receptors,

26

YAMAKURA ET AL

10,

r e = a= Q'=

oo

&

A

&

Ad

D a g

=

0.85

064 4

0.5

2

@

ME 044 { @° oO 0.24 a

=

&

(ez

oa

o

0.04

AAA T

1Oie

SLL

fea

[ee

A

A Tae

T

|

104

[Isoflurane], M

[Isoflurane], M

@

beer Ba

it

a

v

v

v

peta Move

i

O

ee

el

vv

\

084

ag OS

0.64

So

2

\ \

ire)

i

Se

\\

ae)

O=

\\

N Oo move

* RK Geet

ae

ee

RK

*

bie =e oo

TO"

*

oe

[Isoflurane], M

oe

|

|~ Sots

iy cs ee \

Hill = 1.2 QA.

Yo

o44

are

Oo 0.2024 Se

ps adie Wiss ag INA We ah ey een | Ayo):

04

=

bed

9§eS

=

=©

10°

10%

10°

10°

‘

7

r

r

re]

\ esee®

pel 0.0

=

*

ee ee

10°

i

10°

r

T

10%

7

103

[Isoflurane], M

Figure 1 Schematic simulation showing that the slope of categorical responses could become steep independent of the Hill coefficient value of underlying dose-response relationships. (A) Isoflurane dose-inhibition relationships for nACh receptors expressed in a single oocyte (open triangle) and categorical responses [move (1)/no move ((0)] for a single oocyte (or a patient) processed based on the dose-inhibition curve of an oocyte, assuming that relative responses >0.5 produce movement and relative responses T]iodophenyl)diazirine (74). The periodicity of the resulting labeling was consistent with both TM3 and TM4 being a@ helices. The cholesterol-binding domain in the

nACh receptor has been located near TM4 with [!?5I] azido-cholesterol (83). Early studies demonstrated that the TM4 segment could be substituted with homologous

sequences without loss of receptor function (84). However, more recent studies

have shown that substitution of C418 in TM4 by tryptophan altered ion channel function (85). The latter studies suggest that although TM4 is thought to be distant from the channel pore, it is important in gating of the ion channel.

MOLECULAR MECHANISMS OF ANESTHESIA

Tertiary Structure Model of the GABA,

37

Receptor Subunit

Molecular Modeling We obtained the sequences of nine related proteins in the superfamily of ligand-gated ion channels. These were nACh receptor a4, a7, and torpedo a subunits; GABA, receptor @;, @>, 8;, and 6; subunits; glycine receptor a@, subunit, and GABA receptor p, subunit. We predicted the topology of the transmembrane domains of these segments with the bioinformatics techniques PHDhtm and HMMTOP. The ends of the transmembrane a helices were averaged. The sequences were simultaneously aligned with the multiple sequence alignments algorithm ClustalW. The averaged secondary structure predictions were added to the multiple sequence alignment to give a clearer picture of regions of secondary structure and helical limits that were common to all nine sequences. This analysis clearly predicted that the four transmembrane segments (TM 1-4) were all a helices with reasonably well-defined helical ends. This secondary structure information was then used in conjunction with the SeqFold algorithm to search for a modeling template based upon both sequence and fold homology/analogy. Although the SeqFold search of a modified version of the Protein Data Bank produced several well-scored alignments, there was only one template that was of mammalian origin and showed an alignment over all four transmembrane domains. This was the chain C domain | portion of bovine cytochrome oxidase, locc.pdb (Figure 2, see color insert). In a qualitative search of both the SCOP (Structural Classification of Proteins) and CATH (Class, Architec-

ture, Topology, and Homologous superfamily) fold databases, this template was again the only one found that was a tetramer of a helices, mammalian in origin and common to both databases. This template was then aligned with the sequence of GABA, receptor a, subunit. This alignment was initially scored based upon sequence similarity, fold similarity, and the mean force potential of Sippl (86). We used knowledge-based modeling to produce agreement with the hydrophilic and hydrophobic labeling studies (Figure 3A, see color figure) carried out in the homologous nACh receptors (66, 67, 74) and with the proposed juxtaposition of Arg 273 and Asp 286 to forma salt bridge. We also hypothesized that the lack of a halothane effect owing to the L231F mutation (87) was due to its proximity to Ser 269 and Ala 290 (TM2 and TM3 residues critical for anesthetic actions, respectively). Conformer libraries were used to provide reasonable initial orientations of the amino acid side chains. The loops between @ helices were constructed to minimize fraying of helical ends. The TM1-2 and TM2-3 loops were constructed using the loop modeling features within the Swiss Protein Data Bank Viewer (88). These are based upon loop fragment sequence similarities to loops of known three-dimensional structure. The large TM3-4 loop was removed and replaced by a series of 6 glycine residues. This substitution allowed maximum flexibility of the loop, but maintained a reasonable distance between the end of TM3 and the beginning of TM4. The entire structure was subjected to sequentially restrained molecular mechanics energy optimization. Because the sequence of GABA, receptor a subunit is quite different from the primary sequence of the template, we used a previously

38

YAMAKURA

ET AL

described technique to maintain the left-handed supertwist of the template of four a helices without imposing additional restraints on the backbone atoms (89). Each a helix was divided into an upper, middle, and lower third. Then a centroid of the backbone atoms of each third was defined. Distance restraints were applied to maintain an 11 A separation between adjacent centroids of neighboring a helices. These restraints had the appearance of three squares, one each at the upper, middle, and lower levels of the tetramer. The structure was optimized using these restraints with Discover 98 (MSI, San Diego, CA) and the CFF91

forcefield.

In order to

allow the side chains of the tetramer to adjust to optimum packing before the helical backbone was distorted, the initial force constant of the centroid restraints was

set to 100 kcal/A? with subsequent reductions to 10 and 1 kcal/A? in sequential optimizations. We then used a template-forcing algorithm to align the TM2 a helix of this tetramer onto the pore-lining a helix of the pentameric bacterial stretch receptor [1msl (mechanosensitive ion channel) in the Protein Data Bank]. This was repeated with fivefold symmetry to produce a pentamer of tetramers totaling 20 a@ helices. Simulated annealing with restrained molecular dynamics produced our final structure. We used two sets of restraints in these simulations. First, the distance restraints between a helices defined above were used for each subunit with a force constant of 10 kcal/A. Second, the template forcing distances that were used to construct the pentamer of TM2 segments onto the |msl template were retained and assigned a force constant of 100 kcal/A?. We subjected this restrained structure to molecular dynamics with sequential steps (10,000 cycles of 1 femtosecond) of heating by 100°K up to 500°K, an equilibration step of 10,000 cycles at 500°K, and then cooling steps of 100°K. Finally, we re-optimized the structure with molecular mechanics,

and recorded energies of both bonded and nonbonded

interactions.

This procedure produced a structure in which specific amino acid residues from all four transmembrane @ helices were in direct proximity to one another. These residues are Leu 231 (TM1), Ser 269 (TM2), Ala 290 (TM3), and Val 407 (TM4).

Whereas residues that were homologous to Ser 269 and Ala 290 were previously studied in the glycine receptor as being fundamental to anesthetic action, it is only recently that the residue homologous to Leu 231 has been shown to be important as well (87). Lastly, this model clearly predicts that it is the Val 407 on TM4 that should prove to be important for anesthetic binding in conjunction with the other three residues (Figure 3B).

Size of the Anesthetic Binding Pocket We used molecular modeling to visualize the effect of site-directed mutations in the glycine and GABA receptors. Acommon theme was the volume of a site in the transmembrane domain. Homologous nalcohols increase in CNS depressant potency as the length of the carbon chain increases, up to a point at which further increases in molecular size decrease (or no longer increase) alcohol potency. This is termed the cutoff. Our group

found that alcohol regulation of the glycine receptor critically depends on specific amino acid residues in TM2 and TM3 of the a subunit polypeptide (29). They

YAMAKURA ET AL C-1

‘(QOKJANS IV[N][IOVIYXS OY) WOT, MOIA & ‘OSVUIT YS SOURIqUIOUT oY} JO ouRTd SY} WOT MOIA & ‘OSVUTI 19}U99) aye] -UI9} B SP ASN IO} Pasioxd pue (ASCUII Yd]) BSePIXO QUIOIYDIO}JAD UT POTJUAPI Se S[PUNQ XI[IY-INOJ Y ‘WuNgns xI[sy-INoj ev JO} aje[dwisay 7 sans1y

C-2 YAMAKURA ET AL

TM I

f>

ee) TMI: LEU 231

Figure 3 (A) Satisfying restraints based on labeling studies. View of subunit from the extracellular side. The pore-lining TM2 « helix is at the bottom of image. The alignment for lipid-facing and pore-facing labels: blue = hydrophilic label, yellow = hydrophobic label. (B) Proximity of residues in which mutation affects anesthetic potency. In this model of a GABA, receptor a7 subunit, viewed from the extracellular side, three known residues are shown (S269, A290, L231). A fourth residue is predicted to form part of the boundary of the putative binding cavity (V407).

MOLECULAR MECHANISMS OF ANESTHESIA

39

demonstrated that residues in the glycine a, receptors and the GABA p, receptor also control alcohol cutoff. Mutation of Ser 267 to Gln (S267Q) decreased the n-alcohol cutoff in glycine a, receptor. Conversely, mutation of Ile 307 and/or Trp 328 to smaller residues increased the n-alcohol cutoff in the GABA p, receptor. These results are consistent with the suggestion that n-alcohol binding sites within these transmembrane proteins modulate the action of these alcohols, and that amino acid residues present at these critical positions limit the size of alcohol molecules that can interact with the binding cavity. We modeled the double site-directed mutation of Ile to Ser at position 307 (1307S) in TM2, and Trp to Ala at position 328 (W328A) in TM3 in the human

GABA receptor p, subunit. The change in van der Waals volume resulting from the double site-directed mutation, assuming that the inter-helical packing does not allow movement of the backbone atoms of the helices, is 162 — 109 = 53 A? plus 224 — 99 = 125 A?,a total volume change of 178 A®. This volume corresponds to that of CjH,;OH. In the series of alcohols studied, each additional CH, adds approximately 20 A? of van der Waals volume. The volume of 178 A? is equivalent to 8.9 additional methylenes (178 A?/20 A’). On this basis, we predicted that the double site-directed mutation could increase the alcohol cutoff by 8 to 9 methylene groups. We observed that the double site-directed mutation increased the alcohol cutoff by at least 7 methylenes (29) (alcohols greater than C-12 were too insoluble to test).

Interactions with Anesthetics Residues in our current model of GABA , receptor a, subunit were replaced at several positions in an effort to elucidate the effect of side chain volume characteristics on the anesthetic binding regions noted above. These mutations were S269A, S269W,

and S269W + A290W (A Jenkins, EP A Andreasen, A Viner, JR Trudell,

Greenblatt, HJ Faulkner, E Bertaccini, A Light,

NL Harrison, unpublished data). The resulting three new models were re-optimized as above. We defined a site midway between the Ca carbons of residues 269 and 290.

At this site we manually inserted halothane, isoflurane, and either one or

two chloroform molecules to produce a set of 16 ligand-receptor complexes. All protein backbone atoms in these models were constrained to their initial positions. Only movement of the ligand and those amino acids residues within 5 A of this ligand was allowed. Each of these complexes was subjected to simulated annealing with restrained molecular dynamics. The many restraints in these calculations cause the resulting energies of these ligand-receptor complexes to be of only qualitative value. However, they suggest that all three ligands bind well to sites near $269 or A290. The single tryptophan mutant (S269W) binds only one chloroform well, suggesting that the native binding pocket, with a smaller amino acid residue, could contain two chloroform molecules. The double tryptophan mutant (S269W + A290W) produces an unstable complex in the forced presence of all ligands. Therefore, based upon our knowledge of transmembrane mutations and their effects on anesthétic activity, the way in which transmembrane amino acids can

be chemically labeled,

the physicochemical

characteristics

among

series of

40

YAMAKURA ET AL

anesthetics that are favorable for activity, and the ways in which anesthetics bind to generic proteins, one can look for sites in ligand-gated ion channels that meet these criteria. With this, we can show that the region that is both necessary and sufficient for anesthetic activity is the region within the tetrameric subunit of ligand-gated ion channels that is bounded by four specific amino acids: Leu 231 (TM1), Ser

269 (TM2), Ala 290 (TM3), and Val 407 (TM4).

CORRELATION WITH IN VIVO ANESTHETIC ACTION

Stereoselectivity General anesthetic stereoselectivity poses the greatest challenge to traditional lipid theories of anesthetic actions because optical isomers of general anesthetics behave identically with respect to their ability to disorder lipid bilayers, despite significant differences in their in vivo potency (90,91).

Furthermore, stereoselectivity

represents a powerful test for the relevance of a putative anesthetic target, i.e. if a given receptor is one of the important anesthetic targets, the rank order and ratio of potency of the optical isomers should be similar for producing anesthetic actions and for modulating the receptor function. R(+) etomidate is ~17 times as potent as S(—) etomidate to induce a loss of

righting reflex in tadpoles, and the R(+) isomer is 6-10 times more effective than the S(—) isomer at potentiating GABA-induced currents (91).

Steroid anesthetics such as 3a@-hydroxy-Sa@-pregnan-20-one are enantioselective both for inducing a loss ot righting reflex in tadpoles and mice and for potentiating GABA-induced currents, with (+) isomer exhibiting greater potency than (—) isomer (92, 93).

S isomers of barbiturates are generally more potent than R isomers for their anesthetic activities (94). GABA, receptor responses are potentiated by S isomers of hexobarbital, pentobarbital, and thiopental more potently than by their R isomers by a factor of 2~4 (95, 96), whereas neuronal @ 85 nACh receptor is inhibited by R(+) thiopental more effectively than by S(—) thiopental, and inhibition of neuronal @; and muscle a@;8,y5 nACh receptors by thiopental isomers shows no stereoselectivity (97). Furthermore, GABA, receptor function is enhanced by an anesthetic barbiturate R(—) 1-methyl-5-phenyl-5-propyl barbiturate (MPPB), but is inhibited by its convulsant stereoisomer S(+) MPPB (98,99), whereas both R(—) and S(+) MPPB inhibited AMAP (a@-amino-3-hydroxy-5-methyl-4-iso-

xazole propionic acid) receptor current responses, with R(—) MPPB being only slightly potent compared with S(+-) MPPB (99). S(+) isoflurane are slightly (~20-50%) more potent than R(—) isoflurane for measurements of sleep time and MAC (100-102). S(+) isoflurane is ~2 times more potent than R(—) isoflurane to potentiate GABA, receptor function (103— 105), and ~1.5 times more potent to inhibit neuronal nACh receptor currents (90).

In contrast, two optical isomers of isoflurane are equally effective in potentiating glycine receptor currents (106).

MOLECULAR MECHANISMS OF ANESTHESIA

41

S(+) ketamine is 3-4 times more potent than R(—) ketamine as an analgesic

agent (107-108) and 1.54 times more potent in terms of hypnotic activity (107— 109). S(+) ketamine is 2-4 times more potent than R(—) ketamine in inhibiting NMDA receptor currents (110-112), whereas S(+) ketamine is 1.5 times as potent as R(—) ketamine in reducing neuronal excitation by acetylcholine

(rt3), Thus, in general, the consistent correlation of in vivo anesthetic potencies of

optical isomers of several anesthetics with their potencies for modulating receptor function is observed only for potentiation of GABA, receptor activity (except for ketamine and NMDA receptors), supporting the plausibility of GABA, receptors as anesthetic targets.

Nonimmobilizers Certain highly lipid-soluble halogenated cyclobutanes and alkanes are unable to induce immobility at concentrations predicted by the Meyer-Overton rule to be in the anesthetic range (114). The nonimmobilizers provide clues to the relevance of a putative anesthetic target, i.e. if a given receptor is one of the anesthetic targets, anesthetics should affect the receptor, but nonimmobilizers should not modulate the

receptor function. The nonimmobilizer 1,2-dichlorohexafluorocyclobutane (F6) exhibits no modulatory effects on GABA, (115), glycine (13), GABAg p (10), 5-HT; (116), neuronal nACh (117), AMPA (18), kainate (18), or NMDA (T

Yamakura & RA Harris, unpublished data) receptors. Muscle nACh receptors are affected by F6 through mechanisms distinct from anesthetics (118-119). On the other hand, F6 inhibits responses of G protein-coupled receptors such as metabotropic glutamate (120), muscarinic ACh (121), or 5-HT,, (122) recep-

tors the way volatile anesthetics do. These results would seem to exclude these G protein-coupled receptors as viable targets for producing immobility. These receptors may certainly play a role in other actions, such as amnesia, because nonimmobilizers may depress learning and memory (123).

Gene Targeting Gene targeting in mice is very powerful for elucidating the in vivo roles of certain ligand-gated ion channel subunits in mediating diverse behavioral actions of general anesthetics. Several knockout mice lacking specific subunits for ligandgated ion channels have already been created, and anesthetic sensitivities have been determined in some of these mice (124). Mice lacking the 6; subunits of GABA, receptors exhibit some resistance to

the immobilizing actions of halothane and enflurane, but sensitivity to obtunding (loss of righting reflex) effects of volatile anesthetics is not altered (125). These mice are also more resistant to obtunding effects of etomidate and midazolam, but not to those of pentobarbital and ethanol, compared with wild-type mice (125). Mice lacking the a, subunits of GABA, receptors exhibit sensitivities to the

immobilizing actions of enflurane and obtunding effects of halothane, enflurane,

42

YAMAKURA ET AL

pentobarbital, and ethanol, which are not significantly different from those of wild-type mice (126). Mice deficient in the 6 subunits of GABA, receptors exhibit a selective attenuation of sensitivities to obtunding effects of neuroactive steroids such as alphaxalone and pregnanolone (127). This is surprising because the presence of the 6 subunits reduces modulation of recombinant GABA, receptors by neurosteroids (128). Sensitivities to halothane, propofol, midazolam, etomidate, and ketamine are unaffected by the 6 subunit knockouts (127). Mice deficient in GluR2 subunits of AMPA receptors show increased sensitivity to behavioral effects of pentobarbital (loss of righting reflex and loss of corneal, pineal, and toe-pinch withdrawal reflexes) despite reduced pentobarbital inhibition of AMPA receptors in their hippocampal neurons (129). Although knockout mice may provide initial clues to the nature of anesthetic targets, results can be difficult to interpret because compensation for the absent subunits by alterations in other subunits may occur. Furthermore, aberrations in neural development, grossly abnormal motor behavior, and lethality may preclude analyzing the anesthetic sensitivity. These shortcomings of global gene knockouts may be circumvented by conditional gene knockouts, which enable gene knockouts only in limited brain regions and/or specified developmental stages (130). Another strategy overcoming the limitations of global gene knockouts is the gene knock-in technique. This technique introduces, for example, a mutation into receptor subunits that eliminates anesthetic modulation but does not alter other physiological function. This approach has recently applied to benzodiazepines, and revealed that the a, subunit of GABA, receptors mediates sedative, hypnotic, and partly anticonvulsant actions of diazepam, but not anxiolytic, myorelaxant, motor-impairing, and ethanol-potentiating effects of diazepam (131).

CONCLUDING REMARKS AND FUTURE DIRECTIONS As reviewed above, the most plausible target of general anesthetics among ligandgated ion channels is the GABA, receptor. However, there are also some anesthetics that do not affect the GABA, receptors at clinical concentrations, i.e. gaseous anesthetics nitrous oxide and xenon, dissociative anesthetic ketamine, and fluorinated alcohols that have in vivo anesthetic effects (20-22, 24, 25, 132-134). Thus, -

it is unlikely that GABA, receptors are universal anesthetic targets involved in producing general anesthesia. General anesthesia, however, is a behavioral state that requires varying degrees of immobility, amnesia, hypnosis/unconsciousness, analgesia, and so on. Considering that different classes of anesthetics have distinct behavioral effects, it may not be surprising that a wide variety of anesthetics has a diverse pattern of actions not only on ligand-gated ion channels, but also on other brain receptors/proteins. These issues are related to the key question of whether the neuronal function mediating certain behavioral actions of anesthesia is common to every general anesthetic or not (the unitary theory). In vivo studies of specific

MOLECULAR MECHANISMS OF ANESTHESIA

43

anesthetic behaviors are required to address this question, and the most convincing investigation to directly correlate the anesthetic behaviors and neuronal function will be the analysis of gene knock-in mice with a mutation in target proteins that eliminates anesthetic actions but does not otherwise change the physiological function. This experiment may be complicated if multiple subunit isoforms are involved in producing an anesthetic behavior, but recent knock-in studies showing a specific role of the a; subunit of GABA, receptors in certain types of behavioral actions of benzodiazepines (131) are encouraging for future anesthetic research. Through the elucidation of molecular mechanisms of general anesthesia in conjunction with structural biology and rational drug design, it may become possible to develop specific behavior- or target site—oriented anesthetic strategies without interfering with other physiological functions. ACKNOWLEDGMENTS The authors thank Drs. Edmond I Eger, II (University of California, San Francisco) and Neil L Harrison (Weill Medical College, Cornell University) for helpful discussions. Funding was generously provided by National Institutes of Health grants AA06399 and GM47818. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED ie Franks NP, Lieb WR. 1994. Molecular and cellular mechanisms of general anaesthesia. Nature 367:607-14 . Harris RA, Mihic SJ, Dildy-Mayfield JE, Machu TK. 1995. Actions of anesthetics on ligand-gated ion channels: role of receptor subunit composition. FASEB J. 9:1454—62 . Krasowski MD, Harrison NL.

1999. General anaesthetic actions on ligand-gated ion channels. Cell. Mol. Life Sci. 55:1278-303

. Belelli JJ.

I, Pistis

1999.

General

I, Peters

JA,

anaesthetic

Lambert action

at

transmitter-gated inhibitory amino acid receptors. Trends Pharmacol. Sci. 20:496—502 MO, Lunt GG. 1995. Evolutionary history of the ligand-gated ion-channel superfamily of receptors. Trends Neurosci. 18:121-27

. Ortells

. Dingledine R, Borges K, Bowie D, Traynelis

SE. 1999. The glutamate receptor ion channels. Pharmacol. Rev. 51:7-61 . Eger EI II, Saidman LJ, Brandstater B. 1965.

Minimum alveolar anesthetic concentration: a standard of anesthetic potency. Anesthesiology 26:756—-63 . Eger EI I, Fisher DM, Dilger JP, Sonner JM, Evers A, et al. 2000. Relevant concentrations of inhaled anesthetics for in vitro studies of anesthetic mechanisms. Anesthe-

siology. In press . Eckenhoff RG, Johansson JS. 1999. On the

relevance of “clinically relevant concentrations” of inhaled anesthetics in in vitro experiments. Anesthesiology 91:856—60 10. Mihic SJ, Harris RA. 1996. Inhibition of p, receptor GABAergic currents by alcohols and volatile anesthetics. J.Pharmacol. Exp. Ther. 277:411-16 11. Hill-Venning C, Belelli D, Peters JA, Lambert JJ. 1997. Subunit-dependent interaction of the general anaesthetic etomidate with the y-aminobutyric acid type A receptor. Br J. Pharmacol. 120:74956

44

YAMAKURA ET AL

12. Belelli D, Lambert JJ, Peters JA, Wafford K, Whiting PJ. 1997. The interaction of the general anesthetic etomidate with the y-aminobutyric acid type A receptor is influenced by a single amino acid. Proc. Natl. Acad. Sci. USA 94:11031—36 ie Mascia MP, Machu TK, Harris RA. 1996. Enhancement of homomeric glycine receptor function by long-chain alcohols and anaesthetics. Br. J. Pharmacol. 119:133136 14. Pistis M, Belelli D, Peters JA, Lambert JJ. 1997. The interaction of general anaesthetics with recombinant GABA, and glycine receptors expressed in Xenopus laevis oocytes: a comparative study. Br. J. Pharmacol. 122:1707-19 15! Zhang L, Oz M, Stewart RR, Peoples RW, Weight FF. 1997. Volatile general anaesthetic actions on recombinant nACh,7, 5HT 3 and chimeric nACh,7-5-HT3 receptors expressed in Xenopus oocytes. Br. J. Pharmacol. 120:353-55 16. Flood P, Ramirez-Latorre J, Role L. 1997. a462 neuronal nicotinic acetylcholine receptors in the central nervous system are inhibited by isoflurane and propofol, but a7-type nicotinic acetylcholine receptors are unaffected. Anesthesiology 86:859-65 Ve Violet JM, Downie DL, Nakisa RC, Lieb WR, Franks NP. 1997. Differential sensitivities of mammalian neuronal and muscle nicotinic acetylcholine receptors to general anesthetics. Anesthesiology 86:866—74 18. Dildy-Mayfield JE, Eger EI II, Harris RA. 1996. Anesthetics produce subunitselective actions on glutamate receptors. J. Pharmacol. Exp. Ther. 276:1058—65 19% Taverna FA, Cameron BR, Hampson DL, Wang LY, MacDonald JF. 1994. Sensitivity of AMPA receptors to pentobarbital. Eur: J. Pharmacol. 267:R3-—5 20. Jevtovic-Todorovic V, Todorovic SM, Mennerick S, Powell S, Dikranian K, et al. 1998. Nitrous oxide (laughing gas) is an NMDA antagonist, neuroprotectant and neurotoxin. Nature Med. 4:460-63

S, Jevtovic-Todorovic V, Pis\e Mennerick Todorovic SM, Shen W, Olney JW, et al. 1998. Effect of nitrous oxide on excitatory and inhibitory synaptic transmission in hippocampal cultures. J. Neurosci. 18:9716—26 Piper Yamakura T, Harris RA. 2000. Effects of gaseous anesthetics nitrous oxide and xenon on ligand-gated ion channels: comparison with isoflurane and ethanol. Anesthesiology. In press 23% Dzoljic M, Van Duijn B. 1998. Nitrous oxide-induced enhancement of yaminobutyric acid ,-mediated chloride currents in acutely dissociated hippocampal neurons. Anesthesiology 88:473—80 24. Franks NP, Dickinson R, de Sousa SLM, Hall AC, Lieb WR. 1998. How does xenon produce anaesthesia? Nature 396:324 72a% de Sousa SL, Dickinson R, Lieb WR, Franks NP. 2000. Contrasting synaptic actions of the inhalational general anesthetics isoflurane and xenon. Anesthesiology

92:1055-66 S, Roberts RJ. 1998. Postsynaptic inhibitory mechanisms of anaesthesia; glycine receptors. Toxicol. Lett. 100/101:71-76 . Wachtel RE. 1995. Relative potencies of volatile anesthetics in altering the kinetics of ion channels in BC3H1 cells. J. Pharmacol. Exp. Ther. 274:1355-61 . Daniels

. Mihic SJ, Ye Q, Wick MJ, Koltchine VV,

Krasowski MD, et al. 1997. Sites of alcohol and volatile anaesthetic action on GABAy and glycine receptors. Nature 389:38559 . Wick MJ, Mihic SJ, Ueno S, Mascia MP,

Trudell JR, et al. 1998. Mutations of yaminobutyric acid and glycine receptors change alcohol cutoff: evidence for an alcohol receptor? Proc. Natl. Acad. Sci. USA

95:6504-9 . Ye Q, Koltchine

VV, Mihic

SJ, Mascia

MP, Wick MJ, et al. 1998. Enhancement of glycine receptor function by ethanol is inversely correlated with molecular volume

MOLECULAR MECHANISMS OF ANESTHESIA

at position #267. J. Biol. Chem. 273:3314— 19

Sik Krasowski MD, Koltchine VV, Rick CE, Ye Q, Finn SE, et al. 1998. Propofol and other intravenous anesthetics have sites of action on the y-aminobutyric acid type A receptor distinct from that for isoflurane. Mol. Pharmacol. 53:530-38 32. Krasowski MD, Finn SE, Ye Q, Harrison NL. 1998. Trichloroethanol modulation of recombinant GABA,, glycine and GABA , receptors. J. Pharmacol. Exp. Ther. 284:934-42 33. Ueno S, Wick MJ, Ye Q, Harrison NL, Harris RA. 1999. Subunit mutations affect ethanol actions on GABA, receptors expressed in Xenopus oocytes. Br. J. Pharmacol. 127:377-82

Lambert JJ. 1998. The effect of a transmembrane amino acid on etomidate sensitivity of an invertebrate GABA receptor. Br. J. Pharmacol. 124:13—20

40. Belelli D, Pau D, Cabras G, Peters JA, Lambert JJ. 1999. A single amino acid confers barbiturate sensitivity upon the GABA p, receptor. Br. J. Pharmacol. 127:601—4 41. Amin J. 1999. A single hydrophobic residue confers barbiturate sensitivity to yaminobutyric acid type C receptor. Mol. Pharmacol. 55:411—23 42. Birnir B, Tierney ML, Dalziel JE, Cox GB, Gage PW. 1997. A structural determinant of desensitization and allosteric regulation by pentobarbitone of the GABA receptor. J. Membr. Biol. 155:157-66 43. Carlson BX, Engblom AC, Kristiansen U,

34. Koltchine VV, Finn SE, Jenkins A, Nikolaeva N, Lin A, et al. 1999. Agonist gating and isoflurane potentiation in the human y-aminobutyric acid type A receptor determined by the volume of a second transmembrane domain residue. Mol. Pharmacol. 56:1087—93

335). Yamakura T, Mihic SJ, Harris RA. 1999, Amino acid volume and hydropathy of a transmembrane site determine glycine and anesthetic sensitivity of glycine receptors. J. Biol. Chem. 274:23006-12 36. Krasowski MD, Harrison NL. 2000. The actions of ether, alcoho] and alkane gen-

eral anaesthetics on GABA, and glycine receptors and the effects of TM2 and TM3 mutations. Br. J. Pharmacol. 129:731—43 Bie Pistis M, Belelli D,McGurk K, Peters JA, Lambert JJ. 1999. Complementary regulation of anaesthetic activation of human (6637 21,) and Drosophila (RDL) GABA receptors by a single amino acid residue. J. Physiol. 515:3-18

38. Moody EJ, Knauer C, Granja R, Strakhova M, Skolnick P. 1997. Distinct loci mediate

the direct and indirect actions of the anesthetic etomidate at GABA, receptors. J. Neurochem. 69:1310-13

3y)) McGurk KA, Pistis M, Belelli D, Hope AG,

45

Schousboe

44.

A, Olsen

RW.

2000.

A sin-

gle glycine residue at the entrance to the first membrane-spanning domain of the yaminobutyric acid type A receptor > subunit affects allosteric sensitivity to GABA and anesthetics. Mol. Pharmacol. 57:474— 84 Minami K, Wick MJ, Stern-Bach Y, DildyMayfield JE, Brozowski

SJ, et al. 1998.

Sites of volatile anesthetic action on kainate (glutamate receptor 6) receptors. J. Biol. Chem. 273:8248-S5

45. Yamakura T, Sakimura K, Mishina M, Shimoji K. 1995. The sensitivity of AMPAselective glutamate receptor channels to pentobarbital is determined by a single amino acid residue of the w2 subunit. FEBS Lett. 374:412-14 46. Yamakura T, Shimoji K. 1999. Subunitand site-specific pharmacology of the NMDA receptor channel. Prog. Neurobiol. 59:279-98 47. Mori H, Masaki H, Yamakura T, Mishina M. 1992. Identification by mutagenesis of a Mg’*-block site of the NMDA receptor channel. Nature 358:673-75 48. Yamakura T, Mori H, Masaki H, Shimoji K, Mishina M. 1993. Different sensitivi-

ties of NMDA

receptor channel subtypes

46

YAMAKURA ET AL

to non-competitive antagonists. NeuroReport 4:687-90

49. Forman SA, Miller KW, Yellen G. 1995. A discrete site for general anesthetics on a postsynaptic receptor. Mol. Pharmacol. 48:574-81 50. Forman SA. 1997. Homologous mutations on different subunits cause unequal but additive effects on n-alcohol block in the nicotinic receptor pore. Biophys. J. 72:2170-79 ile Colquhoun D. 1998. Binding, gating, affinity and efficacy: the interpretation of structure-activity relationships for agonists and of the effects of mutating receptors. Br. J. Pharmacol. 125:924-47 32) Eckenhoff RG, Johansson JS. 1997. Molecular interactions between inhaled anesthetics and proteins. Pharmacol. Rev. 49:343-67 2k Husain SS, Forman SA, Kloczewiak MA,

54.

SP}

56.

Sl.

58.

Addona

GH,

Olsen

RW,

et

al.

1999. Synthesis and properties of 3-(2hydroxyethyl)-3-n-pentyldiazirine, a photoactivable general anesthetic. J. Med. Chem. 42:3300-7 Pratt MB, Husain SS, Miller KW, Cohen JB. 2000. Identification of sites of incorporation in the nicotinic acetylcholine receptor of a photoactivatible general anesthetic. J. Biol. Chem. In press Mascia MP, Trudell JR, Harris RA. 2000. Specific binding sites for alcohol and anesthetics in ligand-gated ion channels. Proc. Natl. Acad. Sci. USA 97:9305-10 Stowell MH, Miyazawa A, UnwinN. 1998. Macromolecular structure determination by electron microscopy: new advances and recent results. Curr. Opin. Struct. Biol. 8:595—600 Sansom MS. 1998. Ion channels: molecular modeling and simulation studies. Methods Enzymol. 293:647—-93 Scott SP, Tanaka JC. 1998. Use of homology modeling to predict residues involved in ligand recognition. Methods Enzymol. 293:620-47

seh Sutcliffe MJ, Smeeton AH, Wo ZG, Oswald RE. 1998. Molecular modeling of ligand-gated ion channels. Methods Enzymol. 293:589-620 60. Bertaccini E, Trudell JR. 1999. Prediction of the secondary structure of an anesthetic site of action in the glycine alpha 1 receptor subunit. Anesthesiology 91:A363 61. Unwin N. 1993. Nicotinic acetylcholine receptor at 9 A resolution. J. Mol. Biol. 229:1101—24 62. Unwin N. 1995. Acetylcholine receptor channel imaged in the open state. Nature 373:37-43 63. Unwin N. 1996. Projection structure of the nicotinic acetylcholine receptor: distinct conformations of the @ subunits. J. Mol. Biol. 257:586—96 64. Le Novere N, Corringer PJ, Changeux JP. 1999. Improved secondary structure predictions for a nicotinic receptor subunit: incorporation of solvent accessibility and experimental data into a two-dimensional representation. Biophys. J. 76:2329-45 . Ortells MO, Barrantes GE, Wood C, Lunt

66.

67.

68.

69.

GG, Barrantes FJ. 1997. Molecular modelling of the nicotinic acetylcholine receptor transmembrane region in the open state. Protein Eng. 10:511-17 Akabas MH, Karlin A. 1995. Identification of acetylcholine receptor channel-lining residues in the M1 segment of the asubunit. Biochemistry 34:12496—500 Williams DB, Akabas MH. 1999. yaminobutyric acid increases the water accessibility of M3 membrane-spanning segment residues in y-aminobutyric acid type A receptors. Biophys. J. 77:2563-74 Leite JF, Amoscato AA, Cascio M. 2000. Coupled proteolytic and mass spectrometry studies indicate a novel topology for the glycine receptor. J. Biol. Chem. 275:13683-89 Gorne-Tschelnokow U, Strecker A, Kaduk C, Naumann D, Hucho F. 1994. The trans-

membrane domains of the nicotinic acetylcholine receptor contain a@-helical and

47

MOLECULAR MECHANISMS OF ANESTHESIA

B structures.

EMBO

J.

13:338-41

70. Akabas MH, Kaufmann C, Archdeacon P, Karlin A. 1994. Identification of acetylcholine receptor channel-lining residues in the entire M2 segment of the a@ subunit. Neuron 13:919-27 Wile Chang G, Spencer RH, Lee AT, Barclay MT, Rees DC. 1998. Structure of the MscL homolog from Mycobacterium tuberculosis: a gated mechanosensitive ion channel. Science 282:2220-26

1

Wilson GG, Karlin A. 1998. The location

of the gate in the acetylcholine receptor channel. Neuron 20:1269-81 W3- Opella SJ, Marassi FM, Gesell JJ, Valente AP, Kim Y, et al. 1999. Structures of the

M2 channel-lining segments from nicotinic acetylcholine and NMDA receptors by NMR spectroscopy. Nat. Struct. Biol. 6:374-79

74. Blanton MP, Cohen JB. 1994. Identifying the lipid-protein interface of the Torpedo nicotinic acetylcholine receptor: secondary structure implications. Biochemistry 33:2859-72

US. Methot N, McCarthy MP, Baenziger JE. 1994. Secondary structure of the nicotinic acetylcholine receptor: implications for structural models of a ligand-gated ion channel. Biochemistry 33:7709-17 76. Methot N, Baenziger JE. 1998. Secondary structure of the exchange-resistant core from the nicotinic acetylcholine receptor probed directly by infrared spectroscopy and hydrogen/deutertum exchange. Biochemistry 37:14815—22 WY: Rajendra S, Lynch JW, Schofield PR. 1997. The glycine receptor. Pharmacol. Ther. 73:121-46 78. Lynch JW, Rajendra S, Pierce KD, Handford CA, Barry PH, et al. 1997. Identification of intracellular and extracellular domains mediating signal transduction in the inhibitory glycine receptor chloride channel. EMBO J. 16:110-20 3 Ryan SG, Buckwalter MS, Lynch JW, Handford CA, Segura L, et al. 1994. A mis-

sense mutation in the gene encoding the a, subunit of the inhibitory glycine receptor in the spasmodic mouse. Nature Genet. 7:131-35

80. Schofield Pierce

PR,

KD,

Lynch Handford

JW,

CA,

Rajendra

S, et al. 1996.

Molecular and genetic insights into ligand binding and signal transduction at the inhibitory glycine receptor. Cold Spring Harbor Symp. Quant. Biol. 61:333-42

81. Farrens DL, Altenbach C, Yang K, Hubbell WL, Khorana HG. 1996. Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin. Science 274:768-70 82. Lugovskoy AA, Maslennikov IV, Utkin YN, Tsetlin VI, Cohen

JB, et al. 1998.

Spatial structure of the M3 transmembrane segment of the nicotinic acetylcholine receptor@ subunit. Eur. J. Biochem. 255:455-61

83. Corbin J, Wang HH, Blanton MP. 1998. Identifying the cholesterol binding domain in the nicotinic acetylcholine receptor with [!?°I]azido-cholesterol. Biochim. Biophys. Acta 1414:65-74 84. Tobimatsu T, Fujita Y, Fukuda K, Tanaka K, Mori Y, et al. 1987. Effects of substitu-

tion of putative transmembrane segments on nicotinic acetylcholine receptor function. FEBS Lett. 222:56-62

85. Tamamizu S, Lee Y, Hung B, McNamee MG, Lasalde-Dominicci

JA. 1999. Alter-

ation in ion channel function of mouse nicotinic acetylcholine receptor by mutations in the M4 transmembrane domain. J. Membr. Biol. 170:157—64 36. Sipp! MJ. 1990. Calculation of conformational ensembles from potentials of mean force. An approach to the knowledge-based prediction of local structures in globular proteins. J. Mol. Biol. 213:859-83 87. Greenblatt EP, Meng X. 1999. A critical amino acid for halothane action. Anesthestology 91:A807

88. Guex

N,

MODEL

Peitsch

MC.

1997.

SWISS-

and the Swiss-PdbViewer:

an

48

YAMAKURA ET AL

environment for comparative protein modeling. Electrophoresis 18:2714—23

89. Sankararamakrishnan R, Sansom MS. 1995. Water-mediated conformational transitions in nicotinic receptor M2 helix bundles: a molecular dynamics study. FEBS Lett. 377:377-82 20: Franks NP, Lieb WR. 1991. Stereospecific effects of inhalational general anesthetic optical isomers on nerve ion channels. Science 254:427-30 Dik Tomlin SL, Jenkins A, Lieb WR, Franks NP. 1998. Stereoselective effects of etomidate optical isomers on y-aminobutyric acid type A receptors and animals. Anesthesiology 88:708-17 We Wittmer LL, Hu Y, Kalkbrenner M, Evers

AS,

Zorumski

CF,

et al.

1996.

Enantioselectivity of steroid-induced yaminobutyric acid, receptor modulation and anesthesia. Mol. Pharmacol. 50:1581-— 86

O3% Covey

DF,

Nilsson

Nathan

KR,

Hu

D, Kalkbrenner M, Y, et al. 2000. Enan-

tioselectivity of pregnanolone-induced yaminobutyric acid, receptor modulation and anesthesia. J. Pharmacol. Exp. Ther. 293:1009-16 94, Andrews PR, Mark LC. 1982. Structural specificity of barbiturates and related drugs. Anesthesiology 57:314—20 D5), Cordato DJ, Chebib M, Mather LE, Herkes GK,

Johnston

GA.

1999.

Stereoselec-

tive interaction of thiopentone enantiomers with the GABA, receptor. Br. J. Pharmacol. 128:77-82 96. Tomlin SL, Jenkins A, Lieb WR, Franks NP. 1999. Preparation of barbiturate opti-

cal isomers and their effects on GABA, receptors. Anesthesiology 90:1714—22

Dil Downie DL, Franks NP, Lieb WR. 2000. Effects of thiopental and its optical isomers on nicotinic acetylcholine receptors. Anesthesiology. 93:774-83 98. Allan AM, Harris RA. 1986. Anesthetic and convulsant barbiturates alter yaminobutyric acid-stimulated chloride flux

OS;

100.

101.

.

across brain membranes. J. Pharmacol. Exp. Ther. 238:763-68 Kamiya Y, Andoh T, Furuya R, Hattori S, Watanabe I, et al. 1999. Comparison of the effects of convulsant and depressant barbiturate stereoisomers on AMPAtype glutamate receptors. Anesthesiology 90:1704-13 Harris B, Moody E, Skolnick P. 1992. Isoflurane anesthesia is stereoselective. Eur. J. Pharmacol. 217:215-16 Lysko GS, Robinson JL, Casto R, Ferrone RA. 1994. The stereospecific effects of isoflurane isomers in vivo. Eur. J. Pharmacol. 263:25—29 Eger EI I, Koblin DD, Laster MJ, Schurig V, Juza M, et al. 1997. Minimum alveolar anesthetic concentration values for the enantiomers of isoflurane differ mini-

mally. Anesth. Analg. 85:188—92 3. Moody EJ, Harris BD, Skolnick P. 1993. Stereospecific actions of the inhalation anesthetic isoflurane at the GABA, receptor complex. Brain Res. 615:101-6 104. Jones MV, Harrison NL. 1993. Effects of volatile anesthetics on the kinetics of inhibitory postsynaptic currents in cultured rat hippocampal neurons. J. Neurophysiol. 70:1339-49 5. Hall AC,

Lieb WR,

Franks

NP.

1994.

Stereoselective and non-stereoselective actions of isoflurane on the GABA, receptor. Br: J. Pharmacol. 112:906-10 106. Downie DL, Hall AC, Lieb WR, Franks NP. 1996. Effects of inhalational general anaesthetics on native glycine receptors in rat medullary neurones and recombinant glycine receptors in Xenopus oocytes. Br. J. Pharmacol. 118:493-502 107. Ryder S, Way WL, Trevor AJ. 1978. Comparative pharmacology of the optical isomers of ketamine in mice. Eur. J. Pharmacol. 49:15—23 108. Klepstad P, Maurset A, Moberg ER, Oye

I. 1990. Evidence of a role for NMDA receptors in pain perception. Eur. J. Pharmacol. 187:513-18

49

MOLECULAR MECHANISMS OF ANESTHESIA

109. White PF, Schuttler J, Shafer A, Stanski DR, Horai Y, et al. 1985. Comparative pharmacology of the ketamine isomers. Studies in volunteers. Br J. Anaesth. 57:197-203

110. Zeilhofer HU, Swandulla D, Geisslinger G, Brune K. 1992. Differential effects of ketamine enantiomers on NMDA receptor currents in cultured neurons. Eur. J. Pharmacol. 213:155—58 Je Ebert B, Mikkelsen S, Thorkildsen C, Borgbjerg FM. 1997. Norketamine, the main metabolite of ketamine, is a noncompetitive NMDA receptor antagonist in the rat cortex and spinal cord. Eur. J. Pharmacol. 333:99-104 2: Yamakura T, Sakimura K, Shimoji K. 2000. The stereoselective effects of ketamine isomers on heteromeric NMDA receptor channels. Anesth. Analg. 91:225-29 iss Lodge D, Anis NA, Burton NR. 1982. Effects of optical isomers of ketamine on excitation of cat and rat spinal neurones by amino acids and acetylcholine. Neurosci. Lett. 29:281—86 114. Koblin DD, Chortkoff BS, Laster MJ, Eger EI IJ, Halsey MJ, et al. 1994. Polyhalogenated and perfluorinated compounds that disobey the Meyer-Overton hypothesis. Anesth. Analg. 79:1043-48 US). Mihic SJ, McQuilkin SJ, Eger EI I], Ionescu P, Harris RA. 1994. Potentiation

of y-aminobutyric acid type A receptormediated chloride currents by novel halogenated compounds correlates with their abilities to induce general anesthesia. Mol. Pharmacol. 46:851—17

116. Machu TK, Harris RA. 1994. Alcohols and anesthetics enhance the function of 5hydroxytryptamine3 receptors expressed in Xenopus laevis oocytes. J. Pharmacol. Exp. Ther. 271:898—905 We Cardoso RA, Yamakura T, Brozowski SJ, Chavez-Noriega LE, Harris RA. 1999. Human neuronal nicotinic acetylcholine receptors expressed in Xenopus oocytes

predict. efficacy of halogenated compounds that disobey the Meyer-Overton rule. Anesthesiology 91:1370-77 118. Raines DE. 1996. Anesthetic and nonanesthetic halogenated volatile compounds have dissimilar activities on nicotinic acetylcholine receptor desensitization kinetics. Anesthesiology 84: 663-71 i). Forman SA, Raines DE. 1998. Nonanesthetic volatile drugs obey the MeyerOverton correlation in two molecular protein site models. Anesthesiology 88:1535-48 120. Minami K, Gereau RW IV, Minami M, Heinemann SF, Harris RA. 1998. Effects

of ethanol] and anesthetics on type 1 and 5 metabotropic glutamate receptors expressed in Xenopus laevis oocytes. Mol. Pharmacol. 53:148-56

Al

Minami

K, Vanderah

TW,

Minami

M,

Harris RA. 1997. Inhibitory effects of anesthetics and ethanol on muscarinic receptors expressed in Xenopus oocytes. Eur. J. Pharmacol. 339:237-44

AD

Minami

K,

Minami

M,

Harris

RA.

1997. Inhibition of 5-hydroxytryptamine type 2A receptor-induced currents by nalcohols and anesthetics. J. Pharmacol. Exp. Ther. 281:1136—-43 IS:

Kandel L, Chortkoff BS, Sonner J, Laster MJ,

Eger EI I.

can suppress 82:321-26

1996.

Nonanesthetics

learning. Anesth.

Analg.

124. Homanics GE, Quinlan JJ, Mihalek RM, Firestone LL. 1998. Alcohol and anesthetic mechanisms in genetically engineered mice. Front. Biosci. 3:D548—58 1233 Quinlan JJ, Homanics GE, Firestone LL. 1998. Anesthesia sensitivity in mice that lack the 83 subunit of the y-aminobutyric acid type A receptor. Anesthesiology 88:775—80 126. Homanics GE, Ferguson C, Quinlan JJ, Daggett J, Snyder K, et al. 1997. Gene knockout of the @6 subunit of the yaminobutyric acid type A receptor: lack

50

YAMAKURA ET AL

of effect on responses to ethanol, pentobarbital, and general anesthetics. Mol. Pharmacol. 51:588—96 WI Mihalek RM, Banerjee PK, Korpi ER, Quinlan JJ, Firestone LL, et al. 1999. Attenuated sensitivity to neuroactive steroids in y-aminobutyrate type A receptor 5 subunit knockout mice. Proc. Natl. Acad. Sci. USA 96:12905-10 128. Zhu WJ, Wang JF, Krueger KE, Vicini S. 1996. 6 subunit inhibits neurosteroid modulation of GABA, receptors. J. Neurosci. 16:6648-56 120% Joo DT, Xiong Z, MacDonald JF, Jia Z,

1553 Harrison NL, Kugler JL, Jones MV, Greenblatt EP, Pritchett DB. 1993. Positive

136.

ES,

Roder J, et al. 1999. Blockade of gluta-

mate receptors and barbiturate anesthesia. Increased sensitivity to pentobarbitalinduced anesthesia despite reduced inhibition of AMPA receptors in GluR2 null mutant mice. Anesthesiology 91:1329-41 130. Homanics GE, Quinlan JJ, Mihalek R, Firestone LL. 1998. Genetic dissection of

138.

139.

the molecular target(s) of anesthetics with

the gene knockout approach in mice. Toxicol. Lett. 100/101:301—7

SHE

140.

Rudolph U, Crestani F, Benke D, Brunig I, Benson JA, et al. 1999. Benzodi-

azepine actions mediated by specific yaminobutyric acid, receptor subtypes. Nature 401:796—800 152) Eger EI II, Ionescu P, Laster MJ, Gong D, Hudlicky T, et al. 1999. Minimum alveolar anesthetic concentration of fluorinated alkanols in rats: relevance to theories of narcosis. Anesth. Analg. 88:867—76 3. Yamakura T, Chavez-Noriega LE, Harris RA. 2000. Subunit-dependent inhibition of human neuronal nicotinic acetylcholine receptors and other ligand-gated ion channels by dissociative anesthetics

141.

142.

143.

ketamine and dizocilpine. Anesthesiology 92:1144-53 134. Ueno S, Trudell JR, Eger EI II, Harris RA. 1999. Actions of fluorinated alkanols on GABA, receptors: relevance to theories of narcosis. Anesth. Analg. 88:877—83

144.

modulation of human yaminobutyric acid type A and glycine receptors by the inhalation anesthetic isoflurane. Mol. Pharmacol. 44:628—32 Lees G, Edwards MD. 1998. Modulation of recombination human y-aminobutyric acid, receptors by isoflurane: influence of the 6 subunit. Anesthesiology 88:206— 17 Jenkins A, Franks NP, Lieb WR. 1999. Effects of temperature and volatile anesthetics on GABA, receptors. Anesthesiology 90:484-91 Lin LH, Whiting P, Harris RA. 1993. Molecular determinants of general anesthetic action: role of GABA, receptor structure. J. Neurochem. 60:1548-53 Ymer S, Draguhn A, Wisden W, Werner P, Keinanen K, et al. 1990. Structural and functional characterization of the y ; subunit of GABA,/benzodiazepine receptors. EMBO J. 9:3261—-67 Horne AL, Harkness PC, Hadingham KL, Whiting P, Kemp JA. 1993. The influence of the y>, subunit on the modulation of responses to GABA receptor activation. Br. J. Pharmacol. 108:711-16 Sanna E, Garau F, Harris RA. 1995. Novel properties of homomeric £1 yaminobutyric acid type A receptors: actions of the anesthetics propofol and pentobarbital. Mol. Pharmacol. 47:213-17 Puia G, Santi MR, Vicini S, Pritchett DB, Purdy RH, et al. 1990. Neurosteroids act on recombinant human GABA, receptors. Neuron 4:759-65 Shingai R, Sutherland ML, Barnard EA. 1991. Effects of subunit types of the cloned GABA, receptor on the response to a neurosteroid. Eur. J. Pharmacol. 206:77-80 Horne AL, Hadingham KL, Macaulay AJ, Whiting P, Kemp JA. 1992. The pharmacology of recombinant GABA j receptors containing bovine a, 61, vy, sub-units

MOLECULAR MECHANISMS

stably transfected into mouse _fibroblast L-cells. Br J. Pharmacol. 107:73237 145. Thompson SA, Whiting PJ, Wafford KA. 1996. Barbiturate interactions at the human GABA, receptor: dependence on receptor subunit combination. Br. J. Pharmacol. 117:521—27 146. Sanna E, Murgia A, Casula A, Biggio G. 1997. Differential subunit dependence of the actions of the general anesthetics alphaxalone and etomidate at y-aminobutyric acid type A receptors expressed in Xenopus laevis oocytes. Mol. Pharmacol. 51:484-90 147. Lambert JJ, Belelli D, Hill-Venning C, Callachan

H, Peters JA.

1996. Neuros-

teroid modulation of native and recombinant GABA, receptors. Cell. Mol. Neurobiol. 16:155—74

148. Sanna E, Mascia MP, Klein RL, Whiting PJ, Biggio G, et al. 1995. Actions of the general anesthetic propofol on recombinant human GABA, receptors: influence of receptor subunits. J. Pharmacol. Exp. Ther. 274:353-60 149. Flood P, Krasowski MD. 2000. Intravenous anesthetics differentially modulate ligand-gated ion channels. Anesthestology 92:1418—25 150. Krasowski MD, O’Shea SM, Rick CE, Whiting PJ, Hadingham KL, et al. 1997.

a Subunit isoform influences GABA, receptor modulation by propofol. Neuropharmacology 36:941-49 Ilisyile Koltchine VV, Ye Q, Finn SE, Harrison NL. 1996. Chimeric GABA 4/glycine receptors: expression and barbiturate pharmacology. Neuropharmacology 35: 1445-56

2.

Jones

MV,

OF ANESTHESIA

Harrison

51

NL, Pritchett DB,

Hales TG. 1995. Modulation of the GABA, receptor by propofol is independent of the y subunit. J. Pharmacol. Exp. Ther. 274:962-68

153. Rick CE, Ye Q, Finn SE, Harrison NL. 1998. Neurosteroids act on the GABA, receptor at sites on the N-terminal side of the middle of TM2. Neuroreport 9:379-

83 154. Belelli D, Callachan H, Hill-Venning C, Peters JA, Lambert JJ. 1996. Interaction

of positive allosteric modulators with human and Drosophila recombinant GABA receptors expressed in Xenopus laevis oocytes. Biz J. Pharmacol. 118:563— 76 S15). Shimada S, Cutting G, Uhl GR. 1992. y-Aminobutyric acid A or C receptor? y-Aminobutyric acid p, receptor RNA

induces bicuculline-,

barbiturate-,

and benzodiazepine-insensitive y-aminobutyric acid responses in Xenopus oocytes. Mol. Pharmacol. 41:683-87 156. Dilger JP, Boguslavsky R, Barann M, Katz

T,

Vidal

AM.

1997.

Mecha-

nisms of barbiturate inhibition of acetylcholine receptor channels. J. Gen. Physiol. 109:401-14 157. Wachtel RE, Wegrzynowicz ES. 1992. Kinetics of nicotinic acetylcholine ion channels in the presence of intravenous anaesthetics and induction agents. Br. J. Pharmacol. 106:623—27 158. Yamakura T, Sakimura K, Shimoji K, Mishina M. 1995. Effects of propofol on various AMPA-, kainate- and NMDAselective glutamate receptor channels expressed in Xenopus oocytes. Neurosci. Lett. 188:187—90

> jy |

eerie

rec

7

hieq

es

- ev

>

sO

- om

|

eag

a

D

_

aa

ADfb
6 additional alleles)

Figure 2 Patients who inherit TPMT deficiency or heterozygocity accumulate excessive cellular concentrations of active thioguanine nucleotides (TGN), predisposing them to severe hematopoietic toxicity. However, reducing the dosage of thiopurines (i.e. mercaptopurine, azathioprine, thioquanine) in TPMT-deficient or -heterozygous patients permits thiopurine therapy without acute toxicity. Molecular diagnostics, based on TPMT genotype, can now be used to prospectively identify TPMT-deficient patients, minimizing the risk of dose-limiting toxicity.

an autosomal codominant trait. Patients who inherit TPMT deficiency accumulate excessive cellular concentrations of TGN, predisposing them to hematopoietic toxicity (Figure 2), which can be fatal (16). The molecular basis for polymorphic TPMT activity has now been defined for the majority of patients. Whereas 8 TPMT alleles have been identified, 3 alleles (TPMT*2, TPMT *3A, TPMT*3C) account for about 95% of intermediate or low enzyme activity cases (Figure 2) (12, 13). The mutant allele TPMT*2 is defined by a single nucleotide transversion (G238C) in the open reading frame, leading io an amino acid substitution at codon 18 (Ala>Pro) (17).

TPMT*3A

contains

two nucleotide transition mutations (G460A and A719G) in the open reading frame, leading to amino acid substitutions at codon 154 (Ala>Thr) and codon 240 (Tyr>Cys) (18), whereas TPMT *3C contains only the A719G transition mutation (18,19). All three alleles are associated with lower enzyme activity, owing to enhanced rates of proteolysis of the mutant proteins (20). By using allele-specific PCR or PCR-RFLP to detect the three signature mutations in these alleles, a rapid

and relatively inexpensive assay is available to identify >90% of all mutant alleles (21). In Caucasian populations, TPMT*3A is the most common mutant 7PMT allele (3.2—-5.7% of TPMT alleles), whereas TPMT *3C has an allele frequency of 0.2-0.8% and TPMT*2 represents 0.2—0.5% of TPMT alleles (13, 21). Studies in

106

McLEOD

® EVANS

Caucasian, African, and Asian populations have demonstrated the broad utility of this approach (22-24) while revealing that the frequency of these mutant TPMT alleles differs among various ethnic populations. For example, East and West African populations have a frequency of mutant alleles similar to that of Caucasians, but all mutant alleles in the African populations are TPMT *3C (22). Among African Americans,

TPMT *3C is the most prevalent allele, but TPMT *2

and TPMT *3A are also found, reflecting the integration of Caucasian and African American genes in the US population (23). In Asian populations, TPMT*3C is the predominant mutant allele (100% of mutant alleles in published studies to date). The presence of TPMT *2, TPMT *3A, or TPMT *3C is predictive of phenotype; patients heterozygous for these alleles all have intermediate activity, and subjects homozygous for these alleles are TPMT deficient (21, 23). In addition, compound heterozygotes (TPMT *2/3A, TPMT*3A/3C) are also TPMT deficient, as would be expected (21). Whereas most studies have used erythrocytes as a surrogate tissue for measuring TPMT activity, studies have also shown that TPMT genotype determines TPMT

activity in leukemia cells (15,25), as would be expected for

germline mutations. The enthusiasm for TPMT pharmacogenetics has been further stimulated by the finding that TPMT genotype identifies patients who are at risk of toxicity from mercaptopurine or azathioprine. Numerous studies have shown that TPMTdeficient patients are at very high risk of developing severe hematopoietic toxicity when treated with conventional doses of thiopurines (26, 27). More recent studies have shown that patients who are heterozygous at the TPMT gene locus are at intermediate risk of dose-limiting toxicity (28-30). In a study of 67 patients treated with azathioprine for rheumatic disease, six patients (9%) were heterozygous for mutant TPMT alleles (28), and therapy was discontinued in five of the six patients because of low leukocyte count within one month of starting treatment. The sixth patient had documented noncompliance with azathioprine therapy. Patients with wild-type TPMT received therapy for a median of 39 weeks without complications compared with a median of 2 weeks in patients heterozygous for mutant TPMT alleles (28). A second study in patients with Japanese rheumatic disease receiving azathioprine recently confirmed the importance of a heterozygous TPMT genotype for predicting systemic toxicity (29). Futhermore, Relling et al (30) showed that TPMT-deficient patients tolerated full doses of mercaptopurine for only 7% of scheduled weeks of therapy, whereas heterozygous and homozygous wild-type patients tolerated full doses for 65% and 84% of scheduled weeks of therapy, respectively, over the 2.5 years of treatment. The percentage of weeks in which mercaptopurine dosage had to be decreased to prevent toxicity was 2%, 16%, and 76% in wild-type, heterozygous, and homozygous mutant individuals, respectively (30). Collectively, these studies demonstrate that the influence of TPMT genotype on hematopoietic toxicity is most dramatic for homozygous mutant patients, but is also of clinical relevance for heterozygous individuals, who represent about 10%

of patients treated with these medications. TPMT deficiency has also been linked to a higher risk of second malignancies among patients with acute lymphoblastic

PHARMACOGENOMICS

leukemia,

including topoisomerase-inhibitor-induced

107

acute myeloid leukemia

(31, 32) and radiation-induced brain tumors (33). Therefore, prospective knowl-

edge of a patient’s TPMT status permits patient-specific dosages that reduce the risk of acute toxicity from thiopurine medications (Figure 2) and may identify those at higher risk of second malignancies.

Polymorphic Drug Targets Genetic polymorphism of the 62-adrenoreceptor exemplifies a clinically relevant polymorphism in a drug target (34). The 62-adrenoreceptor is a G protein-coupled receptor that interacts with endogenous catecholamines and various medications. These receptors are widely distributed and play an important role in regulating cardiac, vascular, pulmonary, and metabolic functions (34). Studies of such physiologic functions of 62-adrenoreceptor in humans have revealed substantial interpatient variation in receptor function and responsiveness to stimulation. In the heart, activation of 62-adrenoreceptor results in an increased rate and force of cardiac muscle, whereas 62-adrenoreceptor stimulation in the lungs acts to relax airway smooth muscie. Influences on lipolysis in subcutaneous fat have also been described, possibly through regulation of lipid mobilization, energy expenditure, and glycogen breakdown. Understanding the molecular basis for variability in the 2-adrenoreceptor has recently been assisted by the identification of five distinct single nucleotide polymorphisms, each associated with altered expression, down regulation, or coupling of the receptor (34). Alteration at amino acid 16 (Arg>Gly) appears to have relevance in pulmonary disease, with patients homozygous for Arg exhibiting a greater response to 62 agonist medications (35, 36). For example, the FEV, response to oral albuterol was 6.5-fold higher in patients with an Arg/Arg genotype at codon 16 compared with Gly/Gly patients, even though similar plasma drug concentrations were achieved (35). In contrast, the alteration at codon 27 (Gln>Glu) does not appear to influence lung function, but there is an association between the Gln/Gln genotype and an increased incidence of obesity (37, 38). This relationship appeared to be more prominent in men and could be overcome with exercise (38). The mutant allele for codon 16 (frequency 0.61) and codon 27 (fre-

quency 0.43) are relatively common and are therefore under intensive investigation for their clinical relevance. A less common ailele contains a mutation at Codon 164 (Thr>Ile), with a mutant allele frequency of 0.05. The clinical significance of this polymorphism was identified in patients with heart failure: A 42% one-year survival was observed in patients with the Thr/Ile genotype compared with 76% in patients with Thr/Thr (39). This finding led to the suggestion that patients with the Ile164 polymorphism and heart failure should be considered as candidates for early aggressive intervention or cardiac transplantation. More recent findings indicate that 82 receptor haplotype is more informative than individual SNPs in predicting response to beta agonists in asthmatics (39a). Considerable variation in patient response to therapy can also be observed in clinical trials. Understanding the mechanistic basis for differences in drug

108

McLEOD

® EVANS

response can be used to identify disease phenotypes to which specific therapy should be directed. An example of this approach was seen in the treatment of Alzheimer’s disease, where 83% of patients without an apoE4 genotype had an improvement in total response and cognitive response to tacrine therapy, compared with 40% in patients with apoE4 (40). The specific interaction between the apolipoprotein genotype and tacrine therapy has not been elucidated, but this association suggests that apoE4 plays a role in cholinergic dysfunction in Alzheimer’s disease, in a way that cannot be overcome by therapy with acetylcholinesterase inhibitors such as tacrine. Although this relationship between apolipoprotein genotype and improvement in Alzheimer’s disease needs to be confirmed, it provides a putative genetic approach for selecting therapy for this disorder. Although the above examples are illustrative of clinically relevant single nucleotide polymorphism (SNP), many genes with a putative role in the regulation of drug activity do not have clearly defined genetic polymorphisms associated with drug response or disease phenotypes. Therefore, considerable time and money are currently being invested in the production of large libraries of single nucleotide polymorphisms (41) that can be further investigated for an association with drug response. This includes nonprofit ventures (e.g. The SNP Consortium) that release all information to the public free of charge and private SNP efforts from a number of biotech companies (e.g. Genset, Celera Genomics, Incyte). SNPs are the most abundant type of DNA sequence variation in the human genome, with an estimated frequency of 1 in 1000 bases (11,42). A SNP is a site on the DNA in which a single base pair varies among individuals in a population. If a SNP is found within a small, unique segment of DNA, it serves as both a physical landmark and as a genetic marker whose transmission can be followed from parent to child. According to theoretical models, if the genotype of a group of individuals with a common disease and a group without the disease are studied, certain genotypes may be consistently associated with those individuals who have the disease (41). Owing to linkage disequilibrium, alleles of genetic markers in close proximity to a disease-modifying mutation are often found to be associated with the disease, even though they themselves are not involved in disease pathogenesis or drug response (41). Once localized, these specific chromosomal regions can be analyzed further to identify disease-associated genes and mutations. This molecular/population genetic approach also provides a strategy to identify genes associated with other phenotypes, such as drug toxicity or therapeutic benefit. This approach can be used for genome-wide mapping in which no a priori genes or genomic regions are assumed to be associated with the drug effect under investigation. The number of subjects and the numbers of markers needed for such a study depend on the level of contribution of the specific locus to the complex trait (e.g. a single causative mutation is easier to find than an alteration that is one of several contributors to a phenotype). It is estimated that 60,000 markers, at 50-kb spacing, will be needed to blanket the genome in an association study with 1,000 individuals (e.g. 500 patients with toxicity and 500 patients tolerating therapy). If 1,000 individuals are

PHARMACOGENOMICS

109

to be genotyped with 60,000 markers, 60,000,000 genotyping assays will have to be completed for each study. This requires a dramatic advance in high throughput genotyping techniques for this approach is to be used in a timely and cost-efficient manner. An alternative approach (Figure 3) uses an educated guess as to which of the 100,000 genes in the human genome are likely to be important contributors to the clinical phenotype (43), then a search for informative polymorphisms in these genes. This is especially useful for classes of agents with clearly defined biochemistry, allowing candidate gene selection, as exemplified by the recent preliminary studies of clozapine response in schizophrenia (43a). This candidate gene approach substantially reduces the number of loci under evaluation, but will miss genes with no anticipated role in the drug’s in vivo activity. It is through efforts such as these that the next wave of pharmacogenetic predictive tools will emerge, requiring extensive in vitro and in vivo functional analyses to determine the role of each specific SNP in selecting optimal drug therapy.

Pharmacogenetic Discovery Pre-Genomics

Post-Genomics

Drug

Genotype

wa

2: 0'0'0'0'e

‘

| st ee

—-} |

5

a

‘=

et — a= Sil pel TBa) ee

»

5

Genotype

Compare phenotypes

tt

in genotypic groups

Compare genetic polymorphisms in phenotypic groups Figure 3 The evolving paradigm for discovery of genetic polymorphisms associated with aberrant drug disposition or effects. In the pregenomics era (e.g. 1950-2000) discoveries were most often made after an unusual phenotype was observed and family studies established an inherited basis, eventually leading to discovery of the genetic basis of the inherited phenotype. In the postgenomics era many more discoveries will begin with elucidation of genetic polymorphisms in candidate genes (e.g. those known to be involved in the metabolism, transport, or targets of the candidate medication), and then large population studies will identify links between these gene polymorphisms and drug effects in patients. (Adapted from MV Relling, personal communication.)

110

McLEOD

#® EVANS

High-Density Pharmacogenomics Although a candidate gene approach is applicable to agents with clearly defined mechanisms of action, metabolism, and/or toxicity, the pathways influencing anew agent are often unknown. In that circumstance, a genome-wide approach with no a priori assumptions for loci of interest may be the only option. An alternative to low-density genomic approaches, such as SNP analysis, is comparative genomic hybridization (CGH). CGH involves a competitive in situ hybridization of fluorescently labelled test (e.g. tumor) and control (e.g. normal tissue) DNA onto normal metaphase chromosomes from an unrelated healthy donor (44). Computer-assisted fluorescence microscopy is then used to assess the intensity of signal across each chromosome. The differences in test and control fluorescence intensity reflect the change in DNA amount for specific regions of the human genome. If chromosomes or chromosomal subregions are present in identical quantity within both test and control DNA, an equal contribution from each fluorochrome is seen. However, a change in the fluorescent signal is seen if certain chromosomal subregions are gained or lost in the test DNA. CGH is applicable to DNA only and has primarily been used in the context of tumor biology, identifying novel alterations associated with the acquisition of cancer (45). With current technology, CGH has a theoretical limit of detection for gain and loss of genetic material of S-10 Mb. However, gain of DNA in regions as small as 5O kb have been described in situations in which high-level amplification has occurred (44). CGH has been used to evaluate a genomic basis for resistance to anticancer chemotherapy. The genomes of cell lines resistant to raltitrexed and 5-fluorouracil, both antimetabolite anticancer agents that inhibit the thymidylate synthase enzyme, were compared with that found in the corresponding parent, sensitive cell line (45). The genome of cell lines resistant to raltritrexed differed from the sen-

sitive cell lines by only a gain of part of the small arm of chromosome 18 (45). This region was subsequently shown to be an amplification of the thymidylate synthase gene on chromosome 18p11.32 (45). The 5-fluorouracil-resistant cell line also had chromosome 18p gain (45). However, several additional regions were associated with drug resistance, including gain of chromosomes 7p11.1-22 and 6p23-25 and loss of chromosomes 2p11.2 and 9p22-q31 (45). These data are especially interesting in the context of our current understanding of raltitrexed and 5-fluorouracil mechanisms of action. Raltitrexed is a thymidylate synthase— specific agent, with no known alternative mechanisms of action. 5-fluorouracil inhibits thymidylate synthase, but also affects RNA and DNA synthesis through false base incorporation (46). This study has generated chromsome loci that will be used to identify specific genes influencing drug resistance to 5-fluorouracil. Similar findings have been generated for cell lines resistant to the anticancer agent

cisplatin (47). Although the above studies demonstrate proof of principle for use of CGH in pharmacogenomics, this technique now needs to be applied in the context of agents with unknown mechanisms of action.

McLEOD

#® EVANS

C-1

All patients with same diagnosis

Remove

(?

non-responders

ty pede and toxic responders

Treat Responders and Patients Not Predisposed to Toxicity

Figure 1 Pharmacogenomics has the potential to identify the subsets of patients who are genetically predisposed to toxicity from specific medications and those who are likely not to respond. Predisposition to toxicity can occur because of an inherited deficiency in drug metabolism, while mutations in drug receptors can alter a patients response to medications. The subset of patients who are identified as toxic responders or nonresponders would be treated with different dosages or alternative medications.

PHARMACOGENOMICS

111

Genome-wide analysis of copy number has recently been applied to array-based assays to allow a more automated and rapid CGH approach. The published studies confirm the viability of such an approach, with high resolution mapping of regions of loss and gain in human breast cancer (48). The CGH array approach is limited in the extent of the genome that is evaluable on a single chip and has only been applied to specific chromosomal regions (e.g. chromosome 20) or with superficial coverage of the genome (48, 49). CGH arrays have not yet been used in the context of pharmacogenomics, and CGH’s ability to detect only large changes in chromosomal structure represents a substantial limitation of the methodology.

Gene Expression The development of glass and nylon membrane microarrays has revolutionized the way gene expression is evaluated in all areas of medicine, including pharmacology. Initial studies have focused on gene expression along biologic pathways and have provided an increased understanding of the regulation of cellular proliferation and the cell’s response to nutrient stimulation (50). More recently, gene expression arrays have been used in the molecular classification of disease and have highlighted the great genetic heterogeneity among cells with histologically similar appearance (51,52). For example, Alizadeh et al (51) set out to characterize gene expression in

diffuse large B-cell lymphoma (DLBCL), selected because of clinical heterogeneity: 40% of patients respond well to current therapy, whereas the remainder die of disease. By using a “lymphochip” containing 17,856 genes that are preferentially expressed in lymphoid cells, the investigators demonstrated the presence of two molecularly distinct forms of the disease: germinal center B-like DLBCL and activated B-like DLBCL. More important, patients with germinal center B-like DLBCL had a superior overall survival than those with activated B-like DLBCL (5-year survival of 76% versus 16%, respectively) (51). This study provides substance to the proposed use of gene expression arrays as a prospective tool for individualizing patient therapy. Based on the results of Alizadeh, patients with activated B-like DLBCL will not benefit from standard therapy, and experimental treatment approaches should be considered. Alternatively, patients with germinal center B-like DLBCL may be currently “over treated” because of their “good risk” status, and treatment strategies with more manageable side-effect profiles may need to be considered. However, not all evaluations of gene expression array technology have demonstrated usefulness for therapeutics. A molecular evaluation of acute leukemia demonstrated the feasibility of cancer classification based solely on gene expression monitoring rather than the traditional histopathological evaluation (52). The molecular approach is less subjective and not as cumbersome as pathologist-based approaches. This approach was able to differentiate between acute myeloid and acute lymphoid leukemia, and between B-lineage and T-lineage acute lymphoid leukemia within the latter group. However, no strong gene expression signature was evident in this study for those patients achieving

112

McLEOD

® EVANS

disease remission after induction chemotherapy. This should not be surprising, in that the activity of chemotherapy is a dynamic process in which a series of events must occur down specific molecular pathways. The inability of static measures of a “snapshot” of gene expression to satisfactorily predict therapeutic response is not unreasonable. Indeed, evaluation of the patterns of gene expression during serum starvation was only informative when considered in the context of cluster evaluation of data over a time course (53).

The poor predictive power of “static” array analysis in some studies (52) could also reflect the limited number of genes tested, rather than a deficiency of the approach. Indeed, this conclusion is supported by the recent evaluation of the expression of 8000 genes in the National Cancer Institute panel of 60 human cancer cell lines, which has been used over the past 10+ years to test the cytotoxicity of over 70,000 putative anticancer agents (54). In this analysis, gene expression patterns were used to evaluate the relationship between drug-activity patterns and mechanism of action and to assess gene-drug activity correlations for predictive purposes. For example, the antimetabolite chemotherapy agent 5fluorouracil is known to be degraded by dihydropyrimidine dehydrogenase (DPD) (55). High DPD would be expected to decrease exposure of cells to the active form of 5-fluorouracil. Consistent with this hypothesis, a significant negative correlation between DPD mRNA expression and 5-fluorouracil potency was observed among the NCI cell line panel (r = —0.53) (54). Most cell lines with low DPD mRNA were sensitive to S-fluorouracil (14/18), and all seven of the colon cancer

cell lines were in this category. This observation is consistent with the clinical use of 5-fluorouracil as an active agent in colorectal cancer (55).

The utility of

this pharmacogenomic approach for selection of therapy for specific tumor subtypes was further illustrated with the amino acid depletion agent L-asparaginase. This agent takes advantage of the lack of asparagine synthetase in some malignant cells, making them dependent on exogenous L-asparagine (54). Overall, a correlation between asparagine synthetase mRNA expression and L-asparaginase cytotoxicity was observed (r = —0.44). However, further examination noted a much stronger relationship between asparagine synthetase MRNA expression and L-asparaginase cytotoxicity in the 6 leukemia cell lines (r = —0.98) than in the other cell lines (r = —0.32) (54). These findings are consistent with the activity of L-asparaginase in acute leukemia and support the use of asparagine synthetase expression as a predictive marker for guiding use of this agent. These data can be explored further by individual investigators using the NCI Drug Discovery Website (http://dtp.nci.nih.gov/). Many additional publicly and privately funded studies of gene expression and drug sensitivity are being performed and will provide the basis for prospective studies of prognostic prediction and characterization studies of drug-activity relationships. Gene expression array analysis may allow investigators to begin to qualitatively define the elusive “therapeutic index” for specific agents. Both static and dynamic approaches of analyzing gene expression in normal and disease tissues will allow mechanism of action and mechanism of toxicity to be clarified. This will enable

PHARMACOGENOMICS

113

enlightened strategies for modulation of therapy, new agent design, or tissue targeting to be developed, based on direct in vivo observations, rather than theory alone.

Drug Development Growth in the field of pharmacogenomics has been heavily influenced by the pharmaceutical industry and its desire for a “smarter” drug development process. The potential applications of pharmacogenomics extends from identification of novel targets against which new therapies are designed to tools for predicting efficacy or toxicity during clinical development (1). Pharmacogenomics also has the potential to make the drug development process more efficient, by decreasing the number of patients required to show efficacy in early clinical trials (56). Human and mouse SNP projects are being utilized in an attempt to find specific genes or genomic loci that are associated with the disease of interest. Similar approaches are being conducted using gene expression arrays. Disease tissue is used to produce mRNA for comparison with normal reference tissue. The goal of this approach is gene hunting, and arrays covering the broadest range of known and unknown genes are desired. One goal of the SNP and gene array hunting exercise is the identification of novel targets for therapy. These can be putative modulators of the disease phenotype or new mechanisms of disease. After identification of the target, a great deal of effort must be expended to confirm the viability of the target, in terms of normal-disease tissue expression, pattern of normal tissue expression for toxicity prediction, and frequency of expression in the disease tissue. There are not yet any published examples detailing the efficiency and success rate of such an approach in the early drug development process. Gene expression arrays are also being applied to define the mechanism of action for new compounds or to screen for direct influence of an agent on a specific pathway. Even agents developed in the most mechanistically based program can display surprises during in vivo evaluation. For example, inhibitors of HMG-CoA reductase, used to control cholesterol levels, were subsequently found to inhibit farnysyl transferase activity in the cell signalling pathway of the oncogenetic ras (57). By using expression arrays, a profile of the genes altered after drug exposure can be generated and may thereby yield a greater understanding of mechanisms of action. Gene expression arrays can also be used during screening of candidate compounds. By constructing arrays for genes involved in a pathway of interest, in vitro or even in vivo gene dynamics can be used as a functional readout for drug activity. There is now a rapidly growing effort to identify SNPs that will be useful for identifying patients who are likely to benefit from a specific agent or those likely to experience unacceptable toxicity. Examples of TPMT or £2 adrenoreceptor SNPs predicting risk of toxicity or outcome are detailed above. In addition, a SNP in the ALOX5 gene promoter has been found to be associated with the antiasthma efficacy of inhibitors of 5-lipoxygenase, thereby altering leukotriene production (58). There are numerous other genetic polymorphisms in drug-metabolizing enzymes, transporters, and targets, a compilation of which

114

McLEOD

® EVANS

can be found at www.science.org/features/data/1044449.shl. Most, if not all, pivotal phase II/III studies now include collection of blood for genomic DNA, in order to subsequently evaluate whether SNPs can provide a more intelligent use of the medication under evaluation.

Pharmacogenomics as a Public Health Tool Although the promise of pharmacogenomics is enormous, it is likely to have the greatest initial benefit for patients in developed countries, owing to expense, availability of technology and the focus of initial research. However, pharmacogenomics should ultimately be useful to world populations. There is clear evidence of ethnic variation in disease risk, disease incidence, and response to therapy (59). In addition, many polymorphic drug metabolizing enzymes have qualitative and quantitative differences among racial groups (59). For example, the COMT low activity allele is less frequent in African and East Asian populations (60). Because COMT inactivates methyldopa, one of the most commonly prescribed antihypertensive medications in those regions of the world, this has important potential implications (61). In addition, COMT influences the activity of levodopa for Parkinson’s disease and the production of estrogen metabolites associated with breast cancer. One approach to applying pharmacogenomics to public health is through SNP allele frequency analysis in defined populations. For example, TPMT genotype in world populations suggests that TPMT-mediated toxicity from azathioprine or mercaptopurine would be lower in Japanese or Chinese populations than Caucasians (13). In contrast, a higher mutant allele frequency was found in the Ghanaian and Kenyan populations (13). In addition, further analysis of the major tribes of Ghana found distinct differences in TPMT allele frequency, ranging from 9.9% heterozygotes in the Ewe population to 13.8% in Fanti individuals (MM Ameyaw & HL McLeod, submitted). Even greater ethnic differences have been established for

other polymorphic drug-metabolizing enzymes (e.g. NAT2, CYP2D6, CYP2C19), and this will likely be the case for most pharmacogenomic traits, including drug transporters and targets. This general approach needs to be more extensively evaluated, but does offer the potential for generating information that will have broad application to the development of clinical practice guidelines and national formularies in developing countries. Although using knowledge of ethnic differences may be relevant to much of the world’s populations, it is significantly limited in places with extensive genetic mixing. For example, it is well known that the African American population has a great degree of geographic and social mixing that provide a basis for genetic heterogeneity. This is illustrated in evaluation of TPMT mutations between African American and West African populations. Although the TPMT *3C allele was the most frequently observed variant in both populations, it represented 100% of West African mutant alleles and 52% of African American mutant alleles (22,23). The remaining African American mutant alleles were TPMT*2 and TPMT*3A @s))!

PHARMACOGENOMICS

115

alleles that are common in Caucasians. Therefore, great care must be made when applying pharmacogenomics to public health issues, and testing at the genetic level in each patient will remain the most definitive approach.

NONSCIENTIFIC CHALLENGES FOR PHARMACOGENOMICS There are a number of issues influencing the development of pharmacogenomics, including many that are of a practical or nonscientific nature. An important limitation to the wide application of pharmacogenomics is the availability of gene expression arrays, high throughput genotyping, and informatics. Currently, there is considerable growth in the number of companies offering both genomics analysis on a fee-for-service basis and the equipment for user-maintained instruments. As technology and competition bring down the high initial capital costs of array and genotype systems, the potential for general application of these approaches will be further enhanced. A related, and unanswered, question is how much can pharmacogenomics analysis cost and still be a viable adjunct to current medical practice? Currently, the technology for gene expression and genotype assessment is only affordable in the research and development setting or in the context of funded research. Thoughtful pharmacoeconomic analysis is needed to justify and direct the further development of pharmacogenomics for rational therapeutics. On the positive side, once a panel of genotypes has been correctly determined for a given individual, they need not be repeated. It is anticipated that a secured, patient-specific database will be established for each person, into which additional results will be deposited as additional genotypes are determined. This potentially web-based compilation of an individual’s established genotypes would then be available to authorized healthcare providers for the selection of optimal therapy for the treatment or prevention of diseases. Finally, the ethics of genetic analysis is currently under avid discussion and debate. Previously, a system of trust and internal control was utilized to prevent inappropriate use of genetic information. This approach has been very successful, with breach of trust being a rare event. However, the field of bioethics is now focusing on prevention of potential or theoretical abuses of genetic information against individuals. This has led to questions about what information is needed, who should have access to the data, and how they should be used. Issues such as

these are deeply challenging, as the insurance carrier paying for genetic testing is the same entity that could use the information to identify disease or therapy risks that could be used to restrict future coverage. However, the great potential gains from pharmacogenomics, in terms of both patient well-being and cost of healthcare, heavily outweigh the risks. Putting such powerful information in the hands of knowledgeable healthcare providers and those involved in the discovery of new approaches to disease treatment or prevention offers so much promise that society

116

McLEOD

TABLE 1

® EVANS

Examples of pharmacogenomic influences on drug activity*

Gene

Drug

Effect

Reference

£2 adrenoreceptor

Albuterol

Response (FEY) in asthmatics

34, 35, 39a 62-64

5-lipozygenase promoter

ABT-761

Response in asthmatics

58

Enalapril, lisinopril, captopril

Renoprotective effects, cardiac indices, blood pressure, IgA nephropathy

65-69

Quinidine

Drug-induced long QT syndrome Drug-induced torsade de pointes Drug-induced long QT syndrome

74

Drug-induced arrhythmia

Vie

(zileuton)

Angiotensin-converting enzyme (ACE)

70-72

Potassium channels

HERG

Cisapride Terfenadine, Disopyramide, meflaquine Clarithromycin

KvLQT1

hKCNE2

73

ApoE4

Tacrine

Response in Alzheimer’s disease

40

TPMT

Azathioprine, mercaptopurine, thioguanine

Hematopoietic toxicity

12, 28, 30

CYP2C9

Warfarin

Anticoagulant effects

76

— eS “Additional examples can be found at www.science.org/feature/data/ 1044449 shl

must find a way to ensure that inappropriate exploitation does not preclude the vast public good that will emerge from the burgeoning field of pharmacogenomics. ACKNOWLEDGMENTS

Work reported in this chapter was supported in part by The Alvin J. Siteman Cancer Center; The St. Jude Cancer Center Support Grant CA21765: NIH grants NIH R37 CA36401, RO1 CA78224, U01 GM 61393; a Center of Excellence grant from the State of Tennessee; and the American Lebanese Syrian Associated Charities. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Evans WE, Relling MV. 1999, Pharmacogenomics: translating functional genomics into rational therapeutics. Science 286:487— 91 2. Eisen MB, Brown PO. 1999. DNA arrays for

analysis of gene expression. Methods Enzymol. 303:179-205 3. So CW, Caldas C, Liu MM, Chen SJ, Huang QH, et al. 1997. EEN encodes for a member of a new family of proteins

PHARMACOGENOMICS containing an Src homology 3 domain and is the third gene located on chromosome 19p13 that fuses to MLL in hvuman leukemia. Proc. Natl. Acad. Sci. USA 94:2563-68 . Chen XN, Levine L, Kwok PY. 1999. Fluo-

rescence polarization in homogeneous nucleic acid analysis. Genome Res. 9:492-98 . Germer S, Holland MJ, Higuchi R. 2000. High-throughput SNP allele-frequency determination in pooled DNA samples by kinetic PCR. Genome Res. 10:258-66 . Griffin TJ, Smith LM. 2000. Singlenucleotide polymorphism analysis by MALDI-TOF mass spectrometry. Trends Biotechnol. 18:77-84 . Kwok PY. 1998. Genotyping by mass spectrometry takes flight. Nat. Biotech

16:1314—-15 . Iannone MA, Taylor JD, Chen JW, Li MS, Rivers

P, et al. 2000.

Multiplexed

sin-

gle nucleotide polymorphism genotyping by oligonucleotide ligation and flow cytometry. Cytometry 39:131-40 . Kuklin A. 1997. Detection of singlewith the nucleotide polymorphisms WAVE(TM) DNA _ fragment analysis system. Genet. Test. 1:201—6 10. Syvanen AC. 1999. From gels to chips: “minisequencing” primer extension for analysis of point mutations and single nucleotide polymorphisms. Hum. Mutat.

13:1-10 ite Marth GT, Korf I, Yandell MD, Yeh RT, Gu Z, et al. 1999. A genera] approach to singlenucleotide polymorphism discovery. Nat. Genet. 23:452-56 WR Krynetski EY, Evans WE. 1998. Pharmacogenetics of cancer therapy: getting personal. Am. J. Hum. Genet. 63:11—16 13 McLeod H, Krynetski E, Relling MV, Evans WE. 2000. Genetic polymorphism of thiopurine methyltransferase and its clinical relevance for childhood acute lym-

phoblastic leukemia. Leukemia 14:567a2, 14. Weinshilboum RM, Sladek SL. 1980. Mer-

117

captopurine pharmacogenetics: monogenic inheritance of erythrocyte thiopurine methyltransferase activity. Am. J. Hum. Genet. 32:651-62 ay. McLeod HL, Relling MV, Liu Q, Pui C-H, Evans WE. 1995. Polymorphic thiopurine methyltransferase in erythrocytes is indicative of activity in leukemic blasts from children with acute lymphoblastic leukemia. Blood 85:1897-902 16. Schutz E, Gummert J, Mohr FE, Ollerich M. 1993. Azathioprine-induced myelosuppression in thiopurine methyltransferase deficient heart transplant recipient. Lancet 341:436 . Krynetski EY, Schuetz JD, Galpin AJ, Pui C-H, Relling MV, Evans WE. 1995. A

single point mutation leading to loss of catalytic activity in human thiopurine Smethyltransferase. Proc. Natl. Acad. Sci. USA 92:949-53

18. Tai HL, Krynetski EY, Yates CR, Loennechen T, Fessing MY, et al. 1996. Thiopurine S-methyltransferase deficiency: two nucleotide transitions define the most prevalent mutant allele associated with loss of catalytic activity in Caucasians. Am. J. Hum. Genet. 58:694—702

19. Loennechen Relling

MV,

T, Yates Krynetski

CR,

Fessing

MY,

EY, Evans

WE.

1998. Isolation of a human thiopurine S-methyltransferase (TPMT) complementary DNA with a single nucleotide transition A719G (TPMT *3C) and its association with loss of TPMT protein and catalytic activity in humans. Clin. Pharmacol. Ther. 64:46-51

20. Tai HL, Krynetski EY, Schuetz EG, Yanishevski Y, Evans WE. 1997. Enhanced pro-

teolysis of thiopurine S-methyltransferase (TPMT) encoded by mutant alleles in hu-

mans (TPMT *3A, TPMT*2): mechanisms for the genetic polymorphism of TPMT activity. Proc. Natl. Acad. Sci. USA 94:6444-49 Pasi Yates CR, Krynetski EY, Loennechen pebessingm Vi malate Eie ectralye lOO.

118

McLEOD

® EVANS

Molecular diagnosis of thiopurine Smethyltransferase deficiency: genetic basis for azathioprine and mercaptopurine intolerance. Ann. Intern. Med. 126:608—14 Me Ameyaw MM, Collie-Duguid ES, Powrie RH, Ofori-Adjei D, McLeod HL. 1999. Thiopurine methyltransferase alleles in British and Ghanaian populations. Hum. Mol. Genet. 8:367-70 py, Hon YY, Fessing MY, Pui C-H, Relling MV, Krynetski EY, Evans WE. 1999. Polymorphism of the thiopurine S-methyltransferase gene in African Americans. Hum. Mol. Genet. 8:371—76 24. Collie-Duguid ES, Pritchard SC, Powrie RH, Sludden J, Li T, McLeod HL. 1999. The frequency and distribution of thiopurine methyltransferase alleles in Caucasian and Asian populations. Pharmacogenetics 9:37-42 295 Coulthard SA, Howell C, Robson J, Hall AG. 1998. The relationship between thiopurine methyltransferase activity and genotype in blasts from patients with acute leukemia. Blood 92:2856—62 26. Lennard L, Van Loon JA, Weinshilboum RM. 1989. Pharmacogenetics of acute azathioprine toxicity: relationship to thiopurine methyltransferase genetic polymorphism. Clin. Pharmacol. Ther. 46:149-54 Pile Evans WE, Horner M, Chu YQ, Kalwinsky D, Roberts WM.

1991. Altered mer-

captopurine metabolism, toxic effects, and dosage requirement in a_ thiopurine methyltransferase-deficient child with acute lymphocytic leukemia. J. Pediatr. 119:985-89 28. Black AJ, McLeod HL, Capell HA, Powrie RH, Matowe LK, et al. 1998. Thiopurine methyltransferase genotype predicts therapy-limiting severe toxicity from azathioprine. Ann. Intern. Med. 129:71618 2S. Ishioka S. 1999. Thiopurine methyltransferase genotype and the toxicity of azathioprine in Japanese. Intern. Med. 38:944-47 . Relling MV, Hancock ML, Rivera GK,

.

32:

33:

.

3

36.

Cy

.

Sandlund JT, Ribeiro RC, et al. 1999. Mercaptopurine therapy intolerance and heterozygosity at the thiopurine S-methyltransferase gene locus. J. Natl. Cancer Inst. 91:2001-8 Relling MV, Yanishevski Y, Nemec J, Evans WE, Boyett JM, et al. 1998. Etoposide and antimetabolite pharmacology in patients who develop secondary acute myeloid leukemia. Leukemia 12:346-52 Bo J, Schroder H, Kristinsson J, Madsen B, Szumlanski C, et al. 1999. Possible carcinogenic effect of 6-mercaptopurine on bone marrow stem cells: relation to thiopurine metabolism. Cancer 86:1080-— 86 Relling MV, Rubnitz JE, Rivera GK, Boyett JM, Hancock ML, et al. 1999. High incidence of secondary brain tumours after radiotherapy and antimetabolites. Lancet 354:34-39 Liggett SB, Wagoner LE, Craaft LL, Hornung RW, Hoit BD, et al. 2000. beta(2)-Adrenergic receptor pharmacogenetics. Am. J. Resp. 161:S197-201 Lima JJ, Thomason DB, Mohamed MHN, Fberle LV, Shelf TH, Johnson JA. 1999. Impact of genetic polymorphisms of the beta(2)-adrenergic receptor on albuterol bronchodilator pharmacodynamics. Clin. Pharmacol. Ther. 65:519—25 Tan S, Hall IP, Dewar J, Dow E, Lipworth B. 1997. Association between beta 2-adrenoceptor polymorphism and susceptibility to bronchodilator desensitisation in moderately severe stable asthmatics. Lancet 350:995—99 Large V, Hellstrom L, Reynisdottir S, Lonnqvist F, Eriksson P, et al. 1997. Human beta-2-adrenoceptor gene polymorphisms are highly frequent in obesity and associate with altered adipocyte beta-2 adrenoceptor function. J. Clin. Invest. 100:3005-13 Meirheaghe A, Halbecque N, Cottel D, Amouyel P. 1999. Beta 2-adrenoceptor gene polymorphism, body weight, and physical activity. Lancet 353:869

PHARMACOGENOMICS 39) Liggett SB. 1998. Pharmacogenetics of relevant targets in asthma. Clin. Exp. Allergy 28(Suppl. 1):77-79 . Drysdale CM, McGraw DW, Stack CB, Stephens JC, Judson RS, et al. 2000. Com-

plex promoter and coding region beta(2)adrenergic receptor haplotypes alter receptor expression and predict in vivo responsiveness. Proc. Natl. Acad. Sci. USA 97:10483-88

40. Poirier J, Delisle MC, Quirion R, Aubert I, Farlow M. 1995. Apolipoprotein E4 allele as a predictor of cholinergic deficits and treatment outcome in Alzheimer disease. Proc. Natl. Acad. Sci. USA 92:12260-64

41. Kwok PY, Gu ZJ. 1999. Single nucleotide polymorphism libraries: Why and how are we building them? Mol. Med. Today 5:538-43 42. Brookes AJ. 1999. The essence of SNPs. Gene 234:177-86 43. Emahazion T, Jobs M, Howell WM, Siegfried M, Wyoni PI. 1999. Identification of 167 polymorphisms in 88 genes from candidate neurodegeneration path43a.

119

47. Leyland Jones B, Kelland LR, Harrap KR, Hiorns LR. 1999. Genomic imbalances associated with acquired resistance to platinum analogues. Am. J. Pathol. 155:77-84 48. Pinkel D, Segraves R, Sudar D, Clark S, Poole I, et al. 1998. High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat. Genet. 20:207-11 49. Pollack JR, Perou CM, Alizadeh AA, Eisen

MB,

Pergamenschikov

A, et al.

1999. Genome-wide analysis of DNA copy-number changes using cDNA microarrays. Nat. Genet. 23:41-46

50. Ross DT, Scherf U, Eisen MB, Perou CM, Rees C, et al. 2000. Systematic variation

in gene expression patterns in human cancer cell lines. Nat. Genet. 24:227-35

Site Alizadeh AA, Eisen MB, Davis E, Ma C, Lossos IS, et al. 2000. Distinct types

of diffuse large B-cells lymphoma identified by gene expression profiling. Nature 403:503-11

Sy Golub TR, Slonim DK, Tamayo P, Huard

ways. Gene 238(2):315—24

C, Gaasenbeek M, et al. 1999. Molecular

Arranz MJ, Munro J, Birkett J, Bolona A,

classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286:531—37

et al. 2000. Pharmacogetetic predictions of clozapine response. Lancet 355:1615— 16

44. Rooney PH, Murray GI, Stevenson DAJ, Haites NE, Cassidy J, McLeod HL. 1999. Comparative genomic hybridization and chromosomal instability in solid tumors. Br. J. Cancer 80:862-—73 45. Rooney PH, Stevenson DA, Marsh S, Johnston PG, Haites N, et al. 1998. Com-

parative genomic hybridization analysis of chromosomal alterations induced by the development of resistance to thymidylate synthase inhibitors. Cancer Res. 58:5042-45

46. Rustum YM, Harstrick A, Cao S, Vanhoefer U, Yin MB, et al. 1997. Thymidylate synthase inhibitors in cancer therapy: direct and indirector inhibitors. J. Clin. Oncol. 15:389-400

a3: Iyer VR, Eisen MB, Ross DT, Schuler G, Moore T, et al. 1999. The transcriptional program in the response of human fibroblasts to serum. Science 283:83-87 54. Scherf U, Ross DT, Waltham M, Smith LH, Lee JK, et al. 2000. A gene expression database for the molecular pharmacology of cancer. Nat. Genet. 24:236—44 Spy Milano G, McLeod HL. 2000. Can dihydropyrimidine dehydrogenase impact 5-fluorouracil-based treatment. Eur. J. Cancer 36:37—-42 56. Fijal BA, Hall JM, Witte JS. 2000. Clinical trials in the genomic era: effects of protective genotypes on sample size and duration of trial. Control. Clin. Trials 21:7-20 ile Rowinsky EK, Windle JJ, VonHoff DD.

120

McLEOD

#® EVANS

Haas

2000. Ras protein farnesyltransferase: a strategic target for anticancer therapeutic development. J. Clin. Oncol. 17:3631-—52

58. Drazen JM, Yandava CN, Dube L, Szczerback N, Hippensteel R, et al. 1999. Pharmacogenetic association between ALOX5 promoter genotype and the response to anti-asthma treatment. Nat. Genet. 22(2):168-70 oY). Wood AJ. 1998. Ethnic differences in drug disposition and response. Ther. Drug Monit. 20:525—26 60. McLeod HL, Syvanen AC, Githang’a J, Indalo A, Ismail D, et al. 1998. Ethnic dif-

67.

68.

ferences in catechol O-methyltransferase pharmacogenetics: Frequency of the codon 108/158 low activity allele is lower in Kenyan than Caucasian or SouthWest Asian individuals. Pharmacogenetics

8:195-99 61. Weinshilboum

RM. 1999. Methylation pharmacogenetics: catechol O-methyltransferase, thiopurine methyltransferase, and histamine N-methyltransferase. Annu. Rev. Pharmacol. Toxicol. 39:19-52

69.

62. Martinez FD, Graves PE, Baldini M., Solomon S, Erickson. 1997. Association between genetic polymorphisms of the beta2-adrenoceptor and\par response to albuterol in children with and without a history of wheezing. J. Clin. Invest. 100:3184— 88 63. Hancox RJ, Aldridge RE, Cowan JO, Flannery EM, Herbison GP, et al. 1999. Tolerance to beta-agonists during acute bronchoconstriction. Eur. Resp. J. 14:283-87 64. Lipworth BJ, Hall IP, Tan S, Aziz I, Coutie W. 1999. Effects of genetic polymorphism on ex vivo and in vivo function of beta2adraenoceptors in asthmatic patients. Chest 115(2):324—28 65. van der Kleij FG, Navis GJ, Gansevoort RT, Heeg JE. 1997. ACE polymorphism does not determine short-term renal response

to ACE-inhibition in proteinuric patients. Nephrol. Dial. Transplant. 12(Suppl 2): 42-46

70.

71.

U2

M,

Yilmaz

N, Schmidt

A, Neyer

U, Arneitz K, et al. 1998. Angiotensinconverting enzyme gene polymorphism determines the antiproteinuric and systemic hemodynamic effect of enalarpril in patients with proteinuric renal disease. Kidney Blood Press. Res. 21:66—69 Nakano Y, Oshima T, Watanabe M, Matsuura H, Kajiyama G, Kambe M. 1997. Angiotensin I-converting enzyme gene polymorphism and acute response to captopril in essential hypertension. Am. J. Hypertens. 10:1064-68 O’Toole L, Stewart M, Padfield P, Channer K. 1998. Effect of the insertion/deletion polymorphism of the angiotensin-converting enzyme gene on response to angiotensin-converting enzyme inhibitors in patients with heart failure. J. Cardiol. Pharmacol. 32:988—94 Sasaki M, Oki T, Iuchi A, Tabata T, Yamada H, etal. 1996. Relationship between the angio tensin converting enzyme gene polymorphism and the effects of enalapril on left ventricular hypertrophy and impaired diastolic filling in essential hypertension: M-mode and pulsed doppler echocardiographic studies. J. Hypertens. 14:1403-8 Mizuiri S, Hemmi H, Inoue A, Takano M, Kadomatsu S, et al. 1997. Renal hemodynamic changes induced by captopril and\par angiotensin-converting enzyme gene polymorphism. Nephron 75:31014 Yoshida H, Mitarai T, Kawamura T, Kitajima T, Miyazaki, et al. 1995. Role of the deletion of polymorphism of the angiotensin converting enzyme gene in the progression and therapeutic responsiveness of IgA nephropathy. J. Clin. Invest. 96:2162-69 Essen GGV, Rensma PL, Zeeuw DD, Sluiter WJ, Scheffer H, Apperloo AJ. 1996. Association between angiotensinconverting-enzyme gene polymorphism and failure of renoprotective therapy. Lancet 347:94-95

PHARMACOGENOMICS Wake Priort

SG, Barhanin J, Hauer RN, Haverkamp W, Jongsma HJ, et al. 1999.

Genetic and molecular basis of cardiac arrhythmias: impact on clinical management part III. Circulation 99:674-81 74. Donger C, Denjoy I, Berthet M, Neyroud N, Cruaud C, et al. 1997. KVLQT1

C-terminal

missense

mutation

causes

a

forme fruste long-QT syndrome. Circulation 96:2778-81

Ds

121

Abbott GW, Sesti F, Splawski I, Buck ME, Lehmann MH, et al. 1999. MiRP1 forms

IKr potassium channels with HERG and is associated with cardiac arrhythmia. Cell 97:175-87

76. Aithal GP, Day CP, Kesteven PJ, Daly AK. 1999. Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications. Lancet 353:717-19

Annu. Rev. Pharmacol. Toxicol. 2001. 41:123-43

PHENOBARBITAL RESPONSE ELEMENTS OF CYTOCHROME P450 GENES AND NUCLEAR RECEPTORS! T Sueyoshi and M Negishi Pharmacogenetics Section, Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, North Carolina 27709; e-mail: [email protected]. gov

Key Words induction

gene regulation, signal transduction, nuclear translocation, CAR,

@ Abstract Phenobarbital (PB) response elements are composed of various nuclear receptor (NR)-binding sites. A 51-bp distal element PB-responsive enhancer module (PBREM) conserved in the PB-inducible CYP2B genes contains two NR-binding direct repeat (DR)-4 motifs. Responding to PB exposure in liver, the NR constitutive active receptor (CAR) translocates to the nucleus, forms a dimer with the retinoid X receptor (RXR), and activates PBREM

via binding to DR-4 motifs. For

CYP3A genes, a com-

mon NR site [DR-3 or everted repeat (ER)-6] is present in proximal promoter regions. In addition, the distal element called the xenobiotic responsive module (XKREM) is

found in human CYP3A4 genes, which contain both DR-3 and ER-6 motifs. Pregnane X receptor (PXR) could bind to all of these sites and, upon PB induction, a PXR:RXR heterodimer could transactivate XREM. These response elements and NRs are functionally versatile, and capable of responding to distinct but overlapping groups of xenochemicals.

INTRODUCTION In the 1960s, there was only one cytochrome P450 recognized in liver microsomes, and there were two distinct chemicals that induced cytochrome P450: phenobarbital (PB) and 3-methylcholanthrene. Since that time, this induction phenomenon has long been a major driving force to attract many scientists into cytochrome P450 research. It was soon realized that these chemicals induce different forms of cytochrome P450; then molecular biology enabled the recent explosive increase of the number of different cytochrome P450s to > 500 (1). This was also the beginning of intensive efforts that have now resulted in the identification of xenochemical !The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

123

124

SUEYOSHI

® NEGISHI

response elements and transcription factors that regulate these elements and in unveiling induction mechanisms at the molecular level (2-5). This review focuses on recent progress in understanding PB-inducible transcription of cytochrome P450 genes. For some aspects that are not covered in this review, readers should refer to recent articles (6-8).

Cytochrome P450s play the central role in xenochemical metabolism, as members of the heme-thiolate monooxygenase gene superfamily. By increasing the capability for metabolic detoxification and elimination, induction of cytochrome P450 is an integral part of the defense mechanism against xenochemical insult. It was already noted in the early 1950s that 3-methylcholanthrene fed to rats reduced the hepatocarcinogenic activity of certain aminoazo dyes by increasing their N-demethylation (9, 10) and that chronic administration of barbiturates gradually shortened PB-elicited sleeping time by inducing the metabolism of PB. Although induction is advantageous in most instances, inevitably it is also associated with pharmacological and/or toxicological consequences such as alterations in drug efficacy, drug-drug interactions (11), and metabolic activation of procarcinogens (12). Thus, induction of cytochrome P450 can be viewed as “an environmental

friend and foe.” Studies of the induction mechanism may provide insights into understanding general principles of maintaining biological homeostasis against environmental insults and also a way of controlling cytochrome P450 to be more friendly and beneficial to human health. PB is the prototype of a large group of structurally unrelated chemicals that induce a large subset of cytochrome P450 genes: CYP2A, CYP2B, CYP2C, CYP2H, CYP3A, CYPO6A, and CYP102//06. In addition to cytochrome P450s, PB concertedly induces a large number of other enzymes such as NADPH-cytochrome P450 reductase and specific transferases, increasing metabolic capability as a whole (13-15). The induction is largely limited to the liver, although other organs such as the brain can be targets for this induction. Because liver-derived cell lines do not respond to PB in induction of cytochrome P450 genes, the development of suitable primary hepatocyte cultures has led to recent progress in discovering regulatory mechanisms of PB induction. CYP2B has been the main object of studies of PB induction because PB most effectively induces this gene in the liver. With this gene, we begin a journey to search the PB response element (Figure 1).

CYP2B GENES

Search for Phenobarbital Response Element The first cDNA and gene of CYP2B were cloned and sequenced in the 1980s (16, 17). Immediately thereafter, it was demonstrated that PB activates transcription of CYP2B genes (18,19). However, the real progress in identifying a PB response element did not occur until the mid-1990s. First, expression of a chloramphenicol acetyltransferase gene-reporter construct driven by various 5'-flanking sequences in transgenic mice suggested the presence of PB responsiveness in a far (more than —800 bp) upstream region of the CYP2B2 gene (20). The major

PHENOBARBITAL INDUCTION OF P450

HS2

125

HS1

-1.4/-1.2 kbp

-2.3/-2.2 kbp

-150/-50 bo | -96

PBRE-NR GRE AP-1

-69

Se Barbie box

C/EBP site

CCAAT

PBREM (51-bp) CYP2B1; CYP2B2: Cyp2b10:

2216 -

CYP2B6:

-

NR1

Figure 1 DNA elements found in the

NR2

CYP2B genes.

break-through soon came from Anderson’s laboratory, using rat primary hepatocytes, in which PB response activity was associated with a 163-bp DNA sequence at —2318 through —2155 bp of the CYP2B2 gene (21). Subsequently, the PB response activity of this sequence was independently confirmed by using an in situ injection of the reporter gene constructs into rat liver (22). All of these results have agreed that the PB response element must reside within this 163-bp sequence of the CYP2B2 gene designated a phenobarbital-responsive unit (23). With mouse Cyp2b10 and Cyp2b9 genes, PB response activity of the 163-bp DNA was delineated to a 51-bp minimum sequence that could respond to PB. First of all, the corresponding DNA sequence at —2426 through —2250 in the PB-inducible Cyp2b10 gene was identified and proved to be similarly activated by treatment with PB in mouse primary hepatocytes (24). DNase I footprinting on the DNA with mouse liver nuclear extracts defined six regions protected from digestion. Using this information, an additional deletion assay was conducted to delineate the PB response activity of the 163-bp DNA to the 69-bp sequence at —2365 through —2297. When base mutations within the corresponding 69-bp DNA of the noninducible Cyp2b9 gene were introduced into the 69-bp response sequence of the Cyp2b/0 gene, this mutated response sequence completely lost its PB responsiveness. These findings provided genetic evidence that the PB response element is centered on the 69-bp DNA. Sequence analysis of the 69-bp DNA revealed the presence of two possible nuclear receptor (NR) motifs and a nuclear factor 1 (NF1) binding site. Keeping the presence of these motifs and the binding site in mind, further deletions and mutations were used to associate the

PB response activity to a minimum sequence of 51-bp of DNA at —2339 through -—2289 of the Cyp2b10 gene, now called the phenobarbital-responsive enhancer module (PBREM) (25). The PBREM sequence was also found in the rat CYP2B1, CYP2B2, and human CYP2B6 genes (26). Evolutionary conservation of PBREM in the CYP2B genes from mouse to human strongly supports the hypothesis that PBREM is a general PB response element.

126

SUEYOSHI

® NEGISHI

Phenobarbital-Responsive Enhancer Module The PBREM is characterized as a composite element consisting of two nuclear receptor (NR)-binding sites (NR1 and NR2) and an NF1-binding site (25). Both NR1 and NR2 are a DR-4 motif; a DR of imperfect half sites separated by 4 bases. Only the NR sites are essential for the PB response activity, although the NF1 site may be required to confer full PBREM activity. Mutation of either NR1 or NR2 decreased PBREM activity to one third of the wild-type activity in transfected primary hepatocytes, while simultaneous mutations of both NR sites abolished PBREM activity. PBREM sequences containing NF1 mutations retained significant residual activity in the PBREM. The nonessential role of an NFI site was also clearly demonstrated in transgenic mice bearing a S’-flanking region of the CYP2B2 gene with a mutated NF1 site (27). This mutated DNA responded to PB

as efficiently as did the wild-type DNA in the mice. In liver in vivo, however, a large alteration of chromatin structure occurred in the PBREM region after PB treatment, and the NF1 site appeared to be a core region of this alteration (28). The role of the NF1 site in modulating the PBREM-phenobarbital-responsive-unit activity should not be ruled out at the chromatin level. PBREM is capable of responding to various PB-type inducers in addition to PB (25, 26,29). These inducers include clotrimazole,

chlorpromazine(CPZ),

metyrapone, acetone, methyl] isobutyl ketone, isoamyl alcohol, pyridine, 2,2’,4,4’tetrachlorobiphenyl, 2,2’,5,5’-tetrachlorobiphenyl, 2,3,3’,4’,5,6-hexachlorobiphenyl, 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP), dieldrin, 1,1,1-trichloro-1,2-bis(o,p’-chlorophenyl)ethane (o,p'-DDT), methoxychlor, and camphor. On the other hand, neither 3-methylcholanthrene (CYP1A1 inducer), dexamethasone (CYP3A inducer), clofibrate (CYP4A inducer), nor 1,4-bis[2-(3-chloro-

pyridyloxy)]benzene (an inactive derivative of TCPOBOP) activated PBREM. Thus, PBREM has emerged as a versatile response element that can specifically respond to various PB-type inducers, leading to the induction of CYP2B genes in the mouse, rat, and human. To begin to answer to the question of how such diverse

chemicals activate the same response element, a transcription factor that regulates PBREM must first be identified.

Nuclear Constitutive Active Receptor Sequence comparison of PBREMs revealed that the NR1 site (RGGTCAggaaAGTACA) is the most conserved element; mouse and rat NR1 sites are identical, and they differ only by | base from their human counterpart, suggesting that the NR1 site may play the key role in regulating PBREM in response to PB. Since the NR1 site is most conserved within the PBREM, efforts were concentrated on finding a nuclear protein that binds to this site (30). Two different approaches were concur-

rently taken to identify a binding protein to NR1. The first approach was to search for known NRs by using a transient transfection assay. The second was to purify a nuclear protein by using NR1-affinity chromatography. To accomplish this, we first cotransfected PBREM with various expression vectors of NRs, such as retinoid

PHENOBARBITAL INDUCTION OF P450

127

X receptor (RXR), constitutive active receptor (CAR), liver X receptor, thyroid

receptor-a, hepatocyte nuclear factor 4 (HNF4), and chicken ovalbumin upstream promoter-transcription factor (COUP-TF) in HepG2 and HEK293 cells. Among these liver-enriched NRs tested, only CAR was abie to activate PBREM in the co-

transfected cells. The nuclear CAR was originally characterized as a constitutive activator of an empirical set of retinoic-acid response elements, meaning that CAR activated the response element in the absence of retinoic acid (31). Consistent with the fact that CAR activated the response element by forming a heterodimer with RXR, coexpression of RXR resulted in a synergistic increase of PBREM activity in the CAR-transfected cells. As expected from these findings, a gel shift assay confirmed that an in vitro translated CAR alone did not bind to NR1,

but its mixture

with the similarly prepared RXR bound to NRI specifically. Thus, it appeared that a CAR:RXR heterodimer activated the PBREM. To purify a nuclear protein that binds to NR1, two different DNA resins were used for affinity chromatography: NR1-(AGGTCAGGAAAGTACA) and NR1I’conjugated resins (30). NR1’ [AGTTCAGAAAAGTACT (underlined bases differ from NR1] is the mutated NR1 found in the noninducible Cyp2b9 gene. Western blot analysis of the affinity-purified fractions of mouse liver nuclear extracts showed that the NRs CAR and RXR were enriched on NR1 but not NR1’ resins and only from the PB-induced nuclear extracts. Moreover, protein microsequencing of the stained bands confirmed that CAR and RXR were included in the purified fraction from NR1 resin of PB-induced nuclear extracts. By forming a heterodimer with RXR, CAR appeared to be a nuclear factor capable of binding to NR1, and the binding occurred after treatment with PB. Transfection assays using the NR1-tk-reporter plasmid revealed that NR1 alone was activated by CAR, indicating that the binding of CAR:RXR to NRI1 could be sufficient to activate PBREM.

Because a mixture of the in vitro-translated CAR and RXR also exhibited a weaker binding to NR2, CAR:RXR may bind concurrently to both NR1 and NR2 during the activation of PBREM in the transfected cells. The cotransfection of mCAR resulted in expression of the endogenous CYP2B gene and the activation of PBREM in HepG2 cells (26). Moreover, a CAR-null mouse has been produced by using gene targeting, in which the induction of the Cyp2b/0 gene by PB or TCPOBOP was impaired (32). These results have provided the ultimate evidence that CAR is the receptor that can regulate PB induction of the CYP2B gene. A major enigma

with this activation mechanism, however, was that CAR activated absence of PB or PB-type inducers in HepG2 cells. If CAR is to tion factor that activates PBREM in response to PB, the receptor be repressed in unexposed liver in vivo. We discuss this in a later article.

PBREM in the be a transcripfunction must section of this

DNA Elements Other Than the Phenobarbital-Responsive Enhancer Module Before the in vitro and in vivo systems (e.g. primary hepatocyte, transgenic mouse, and direct injection of DNA) were readily available for assaying PB-responsive

128

SUEYOSHI

® NEGISHI

activity of DNA functionally, a search for a PB response element relied on nuclear protein-binding assays and was largely limited to the proximal promoter regions of CYP2B genes. Although various sequences have been defined as nuclear proteinbinding sites in the promoter regions, a peer consensus has not yet emerged with respect to functional roles of these sequences and elements in PB induction. We have only briefly described DNA elements in the proximal promoter regions of CYP2B genes, because these have already been summarized in recent reviews (13, 33). Two DNase I-hypersensitive sites appeared on the CYP2B2 gene after PB induction; one was in a proximal promoter (HS1), and the other one was at a distal —2.2-kbp region (HS2) (34). Although the HS2 region includes PBREM, the HS1 region contains the previously reported sequences and elements, such as a Barbie box (see below), CCAAT box, CCAAT-enhancer binding protein (C/EBP)-binding

site, and other nuclear protein-binding sites. Some findings with these proximal sequences are quite consistent. Deletion or mutation of a putative C/EBP-binding site (at around —65 through —45 bp) dramatically decreased promoter activity of CYP2B genes in transient transfection assays (25). Although this site was neither responsive to PB nor did its mutation affect PB-enhanced reporter gene activity, it may regulate a basal transcription activity of the genes. No specific protein binding was observed to the Barbie box and deletion or mutation of the Barbie box did not affect PB responsiveness of CYP2B genes. A 27-bp DNA sequence [called a positive element (PE) and located at —96 through —69 bp] overlapping with the Barbie box and C/EBP site was repeatedly shown to have PB-responsive activity by using in vitro transcription and in vivo gene delivery systems (35-37). Other possible regulatory elements were also located in the —1.4- through —1.2-kbp region of CYP2B genes: a functional glucocorticoid response element (GRE) at position —1357 bp (38) and an Activator protein-1 site at —1441 bp (39). A DNA sequence (—1404 through —971) of the Cyp2b/0 gene was associated with a minor PB response in mouse primary hepatocytes, in which there is an NR-binding motif (PBRE-NR) similar to NR1 within PBREM (40). Yet another 1.3-kbp 5’-flanking sequence of the CYP2B/gene also exhibited a PB response

in transgenic mice (41). Thus, the elements within the —1.4- through —1.2-kbp region may account for some degree of total PB responsiveness of CYP2B genes.

CYP3A GENES

Dexamethasone/Rifampicin Response Element/Xenobiotic-Responsive Module The CYP3A genes were originally characterized by their nonclassical induction by glucocorticoids (42, 43), and they are also known by their induction by macrolide antibiotics

such as rifampicin

(44). In addition,

PB

often induces

these

CYP3A genes. Response activity of various proximal promoter sequences to the typical CYP3A inducers (e.g. dexamethasone, pregnenolonel6-carbonitrile and

PHENOBARBITAL INDUCTION OF P450

-7.8/-7.6 kbp

129

-170/-140 bp 7273

DR-3 " AGCTCAagaAGGTCA™®

ER-6

—:CYP3A2

18° TGAACT caaaggAGGTCA"** :CYP3A4

ER-6 “7! JTAACTcaat ggAGGTCA™ :CYP3A7

CYP3A4:

“TGAACTtgcJGACCC

Figure 2

T*TGAAATcatgtcGGTTCA

DR-4

AACTCAaaggAGGTCA™? enue

—:CYP3Aza

DNA elements found in the CYP3A genes.

rifampicin) was examined in primary hepatocytes or liver-derived H4IIE cells. As depicted in Figure 2, common enhancer elements emerged: 68 A-C in CYP3A2 (45), consensus IT in CYP3A4 (46), and DexRE-1 in CYP3A23

(47,48). These

common elements are found in approximately —150-bp regions of CYP3A promoters, and they contain NR sites as DR-3, DR-4, or ER-6 motifs (49-51). Having

the relatively low activation of these proximal elements by rifampicin, Goodwin et al took a courageous step to search for an additional response element up to the —13-kb region of the CYP3A4 gene, and they identified a 230-bp distal element [called the xenobiotic-responsive enhancer module (XREM)] at —7836 through

—7607 of the CYP3A4 gene (52). A 417-bp proximal promoter sequence (—365 through +52) of the CYP3A4 gene did not respond to either dexamethasone or rifampicin in HepG2 cells. Placing XREM in front of the 417-bp DNA conferred a strong response to both inducers. Moreover, PB also activated XREM in HepG2 cells. XREM contains two NR-binding sites—dNR1 (DR-3) and dNR2 (ER-6)— that are separated by 29 bases. Neither of these NR sites alone seems to regulate XREM; a mutation of dNR1 decreased response activity only 30-40%, whereas that of dNR2 increased it ~20--30%. Even when both dNR1 and dNR2 were simultaneously mutated, XREM retained >50% of the original activity. Mutation of all dNR1, dNR2, and proximal consensus II/prPXRE (ER-6) sites reduced the activity to ) (16). Thus, an overproduction of either NO or O,° will decrease the reactivity of ONOO ; accordingly, the maximum activity of ONOO™ occurs when O,°~ and NO are produced in equivalent amounts. When there is more production of NO than O,°~, NO, is produced, which can then react with NO to produce the nitrosating agent N,O3. When there is an overproduction of O,°, oxidative chemistry prevails. There are two likely sources of O,°°: mitochondria and immune cells. Mitochondria produce O,*° during the course of aerobic respiration. As NO is more soluble in lipid layers, ONOO™ may be formed in the hydrophobic regions of mitochondria.

However, the mitochondrial manganese

SOD

(MnSOD)

may

play a role in limiting the production of ONOO™ under normal conditions. Under inflammatory conditions immune cells such as neutrophils and macrophages produce large quantities of O,° through either NADPH oxidase or xanthine oxidase (22). Macrophages also produce large quantities of NO; thus, ONOO7 is likely to be produced under inflammatory conditions. Immune cells may also produce nitrating species through the myeloperoxidase system. In summary, the potential reactions of NO are numerous and dependent on many different factors. The site and source of production, as well as the concentration of NO, collectively determine whether NO will elicit direct or indirect effects. In addition, a relative balance between oxidative and nitrosative stress exists that will determine the indirect effects of NO.

NO MODIFICATIONS OF PROTEINS

S-Nitrosylation S-nitrosylation of cysteine residues resulting from the addition of aNO+ group has been shown to modify the activity of several proteins. Although it is unlikely that NO acts directly on the cysteine residue, NO interacts with O, or O,*~ to produce

NON-cGMP MEDIATED EFFECTS OF NO

207

RNS capable of nitrosylating cysteine residues. Nitrosylation is a chemical reaction, not an enzymatically catalyzed reaction. However, there seems to be some specificity in nitrosylation. First , not every protein with available cysteine residues becomes nitrosylated. Subcellular location and the local chemical environment (i.e. the local concentration of NO and molecules that react with NO such as O,°~ and heme proteins) may dictate to some extent which proteins become nitrosylated. Furthermore, not every available cysteine residue within a given target protein becomes nitrosylated. Of the five cysteine residues in p218*, only one cysteine is nitrosylated (23). In the ryanodine receptor there are 84 cysteines with free —SH groups, but only 12 cysteines appear to be susceptible to nitrosylation (24). The tertiary structure of a protein may make some cysteine residues more susceptible to nitrosylation. A consensus sequence, though somewhat degenerate, for nitrosylation has also been postulated. The proposed motif is XYCZ, where X is Gly, Ser, Thr, Cys, Tyr, Asn, or Gln; Y is Lys, Arg, His, Asp or Glu; and Zis Asp or Glu (25).

Nitrosylation has been shown to modify the function of several proteins including the N-methyl-D-aspartate (NMDA) receptor, p218*5, caspase-3, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH), which suggests that nitrosylation may be an important cellular regulatory mechanism. The NMDA receptor is a glutamate receptor critical for development, learning, and memory in the central nervous system (26). Activation of the NMDA receptor triggers Ca?* influx, which causes nNOS activation and subsequent NO production (27, 28). Manzoni & Bockaert showed that NMDA receptor activity is downregulated by endogenous NO in primary neurons (29). This downregulation was later shown, through site directed mutagenesis experiments, to be due to specific nitrosylation of cysteine 399 in the NR2A subunit of the NMDA receptor (30). Downregulation of NUDA receptor function by NO may act as a negative feedback mechanism to prevent excessive activation of the NMDA receptor and associated neurotoxicity (31). Although NO has been shown to be neuroprotective, it has also been shown to be neurotoxic in some circumstances. The paradoxical effects of NO in neurons are an example in which the effects of NO are dictated by the intracellular milieu. When NO is produced under conditions in which O,°~ is available, NO has neurotoxic effects in rat cerobrocortical cultures (a mixture of neuronal and

glial cells) (32). When cells are treated with SOD, NO does not produce these toxic effects; however, exogenous treatment of the cerobrocortical cultures with peroxynitrite and SOD causes neurotoxicity (32). Furthermore, treatment of neurons with NO congeners in different redox states resulted in different effects. 3-morpholinosydnonimime (SIN-1), which produces both NO. and O,*”, caused neurotoxicity; nitroglycerin and sodium nitroprusside, which both produce NO+ equivalents, cause neuroprotective effects (32). The neurotoxic effects of NO seem to be related to the formation of peroxynitrite; however, it is unclear whether the deleterious effects of peroxynitrite are on the NMDA receptor or elsewhere in the cell. It is clear that when cellular conditions favor nitrosylation,

NO has neuropro-

tective effects; however, if NO is produced under conditions in which there is also O,°~ production, NO can be neurotoxic.

208

DAVIS ET AL

In addition to the NMDA receptor, nitrosylation has also been shown to modulate the activity of several other channels. Nitrosylation of a single cysteine residue on the intracellular face of the cyclic nucleotide-gated channel involved in olfactory and visual transduction results in activation of this channel (33,34). Polynitrosylation of the cardiac calcium release channel (ryanodine receptor) also results in activation of this channel (24). In this case, nitrosylation of up to 12 cysteine residues of the available 84 leads to progressive, reversible activation of the channel (24). Recently, NO has been shown to modulate cell survival by interacting with several proteins in proapoptotic pathways. One such protein whose function is modified by nitrosylation is p21"*. Lander and coworkers showed that NO activates p21" in human T cells (23). Furthermore, in vitro experiments showed that this activation is a direct, reversible effect mediated by NO (23). The concentration

of NO necessary to elicit maximal activation in vitro was more than 1000 times higher compared to that required in whole cells, which suggests that intracellular conditions are more favorable for nitrosothiol formation. Lander subsequently showed that p21" is specifically nitrosylated on cysteine 118 (35). When cysteine 118 is mutated to a serine (C118S mutant), NO mediated activation of p2Iras is eliminated (35).

In PC12 pheochromocytoma cells, p21"* mediates neuronal growth factordriven differentiation and survival through the recruitment of the mSOS-RasMAP kinase cascade (mSOS-Ras-Raf-MEK-ERK) and phosphatidylinositol (PI) 3—kinase,

respectively

(36-39).

The mSOS-Ras-MAP

kinase pathway

is nor-

mal in the C118S- expressing PC12 cells compared to wild type Ras-expressing PC12 cells. However, after long-term neuronal growth factor treatment, C118Sexpressing PC12 cells are unable to maintain PI 3-kinase activation resulting in apoptosis of these cells (40). Thus, while nitrosylation of Ras does not affect the mSOS-Ras-MAP kinase pathway, it does modulate the interaction of Ras with the PI 3-kinase pathway. These experiments demonstrate that NO, acting as an anti-apoptotic agent, activates p21" to modulate neuronal PC12 survival but not differentiation (40). NO has also been shown to nitrosylate several members of the caspase family of proteins, which functions in the apoptotic pathway. However, in this case nitrosylation results in inhibition of activity instead of activation, as observed with

p21". NO has been shown (in vitro) to inhibit seven members of the caspase family, including caspases-1, -2, -3, -4, -6, -7, and -8, through mitrosylation (41). Caspases contain a reactive cysteine residue in the active site of the enzyme, which is specifically nitrosylated by NO donors (42). Nitrosylation of caspase appears to be reversible, probably depending on the redox state of the cell, which suggests that NO-mediated inhibition of apoptosis is reversible (43). Caspase-1 is part of the caspase subfamily that participates in cytokine mat-

uration. In human umbilical vein endothelial cells (HUVEC), NO was shown to

inhibit tumor necrosis factor (TNF)-a-induced apoptosis through s-nitrosylation of caspase-1 (44). Both exogenous treatment of cells with NO donors and endogenou s

.

NON-cGMP MEDIATED EFFECTS OF NO

209

activation of NOS by shear stress antagonized TNF-a-induced apoptosis in a cGMP-independent manner. Interestingly, whereas low concentrations of NO donors (300 uM) were pro-apoptotic, revealing once again the paradoxic effects of NO (44). Treatment of purified caspase-1 with NO resulted in nitrosylation of the active site cysteine residue and inhibition of caspase activity (44). Site-directed mutagenesis experiments showed that caspase-3 is also nitrosylated on its active site cysteine (45). Furthermore, endogenous caspase-3 nitrosylation has been documented. Experiments in human B and T cell lines have demonstrated a biological role for caspase nitrosylation in apoptosis. A significant portion of caspase-3 was found to be nitrosylated in human B and T cell lines (including 10C9 and Jurkat cells) inhibiting caspase activity (45). Activation of Fas-dependent apoptosis in these cells causes a denitrosylation of caspase-3 within 1.5—2.0 hours (45). In-

cubation of Jurkat and 10C9 cells with NOS inhibitors for long periods of time (24 hours) resulted in denitrosylation of caspase-3 but not activation. However, the

denitrosylation caused an increase in Fas-induced activation of caspase-3. Thus, both denitrosylation and cleavage seem to be required for caspase activation. When the Fas apoptotic pathway is induced, caspase-3 is denitrosylated to expose the active site cysteine so that, upon Fas-induced cleavage, caspase-3 becomes active. In this way, NO regulates the Fas apoptotic pathway through the balance of nitrosylation/denitrosylation (45, 46). GAPDH was among the first proteins shown to be nitrosylated. A correlation was found between NO production and inhibition of GAPDH in rat liver during chronic inflammation (47). GAPDH

contains a cysteine residue in its active site

that is thought to be nitrosylated, leading to reversible inhibition of GAPDH activity (47). Moreover, the S-nitrosylation of GAPDH

facilitates further covalent

modification of the enzyme by NADH (48). Nitrosylation of GAPDH is reversible and may be involved in the regulation of glycolysis; however, NADH modification of GAPDH is irreversible and is likely to be involved in pathological events (48). Several other enzymes, including aldolase, aldehyde dehydrogenase, cathepsin B, and y-glutamylcysteinyl synthetase, are nitrosylated on their active site cysteine residues, in a manner similar to GAPDH, and are reversibly inactivated(49).

Tyrosine Nitration Nitrotyrosine formation has been demonstrated by immunohistological staining in numerous human diseases and animal models. Nitration of tyrosine residues is thought to be selective. Western blotting of nitrotyrosine-containing tissues in our laboratory revealed that only a few proteins appear to be nitrated in vivo. In skeletal muscle, the SERCA2a isoform of the sarcoplasmic reticulum Ca-ATPase

is nitrated. In in vitro experiments SERCA2A is preferentially nitrated, even when the SERCA1 isoform is added in excess of SERCA2A (50). It is worth noting that

210

DAVIS ET AL

although tyrosine nitration has been demonstrated in many pathological conditions, nitrotyrosine formation also seems to occur under basal conditions. In addition to selectivity, there is increasing evidence that tyrosine nitration is reversible. In a paper recently published in Proceedings of the National Academy of Sciences, our laboratory described an activity in crude rat spleen and lung homogenates that removes the nitrotyrosine epitope of nitrated proteins (51). It is unclear what type of modification this factor is making to the nitrotyrosine epitope (the product of the reaction is currently being determined). We have found this activity not only in spleen but also in several other tissues. Other investigators have also shown evidence that tyrosine nitration is reversible. When Gow et al incubated nitrotyrosine-containing peptides with plasma, the amount of nitrated peptides was reduced, which suggests that tyrosine nitration may be reversible (52). The in vivo pathway of tyrosine nitration has been a source of controversy for several years. Figure 2 shows the potential pathways for tyrosine nitration. Peroxynitrite, formed from the reaction of NO with O,°-, was thought to be the major nitrating agent in vivo. As a matter of fact, nitrotyrosine is routinely used as a marker for peroxynitrite production. However, there may be peroxynitrite-independent mechanisms of tyrosine nitration; thus, nitrotyrosine may not be a reliable marker for peroxynitrite production. There are a number of in vitro chemical studies showing that peroxynitrite can nitrate tyrosine residues (53,54). Furthermore, peroxynitrite has been shown to nitrate tyrosine residues in intact cells (55). In contrast to these studies, Pfeiffer and Mayer observed very little tyrosine nitration at physiologic pH when NO® (spermine NONOate) and O,°~ (xanthine oxidase) were generated simultaneously to form peroxynitrite compared to treatment with preformed peroxynitrite causing these authors to suggest that peroxynitrite may not be a nitrating agent in vivo (56). However, several investigators have questioned Pfeiffer and Mayer’s studies in two recently published papers where efficient tyrosine nitration by the simultaneous production of NO* and O,* was demonstrated (56a, 56b).

Sawa and Reiter show that the accumulation of urate and the rapid

consumption of oxygen by xanthine oxidase in the Pfeiffer and Mayer study may have led to erroneous conclusions (56a, 56b).

An alternative pathway for peroxynitrite-mediated tyrosine nitration is the in-

teraction of peroxynitrite with CO, (rate constant, 5.8 x 10* M7! s~!) to form nitrating species. Squadrito & Pryor proposed that the reaction of peroxynitrite with CO, yields the free radicals NO*, and CO;*~, which can nitrate phenolic compounds such as tyrosine (57). Recent evidence suggests there may be peroxynitrite-independent mechanisms by which nitrotyrosine can form in vivo. Myeloperoxidase (MPO) uses H,O, and Cl” to produce HOCI. Eiserich proposed that NO,~ formed from the oxidation of NO can be oxidized by either HOCI or myeloperoxidase to form reactive nitrogen species, NO,Cl and NO3, which may be capable of nitrating tyrosine residues (58,59). Eiserich et al used human polymorphonuclear neutrophils to show that the MPO system can generate RNS capable of nitrating the tyrosine residues

NON-cGMP MEDIATED EFFECTS OF NO

Oo°+ NO —®

OONO

Tyr:

211

COs

NO»o:+ CO3"

tyrosine nitration =~

-||

Cl

fee HOCI

Figure2 The proposed mechanisms for in vivo nitration of tyrosine residues. The MPO-mediated mechanisms are demonstrated on the bottom, and peroxynitrite and tyrosy] radical mechanisms are shown at the top. MPO, myeloperoxidase; tyre, tyrosyl radical.

212

DAVIS ET AL

of several synthetic peptides (58). There is also recent evidence that eosinophil peroxidase may catalyze the production of nitrating species (60). In contrast to Eiserich’s proposed mechanism, van Dalen et al showed that NO,~ is a poor substrate for MPO (61). Thus, in the presence of physiological levels of NO,~ and Cl-, MPO catalyzes very little tyrosine nitration. However, van Dalen proposed an alternative mechanism for MPO-mediated tyrosine nitration (61). MPO may catalyze the oxidation of free tyrosyl residues to tyrosyl radicals that can then exchange for tyrosyl residues in proteins. These tyrosyl radicals can then be nitrated by NO,. Thus, free tyrosine acts as a cosubstrate in MPO-mediated tyrosine nitration. Another pathway for the formation of nitrotyrosine is through the direct reaction of NO with a tyrosyl radical. NO can form an unstable complex with the tyrosyl residue of prostaglandin H synthase-2 (62). This complex can be oxidized to forma nitrotyrosine (62). This pathway necessitates the presence of a tyrosyl radical in the protein and thus, may be of limited consequence in biological systems. However, there are a number of proteins that contain tyrosy] radicals and may undergo similar chemistry, namely ribonucleotide reductase and photosystem II (63, 64). Nitrotyrosine formation has gained increasing attention. The observation of tyrosine nitration in a number of human diseases has brought the idea that nitration of tyrosine residues in proteins has functional consequences to the forefront of NO research in recent years. Nitrotyrosine has been detected in atherosclerotic plaques of coronaries by both immunohistochemistry and Western blotting (65,66). Increased nitrotyrosine staining was found in motor neurons of patients with ALS (9). Tyrosine nitration has also been found in rejected renal allografts and chronic renal failure, inflammatory bowel disease, the synovial fluid of arthritis patients, and the placental tissues from preeclamptic pregnancies (67-71). Nitrotyrosine formation is also found in numerous animal models of disease. In the heart, nitrotyrosine formation has been detected in ischemia-reperfusion injury and myocardial inflammation (72, 73). Tyrosine nitration has also been found in the kidney in endotoxin-induced injury and in renal hypertension (74,75). Nitration is evident in an MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine)-induced model of Parkinson’s disease and in a transgenic model of ALS (76, 77). The list of

pathological conditions in which nitrotyrosine formation is observed is extensive; however, in very few of these diseases have the modified protein(s) been identified.

Furthermore, many proteins that have been shown to be nitrated by peroxynitrite in vitro have never been conclusively shown to be nitrated in vivo. Thus, the physiological consequences of tyrosine nitration remain poorly understood. One protein that has been shown to be nitrated in vivo is prostacyclin synthase. Nitration of prostacyclin synthase has been demonstrated in both interleukin 18 (IL-18)-stimulated rat mesangial cells and in atherosclerotic bovine coronary arteries (78,79). In rat mesangial cells prostacyclin synthase was immunoprecipitated from IL-1 6-treated cells using antinitrotyrosine antibodies. Decreased prostacyclin synthase activity in these cells as well as decreased activity of prostacyclin synthase, treated in vitro with peroxynitrite, indicates that nitration

NON-cGMP MEDIATED EFFECTS OF NO

213

causes inhibition of prostacyclin synthase (78, 80). In bovine coronary arteries immunoprecipitations using prostacyclin synthase antibodies revealed increased staining for nitrotyrosine in atherosclerotic arteries compared with normal tissue. Vasodilation was impaired in the atherosclerotic vessels compared with normal vessels, which suggests that inhibition of prostacyclin synthase by nitration was at least partially responsible for impaired relaxation in these vessels. Manganese superoxide dismutase (Mn SOD) was also found to be nitrated in vivo. Mn SOD was extracted from human renal allografts by immunoprecipitation using antinitrotyrosine antibodies (67). More Mn SOD was immunoprecipitated from chronically rejected allografts compared with control allografts; however, it was unclear if the two groups of allografts contained the same amount of total Mn SOD (67). In addition to increased nitration, the chronically rejected al-

lografts showed decreased Mn SOD activity, which suggests that nitration inhibits Mn SOD activity (67). Consistent with this hypothesis, Mn SOD is inactivated by peroxynitrite treatment in vitro. Furthermore, ONOO™ specifically nitrates only one tyrosine residue, Tyr34, located near the bound manganese (81). Nitration of Mn SOD, responsible for scavenging 02° in the mitochondria, may cause mitochondrial dysfunction under inflammatory conditions such as those associated with chronic organ rejection. Another protein known to be nitrated in vivo is the low-molecular-weight neurofilament subunit protein (82). This protein isolated from the cervical spinal cords of ALS patients was found to be nitrated. Nitration was associated with neurofilamentassembly derangement in ALS patients investigated by Chou et al (83). However, Strong et al found no differences in the quantity or quality of neurofilament nitration in familial ALS patients compared with control patients (82). Thus, the functional consequence of neurofilament nitration remains unclear. As mentioned above, the SERCA2a isoform of the skeletal muscle sarcoplasmic reticulum ATPase, found predominantly in slow-twitch muscle, is nitrated in vivo (84). SERCA2a, which has a relatively long protein half-life, has been shown to accumulate nitrotyrosine with biological aging in a rat model (84). This accumulation of nitrotyrosine is associated with decreased Ca-ATPase activity. In vitro treatment of SERCA2a with peroxynitrite results in both increased tyrosine nitration and decreased Ca-ATPase activity. Both the in vitro and in vivo data suggest that nitration of SERCA2a inhibits Ca-ATPase activity. Tyrosine nitration is localized to Tyr 294 and Tyr 295 by tryptic digestion and V8 protease treatment of the protein followed by analysis for nitrotyrosine content. In contrast to the SERCA2a

isoform of the SR CaATPase, the SERCA1

isoform, found predomi-

nantly in fast-twitch muscle, is not nitrated in vivo. The preferential nitration of the SERCA2a isoform over SERCA1a suggests that tyrosine nitration is at least somewhat selective (50). Exposure of pulmonary surfactant protein A (SP-A) to nitrating agents such as

peroxynitrite or tetranitromethane causes tyrosine nitration of this protein (85). Nitration of SP-A results in decreased mannose-binding ability and decreased ability to aggregate lipids (86). Sequencing of nitrated SP-A tryptic peptides showed that

214

DAVIS ET AL

the protein is specifically nitrated on two tyrosine residues, Tyr 164 and Tyr 166, in the carbohydrate recognition domain (87). Although to our knowledge specific in vivo nitration of SP-A has not been shown, increased iNOS activity and increased

tyrosine nitration in the airway epithelium of asthmatic patients and patients with adult respiratory distress syndrome have been demonstrated histologically (85, 88). In addition to modification of the proteins mentioned above, tyrosine nitration may also function in cellular signaling. We and others have hypothesized that nitration of tyrosine residues in tyrosine kinase substrates may prevent phosphorylation and therefore inhibit tyrosine kinase signaling (52, 89). Using a synthetic peptide, Gow and his colleagues showed that peroxynitrite-mediated nitration of the tyrosine residue in this peptide resulted in about 50% inhibition of tyrosine phosphorylation by the tyrosine kinase, c-sre (52). In addition, Kong and his coworkers showed that peroxynitrite treatment of the pentadecameric peptide cdce(6-20)NH2, which corresponds to the tyrosine phosphorylation site of p34cdce2 kinase, increased nitration and inhibited phosphorylation of the peptide compared with untreated peptide (89). Crosstalk between NO and other signaling cascades has been the subject of numerous publications in recent years. The inhibition of tyrosine kinase signaling by tyrosine nitration represents a novel mechanism of NO interaction with tyrosine kinase signaling. However, to our knowledge, no tyrosine kinase substrate has been shown to be nitrated in vivo. Furthermore, in some systems, treatment with peroxynitrite has been shown to increase tyrosine phosphorylation (90, 91). Thus, the inhibition of tyrosine phosphorylation by tyrosine nitration remains highly speculative.

INTERACTION OF NO WITH TRANSITION METALS The physico-chemical properties of NO govern its interaction with transition metals. NO can form a o-bond through its nitrogen pair of electrons and a 2-bond through the antibonding 2pz* unpaired electron with the d-electrons of transition metals (92), acting as a three electron donor.

Effect of NO on Heme Proteins Iron is by far the most abundant transition metal in biological systems. Not surprisingly, investigators have given much attention to the interaction of NO with iron and to the biological function of this interaction. The geometry of heme determines the character of the NO-heme interaction. As a rule, NO does not interact efficiently with six-coordinated heme (93) and has a limited effect on the function of these proteins. The character and the outcome of the NO interaction with iron depends on various factors, including the oxidation state of the iron (ferrous or ferric), the microenvironment in the heme-binding pocket (residues surrounding the prosthetic group), and the availability of oxygen or other radical species. Soluble guanylyl cyclase (SGC) is a heterodimeric hemoprotein that converts guanosine 5’-triphosphate (GTP) into cGMP and pyrophosphate. The enzyme has

NON-cGMP MEDIATED EFFECTS OF NO

215

some residual basal activity in the resting state but is activated up to 500-fold upon interaction of the ferrous iron of the enzyme’s heme moiety with NO (94, 95). The activation of sGC has diverse physiological effects in cardiovasculature, platelet function, neurotransmission, and other cellular aspects (18). sGC is activated by NO at a fairly low concentration (10-100 nM). Many researchers regard the effective NO concentration as crucial in determining whether NO effects are physiological or pathophysiological. However, the effective concentrations of NO in vitro or the concentrations of exogenously added NO donors in tissue cultures do not always reflect the local effective concentration of NO. In addition, the effective concentration of NO can be both positively and negatively influenced by the presence of other radicals or reactive oxygen species. The low effective concentration of NO, in the case of sGC, reflects the high affinity of NO for the sGC heme

moiety. The on-rate of NO-heme binding for sGC is rather fast, in the order of

10’-10° M~'s~! (96, 97), which is in the same range as many ferrous-heme proteins (98). The geometry of NO binding to ferrous heme is similar to oxygen binding. Both NO and O, bind in a 130-150° angle towards the ligand-iron axis. NO binding to sGC disrupts the bond between the iron and coordinating His105 residue of sGC (94, 99), which may be involved in the catalytic process. The disso-

ciation rate of NO from sGC depends on the state in which the enzyme is present. Early measurements indicated that the iron-heme complex is rather stable (93), in contrast with the biological function of sGC as a fast regulator (93). However, in

the presence of GTP and Mg?* the dissociation rate is increased at least 100-fold (96, 97), which indicates that the heme microenvironment changed. This is cor-

roborated by the disturbance in the Raman spectrum of the NO-iron bond (100) upon treatment with GTP. Thus, the activation and deactivation of sGC are both likely to be very fast processes, which is in agreement with the rapid increases and decreases in cGMP levels after hormonal activation of NO/sGC (101).

Another well-documented effect of NO at low concentrations is the inhibition of the terminal complex IV (cytochrome oxidase) from the mitochondrial respiratory chain. NO-dependent inhibition of the respiratory chain is regarded as one mechanism of macrophage-derived cytotoxicity. Cytochrome c oxidase (CcO) is a complex of 13 subunits, containing 2 hemes (cyt a and cyt a3) and 2 copper centers (CuA and CuB). All four redox-active metal centers have different functions in

CcO catalytic activity. Low concentrations of NO cause immediate inhibition of oxygen consumption (102-104). NO acts as a potent, rapid, and reversible inhibitor with a half-inhibitory concentration in the range of 60-270 nM (104), depending on the oxygen concentration. The binding of NO to the reduced cyt a* is very fast

and comparable to oxygen binding, with a rate of 0.4-1.0 x 10° M~'s~! (105). The dissociation of NO is also rapid (0.13 s~') (106). The exact mechanism of CcO inhibition by NO is not completely defined. Some authors have argued that the rapid inhibition could be explained entirely by NO binding to the reduced cyt a; site (106), which is also the site of O, binding. Others have suggested that NO binds to the CuB site, which gives NO advantage over oxygen in binding to cyt a; (107). The formation of a stable Cu’*-NO complex as a mechanism of CcO inhibition has also been

216

DAVIS ET AL

proposed (108). NO also reacts with the oxidized CcO; however, it has been suggested that this interaction forms an irreversible bridged complex, a;7*-NO-Cup~*, rather than a reversible complex, a,°*-Cu,?*-NO, after NO binding to reduced CcO (109). Although inhibition of CcO with NO occurs in the physiological range of NO

concentration, the biological significance of such inhibition has not been clar-

ified. Some researchers hypothesize that the balance between intracellular levels of O, and NO dictate the ultimate respiratory rate of the mitochondria (110). Catalase is a ferric heme protein that is critical for intracellular degradation of hydrogen peroxide. As a rule, NO binds less tightly to ferric iron (98); however, when the iron coordination with the distal ligand is weak or absent, as in the case

of catalase, the on-rate for NO can be reasonably fast:3 x 10’M~'s~! (111). NO binding to ferric heme is less reversible, which in the case of catalase results in the NO-dependent inhibition of the enzyme with a K; of 0.18 uM (112). An increase in NO production or addition of NO-donors results in a decrease of cellular hydrogen peroxide consumption (113, 114). Although the role of catalase inhibition is unclear, NO-dependent inhibition of catalase resulting in increased hydrogen peroxide concentrations may enhance the cytotoxic effect of NO.

Nonheme Iron Proteins NO can also react with iron in FeS clusters. The early studies of macrophage cytotoxicity demonstrated the inhibition of mitochondrial respiration (115) in the complex I and II (116) in tumor cells cocultured with activated macrophages. Changes in activities indicated that aconitases were most sensitive and were inhibited first, followed by complex I and complex II (117,118). The richness of these complexes in [4Fe-4S] clusters designated them as primary targets of what was later shown to be a nitric oxide—mediated effect. Aconitases are a family of dehydratases that catalyze the reversible isomerization of citrate and isocitrate via cis-aconitate (119). These enzymes contain unique [4Fe-4S] clusters in which one of the irons, Fe,, is not ligated to a protein residue but rather to a hydroxide from solvent. The substrate reacts with this apical iron. A direct NO interaction with the mitochondrial FeS clusters has been proposed, based on the detection of a paramagnetic g = 2.03 electron paramagnetic resonance (EPR) signal interpreted as the formation of nitrosyl-iron-sulfur complexes (120). However, the question of whether mitochondrial FeS clusters participate in the formation of a P= 23 complex [also referred to as dinitrosyl-iron-dithiol (DNIC)] remains unsettled. The cytosolic form of aconitase is a bifunctional protein, carrying a [4Fe-4S] cluster that acts as a regulator of enzyme function (121). As a holoenzyme form, the enzyme has aconitase activity, which can be inhibited by NO-treatment and protected by the presence of the enzyme’s substrate, citrate. This suggests a direct interaction of NO with the apical iron, Fe,. In an apoenzyme form, lacking the [4Fe-4S], the enzyme acts as a posttranscriptional regulator, known as iron regulatory protein (RP) (122). IRP acts as a translational regulator of the intracellular

iron pool. IRP binds to the iron response element (IRE) of the 5’ untranslated

NON-cGMP MEDIATED EFFECTS OF NO

217

regions of ferritin and erythroid 5-aminolevulinilate synthase to attenuate their translation (123) or to the 3’ region of the transferrin receptor to stabilize its mRNA

(124). In the iron-depleted cells, the IRP is in the opened apo-conformation and binds to RNA, whereas in the iron-supplied cells the IRP retains its [4Fe-4S] cluster and acts as an aconitase (125). The activated macrophages produce NO, exhibit the g = 2.03 EPR signal, and display a decrease in aconitase activity and an increase in IRE-binding activity of IRP (126, 127). Thus, it is compelling to assume that NO has a direct effect on the conversion of [4Fe-4S]-containing cytosolic aconitase into a [4Fe-4S]-deficient IRP as a result of nitrosylation of the apical Fe,. However, there is little evidence indicating that the direct NO binding to aconitase strips it of its [4Fe-4S]-cluster and converts it directly to IRP. Although the direct binding of NO to the apical Fe, converts aconitase into inactive [3Fe-4S]-enzyme, these forms of aconitase do not bind IRE (125), which suggests that NO-dependent modulation of aconitase-IRP conversion can either be nondirect or involve more potent reactive nitrogen species. It is also possible that NO prevents the assembly of a [4Fe-4S]-cluster rather than dismantling of the aconitase into the iron-free

enzyme. Some of the possible mechanisms are discussed in more detail in recent reviews (122, 128).

An interesting example of direct interaction between NO and FeS clusters is presented by the Escherchia coli SoxR activator protein. SoxR is a 17-kDa DNAbinding factor regulating the expression of the SoxS protein, which activates the transcription of all protein members of the oxidative stress soxRS regulon. The upregulation of proteins by the SoxRS regulon results in increased resistance to oxidative stress. SoxR protein forms a homodimer containing a [2Fe-2S] cluster. This cluster exists in the reduced state under normal aerobic growth, but can be

oxidized upon cell exposure to a number of agents, including superoxide or NO (129, 130). Oxidation by NO or superoxide switch the protein into an activation mode, which leads to a 100-fold stimulation of soxS expression (131, 132). SoxR protein treated with NO displays a g = 2.03 EPR spectrum (130, 133) typical for dinitrosyl-iron-dithiol complexes (120). NO activation of the soxRS regulon results in an increased bacterial resistance to activated macrophages (133) and

represents one of the potential defensive mechanisms of bacterial cells against oxidative stress in general and NO-mediated toxicity in particular. Ferritin, a protein crucial in the regulation of cellular iron pool availability, displays three types of EPR signals attributed to iron-nitrosy! complexes at imidazole groups of histidine, thiol groups of cysteine, and carboxylate groups of aspartate and glutamate (134). When the reaction between NO, apo-metallothionein (or Zn-metallothionein), and iron was examined by electron spin resonance spec-

troscopy, paramagnetic products with g values of 2.013 and 2.039 were detected. These EPR spectra are similar to dinitrosyl-iron-dithiol complexes (135). However,

it is still not clear whether this in vitro treatment with NO

reflects

the physiological or pathophysiological events taking place under NO stress. The modification of ferritin by NO may be an important step in the NO-dependent regulation of the intracellular iron pool.

218

DAVIS ET AL

Interaction of NO with Other Transition Metals The role of biological Zn is best characterized for the Zn-finger proteins. Zn is coordinated mainly by conserved cysteine and histidine residues. Nearly 1% of human genes may encode Zn-finger proteins (136). Other Zn-chelating structures have also been

described

(for review

see e.g.

137). The

effect of NO

on the function of Zn-finger-containing nuclear receptors was recently reported (138). The specific interaction of the heterodimeric complex of two Zn-finger transcription factors, 1,25-dihydroxyvitamin D3 [1,25(OH)2D3] receptor (VDR) and retinoid X receptor (RXR) with 1,25(O0H)2D3 response elements (VDREs), was used as a model system. NO caused a dose-dependent inhibition of VDRRXR-VDRE

complex formation (138).

It should be noted, however, that these

in vitro effects were observed at a rather high concentration of NO (ICSO values 0.5—0.8 mM). In vitro treatments of the Zn-finger containing DNA repair enzyme formamidopyridine-DNA glycolyase (Fpg) resulted in inactivation of Fpg (135a). The enzyme was protected, if free cysteine was present during incubations, suggesting that the cysteine-residues coordinating Zn may be the primary targets of NO and RNS rather the Zn-ion itself. Considering the staggering amount of proteins potentially containing coordinated Zn (136) and their function as regulators of expression, such NO-dependent changes of their properties could have an important role in the regulation of transcription under NO stress. In recent years, interesting data have emerged that demonstrate a rather complicated crosstalk between nitric oxide/RNS and the tumor suppressor p53. p53 is a transcriptional regulator that has a unique DNA-binding domain structure. The DNA-binding domain is made up of an array of two-beta sheets supporting large loop-helix structures directly involved in contacting DNA (139). These loops are bridged together by the coordination of the divalent Zn atom with three cysteines and one histidine. Expression of p53 in a variety of human cell lines and in murine fibroblasts downregulates the transcription from the NOS2 promoter, which suggests a negative feedback loop that protects cells from NO-induced damage (140, 141). However, NO donors have also been shown to induce conformational

changes in p53, resulting in the inactivation of the DNA-binding properties of the protein (142). Although the exact nature of these NO-dependent changes has not been elucidated, nitrosylation of the Zn ion is one possible mechanism. Such disruption of p53 function by NO may represent a mechanism for its inactivation in some cancer or precancer conditions.

INTERACTION OF NO WITH RADICAL RESIDUES Reaction with Protein Radicals NO is a paramagnetic molecule that is capable of reactions with other radicals. The interaction of NO with O,*~ and other free radical molecules has a significant

biological importance, as discussed elsewhere in this review. A vast number of

NON-cGMP MEDIATED EFFECTS OF NO

219

biochemical reactions proceed via radical intermediates. Some of these reactions utilize tightly controlled protein-bound amino acid radical residues as cofactors. The amino acid involved can be a tyrosine (ribonucleotide reductase class I, photosystem II, prostaglandin H synthase), a modified tyrosine (amine oxidase, galactose oxidase), a tryptophan (cytochrome c peroxidase), a modified tryptophan (methylamine dehydrogenase) or a glycine (ribonucleotide reductase class III, pyruvate formate lyase) (for review see 143). Such protein radicals are likely targets for NO and RNS.

Ribonucleotide Reductases Ribonucleotide reductases (RNRs) catalyze the reduction of ribonucleotides to

deoxyribonucleotides crucial for the synthesis of DNA. The class I RNR requires a tyrosyl radical for its enzymatic activity. The tyrosyl radical of RNR is stabilized owing to its delocalization over the aromatic ring and its position in the protein core with the lack of oxidizable amino acid side chains in close proximity (144, 145). The tyrosyl radical is magnetically coupled with a pair of nonheme iron atoms involved in the production and stabilization of the radical (146). Al-

though the diferric center of RNR is not very sensitive to NO (147), the enzyme is rather susceptible to NO (148, 149), presumably owing to NO scavenging of the tyrosyl radical (150). It was estimated that 0.5—3.0 uM NO scavenge 100% of the RNR tyrosyl radicals available in mammalian cells (151). This amount of NO

can be attained through activation of iNOS. The inhibition of RNR by NO results in inhibition of DNA synthesis and is widely regarded as one of the important mechanisms of macrophage-dependent cytotoxicity.

Prostaglandin H Synthase Prostaglandin H synthases (PGHS) (cyclooxygenases) catalyze the first two steps in the biosynthesis of prostanoids. The involvement of the tyrosyl radical of PGHS in catalysis was demonstrated by EPR studies (152) and later confirmed by other methods (153, 154). The PGHS tyrosyl radical is considerably less stable [halflife =

20s at —12°C (152)] compared with the RNR tyrosyl radical. The inter-

action between NO and PGHS tyrosyl radical was documented only recently by low-temperature EPR spectroscopy and by the presence of nitrotyrosine modification of the catalytically active tyrosine residue (62, 155). In contrast to ribonucleotide reductase, RNS may activate PGHS (156). Some investigators showed that PGHS-1 is inhibited by NO donors but stimulated by compounds capable of generating peroxynitrite (157).

NO MODIFICATIONS OF DNA NO is probably insufficiently reactive to attack DNA directly, but numerous RNS and their CO,” or Cl” adducts can oxidize, nitrate, or deaminate genomic DNA, resulting in strand breaking and mutations (158-160). It is unlikely that O,, O,°~,

220

DAVIS ET AL

and HO, are capable of reacting directly with DNA, at least at their physiological levels, but the hydroxyl radical (HO*) derived from H,O, generates a multiplicity of products from all four DNA bases (158-160). However, the physiological role of HO® is uncertain because of its extremely high reactivity. Because HO® production in living cells requires the presence of a transition metal (such as Fe7*), it could damage DNA only when the transition metal is immediately adjacent to the DNA. Because metal ions appear to be mainly in a form that is unable to catalyze free radical reactions, their availability for reaction with HO® is restricted in vivo. In contrast, peroxynitrite can travel up to 9 wm (161) and easily pass through biological membranes (162). Thus, peroxynitrite is more likely to modify DNA than OH’. Studies have shown that peroxynitrite preferentially reacts with guanine, but because of the various reactions that peroxynitrite can undergo, there are also a variety of products observed. The reaction of peroxynitrite with guanine yields 8-oxo-deoxyguanosine (8-oxo-dG), one of the most abundant products of DNA oxidation by peroxynitrite (163). 8-oxo-dG is known to cause GC —

TA transi-

tions, mispairing, and ultimately, mutations. Once 8-oxo-dG is formed, it becomes approximately 1000-fold more reactive with peroxynitrite than its precursor, dG (164). Depending on the ratio of peroxynitrite over 8-oxo-dG, the oxidation reaction may continue until the oxidized product, oxaluric acid, is formed (165). Cyanuric acid, oxazolone, and 4-hydroxy-8-oxo-4,8-dihydro-2 -deoxyguanosine may also be generated as a result of secondary oxidation of 8-oxo-dG (158). 8-oxo-dG is formed by many DNA-damaging agents in vivo, as well as by ONOO-. In contrast, 8-nitro-dG formed by the reaction of peroxynitrite with DNA is more specific for peroxynitrite-induced damage to DNA. Recently, another proteinnitrating agent, nitryl chloride, has been reported (166). However, it is unclear whether nitryl chloride can nitrate guanine in the same manner as peroxynitrite. Peroxynitrite treatment also causes DNA strand breaks. Generally, peroxynitrite can abstract hydrogen from sugar moieties and form sugar radicals leading to sugar fragmentation and DNA strand breaks (167). The formation of 8-nitro-dG by reaction with peroxynitrite may also favor the creation of basic sites favorable for cleavage by endonucleases Another powerful reactive nitrogen species, N,O3, acts as a nitrosating agent. NO; may damage DNA directly through the nitrosation of primary amines on DNA bases or indirectly through reactions with primary and secondary amines, ultimately leading to DNA deamination. This chemistry was recently reviewed in detail (158). Deamination is the replacement of an exocyclic amino group by a hydroxyl group. Therefore, any DNA base containing such an amino group can be deaminated in the reaction with N,O;. The major consequences of this deamination reaction are GC + AT, GC — TA, and AT - GC transitions and single-strand breaks (168, 169). N50; may also cause intra- and interstrand cross-links that, even formed in very small amounts, disrupt gene expression. Direct chemical modification of DNA by reactive nitrogen species derived from NO may be an important contributor to the age- and inflammation-related

NON-cGMP MEDIATED EFFECTS OF NO

221

development of cancer or other diseases. There are three major techniques used to identify and quantify oxidized DNA products: gas chromatography/mass spectroscopy with selected ion monitoring (170), reverse phase HPLC with electrochemical detection (171), and liquid chromatography/electrospray ionization mass spectrometry (165). Each of these techniques may produce artifacts that have created some discrepancy in the field, especially in terms of quantities of oxidized DNA products. These discrepancies have also raised the question—how high is the steady-state level of DNA damage in normal and pathological conditions? Keeping in mind the existence of nonexpressed DNA and well-balanced DNA repair systems, the contribution of DNA oxidation to age-related development of cancer is unclear. Despite debatable quantitative issues, 8-oxo-dG generation was used for years to monitor DNA damage resulting from treatment with various chemical and physical agents. Although there is no direct evidence that 8-oxo-dG causes cancer, several studies have shown a correlation between elevated levels of 8-oxo-dG and carcinogenesis (172-174). It is generally accepted that the high frequency of point mutations in certain common human tumors can be induced by exposure to reactive nitrogen species. It was shown that excess production of NO in chronic inflammation causes DNA damage, inhibits DNA repair, and may link inflammation and cholangiocarcinoma (175). RNS production has been linked to human colon adenomas and carcinomas (176) and to breast (177) and gastric (178) cancer. Reactive nitrogen species can also cause mutations in cancer-related genes, such as tumor suppressor gene p53 (179). In addition to nuclear DNA, there is also mitochondrial DNA (mtDNA). Mammalian mtDNA codes 13 subunits of respiratory chain complexes and its own structural rRNAs

and tRNAs.

MtDNA

is a 16,569-bp double-stranded circular

DNA that is mutated much faster than nuclear DNA, presumably because mtDNA is not protected by organization along histones. It has no introns, and a random hit will inevitably cause damage or mutation with serious consequences. It is important to note that mtDNA contains unusually high amounts of direct repeats that may give rise to large-scale deletions by mispairing during oxidized DNA replication. So far, more than 50 pathogenic mtDNA mutations have been found that are associated with or responsible for specific human diseases (159, 180, 181). Five different types of mtDNA mutations were found: deletions, point mutations, insertion, tandem duplications, and DNA rearrangements. This broad spectrum

of mutations of mtDNA accumulates in various human tissues and accompanies aging. Because mtDNA is located in the vicinity of the reactive oxygen species generation site in the mitochondrial inner membrane, investigators hypothesized that reactive oxygen species—associated oxidation causes damage to this genome (180). However, recent evidence that mitochondria may be a significant intracellular source of peroxynitrite may to a large degree refocus research. The existence of a mitochondrial NOS has been reported for several rat tissues (182, 183). The actual production of NO in mitochondria has also been demonstrated (184, 185).

Furthermore, NO can diffuse freely and is more soluble in organic phases than in water. Mitochondria are also the major source of O,°” production in cells. Any

222

DAVIS ET AL

perturbation capable of decreasing the coupling efficiency of the mitochondrial respiratory chain generates O,°-. The well-known reaction between NO and O,°~ generates ONOO ; however, ONOO™ may also be formed by the reaction of nitroxyl anion (NO~) with O,*~ (186). Recently, it was reported that cytochrome c can catalyze the reduction of NO to NO™ (187), which raises the possibility that the formation of NO~ may be an important source for the production of ONOOin mitochondria. Recent studies indicate that oxidative damage of mtDNA results in fragmentation (188). However,

the level of the oxidized bases in full-size mtDNA,

the

only template for replication, is rather low. This finding indicates that, despite the extensive mtDNA oxidation, efficient mtDNA repair and degrading systems decrease accumulation of negative changes. Although the general consensus is that mtDNA is subject to severe oxidative damage, the critical questions of what type of reactive species participate in this damage and what kind of contribution oxidative damage makes to human diseases and aging remain to be addressed.

NO MODIFICATIONS OF LIPIDS Numerous mechanisms of lipid oxidation, both enzymatic and nonenzymatic, have been thoroughly studied in vitro. It is also well established that lipid oxidation is a typical feature of inflammatory diseases. However, there is a very little understanding of the oxidative mechanisms in vivo and how they impinge upon lipidmediated signal transduction, integrity, and fluidity of biological membranes, and enzymatic properties. Identification of specific oxidants that are responsible for lipid oxidation in inflammatory conditions remains a major priority. Reactive oxygen species derived from NO may interact with unsaturated lipids (189). Chemical mechanisms for these reactions generally fall into two categories: oxidation and nitration. A complicated and incompletely understood set of factors determines which lipid oxidation products are generated in vivo in each particular case. However, it is clear that the role of NO is different in the presence of o> compared with the absence of O,°~ (22). In the absence of O,°~, NO may terminate lipid oxidation (190). Several mechanisms for this termination have been described, namely (a) NO trapping of alkyl, aloxyl, and peroxyl lipid—derived radicals; (b) NO regulation of the activity of enzymes such as cyclooxygenase, lipoxygenase, and cytochrome P-450; (c) NO regulation of cell signaling not directly associated with lipid oxidation; and (d) NO binding to redox-active metal centers, which inhibits metal-catalyzed HO® generation and lipid peroxidation. In the presence of O,°~, the NO concentration is decreased and a variety of oxidant and nitrating agents are formed. These nitrating and oxidizing agents react with lipids to form several oxidation products, which may be subsequently nitrated to form different products. Because O, may also react with lipid oxidation products, the concentration of O, may determine whether nitration occurs (191).

NON-cGMP MEDIATED EFFECTS OF NO

223

In vivo the high concentration of thiols may successfully compete with lipids for reactive nitrogen species. ONOO7“ is one of the major nitrating species. ONOO~ may be formed in either the aqueous or the hydrophobic phase of the cell. Presumably, peroxynitrite undergoes different chemical reactions that contribute differently to lipid oxidation in these two compartments. It is well established that ONOO™ rapidly permeates model phospholipid membranes in vitro (162) and moves through intact erythrocyte or mitochondrial membranes (192). Because ONOO™ reacts rapidly with many substances, it is not clear how much peroxynitrite, formed in the aqueous phase, contributes to lipid oxidation. Regardless of the source of reactive nitrogen species, the hydrophobic phase has been shown to be more favorable for their reactions than the aqueous phase (54, 162, 193). It was shown that NO is more soluble in the hydrophobic phase (194), and the reaction of NO witii O, within membranes is approximately 300 times more rapid than in the surrounding aqueous medium (193). The membrane permeability coefficient for peroxynitrite is close to that reported for water (195). Therefore, peroxynitrite can be expected to have free access to hydrophobic compartments in cells and to hydrophobic structures like atherosclerotic plagues, myelin sheaths, or the lining of the lung. The life span for peroxynitrite is expected to be longer in the hydrophobic phase than the aqueous phase. Thus, its capacity to accumulate in the hydrophobic phase can play a critical role in regulating membrane and lipoprotein lipid oxidation reactions. Besides oxidation and nitration, other reactions of reactive nitrogen species with lipids that may occur in vitro include decarboxylation of free fatty acids (196). Reactions with aliphatic or aromatic alcohol groups, resulting in alkyl or aryl nitrites, are also possible. Because these reactions are rather slow, their physiological relevance is unclear. Oxidation and nitration convert low density lipoproteins into an atherogenic high-uptake form (197). It has been shown (198, 199) that a variety of halogenating and nitrating intermediates may be generated by the myeloperoxidase-H,O,Cl system in the presence of nitrite (NO,), the autoxidation product of NO. Their potential roles are still obscure. The latest statement is that nitrating intermediates are more effective than chlorinating intermediates in promoting oxidative conversion of LDL into a stable high-uptake form (197). However, the biological consequences of oxidative modification of LDL, as well as the precise oxidative and nitrating intermediates, remain to be established.

MISCELLANEOUS Various

antioxidants

may react with peroxynitrite and inhibit peroxynitrite-

mediated oxidation and nitration reactions (200). The reported list includes ascorbate (201, 202), uric acid (203), bilirubin (204), vitamin E (205), catecholamines (206), flavonoids (207), melatonin (208), glutathione (209), and B-carotene (210).

224

DAVIS ET AL

Although the apparent rates of these reactions are not very fast, they may occur in vivo and accelerate ONOO™ decay. For example, ONOO-mediated hemolysis is effectively inhibited by glutathione (211). Another point of interest is the generation of stable NO donors in peroxynitritedependent reactions. For example, peroxynitrite can react with compounds containing an alcohol functional group, e.g. D-glucose (212). This could be an additional pathway for peroxynitrite decay. However, this also could represent an additional way to recycle reactive nitrogen species and generate compounds with cytoprotective properties, namely longer-lived NO donors. Formation of low— molecular weight nitrosothiols probably falls in this category.

Summary The chemistry of NO, with its diverse array of modifications and products, is obviously complex. This unique free-radical gas can participate in cellular signaling and regulation in a variety of ways, which we have attempted to summarize in this review. The reactions of NO, and consequently the effects elicited by NO, are a function not only of the concentration and location of NO but also the surrounding milieu in which NO is produced. NO or the products of the reaction of NO with O, and O,°~ can modify many different macromolecules, including proteins, lipids, and nucleic acids, to produce both physiological and pathophysiological effects. Although there are more than 30,000 publications on NO, numerous questions remain unanswered. ACKNOWLEDGMENTS The authors would like to thank the John S. Dunn Foundation, the Mathers Foundation, the Welch Foundation, the National Aeronautical Space Association, the

Army Defense Research Program, and the University of Texas for their generous support. Karen L Davis is supported by a National Research Service Award from the National Institutes of Health (HL10046-02). Visit the Annual Reviews home page at www.AnnualReviews.org LITERATURE CITED 1. Murad F. 1996. The 1996 Albert Lasker Medical Research Awards. Signal transduction using nitric oxide and cyclic guanosine monophosphate. JAMA 276(14):1189-92 2. Schmidt HH, Walter U. 1994. NO at work. Cell 78(6):919-25 3. Marletta MA. 1994, Nitric oxide synthase: aspects concerning structure and catalysis. Cell 78(6):927-30

4. Jang D, Murrell GA. 1998. Nitric oxide in arthritis. Free Radic Biol. Med. 24(9):151119 5. Kooy NW, Lewis SJ, Royall JA, Ye YZ, Kelly DR, Beckman JS. 1997. Extensive tyrosine nitration in human myocardial inflammation: evidence for the presence of peroxynitrite. Crit. Care Med. 25(5): 812-19

~

225

NON-cGMP MEDIATED EFFECTS OF NO . Iwashita E, Miyahara T, Hino K, Tokunaga T, Wakisaka H, Sawazaki Y. 1995. High nitric oxide synthase activity in endothelial

biology of nitric oxide: insights into regulatory, cytotoxic, and cytoprotective mechanisms of nitric oxide. Free Radic. Biol. Med. 25(4-5):434-56

cells in ulcerative colitis. J. Gastroenterol.

30(4):55 1-54 . Hogaboam CM, Jacobson K, Collins SM, Blennerhassett MG. 1995. The selective beneficial effects of nitric oxide inhibition in experimental colitis. Am. J. Physiol. Gastrointest. Liver Physiol. 268(4 Pt 1):G673-84 . Kelly CJ, Gold DP. 1999. Nitric oxide in interstitial nephritis and other autoimmune diseases. Semin. Nephrol. 19(3):288—-95 . Abe K, Pan LH, Watanabe M, Kato T, Ito-

yama Y. 1995. Induction of nitrotyrosinelike immunoreactivity in the lower motor neuron of amyotrophic lateral sclerosis. Neurosci. Lett. 199(2):152-54

10. Wink DA, Vodovotz Y, Laval J, Laval F, Dewhirst MW, Mitchell JB. 1998. The multifaceted roles of nitric oxide in cancer. Carcinogenesis 19(5):711-21 le Moochhala S, Rajnakova A. 1999. Role of nitric oxide in cancer biology. Free Radic. Res. 31(6):671-79 12; Chan NN, Vallance P, Colhoun HM. 2000. Nitric oxide and vascular responses in Type I diabetes. Diabetologia 43(2):137-47

13). Sjoholm A. 1998. Aspects of the involvement of interleukin-1 and nitric oxide in the pathogenesis of insulin-dependent diabetes mellitus. Cell Death Differ. 5(6): 461-68 14. Bolanos JP, Almeida A, Stewart V, Peuchen S, Land

JM, et al. 1997.

Nitric

17. Grisham MB, Jourd’Heuil D, Wink DA. 1999. Nitric oxide. I. Physiological chemistry of nitric oxide and its metabolites: implications in inflammation. Am. J. Physiol.

Liver Physiol.

18. Martin E, Davis K, Bian K, Lee YC, Murad F. 2000. Cellular signaling with nitric oxide and cyclic guanosine monophosphate. Semin. Perinatol. 24(1):2-6 19. Wink DA, Osawa Y, Darbyshire JF, Jones CR, Eshenaur SC, Nims RW.

1993. In-

hibition of cytochromes P450 by nitric oxide and a nitric oxide-releasing agent. Arch. Biochem. Biophys. 300(1): 115-23 20. Ford PC, Wink DA, Stanbury DM. 1993. Autoxidation kinetics of aqueous nitric oxide. FEBS Lett. 326(1-3):1-3 20a.

Goldstein S, Czapski G. 1996. Mechanism of the nitrosation of thiols and amines by oxygenated NO® solutions; the nature of the nitrosating intermediates. J. Amer. Chem. Soc. 118:3419-25

20b. Liu X, Miller MJS, Joshi MS, Thomas DD, Lancaster JR. 1998. Accelerated re-

action of nitric oxide with O, within the hydrophobic interior of biological membranes. Proc. Natl. Acad. Sci. USA 95:2175-79 20c.

oxide-

mediated mitochondrial damage in the brain: mechanisms and implications for neurodegenerative diseases. J. Neurochem. 68(6):2227—-40

15! Schulz JB, Matthews RT, Klockgether T, Dichgans J, Beal MF. 1997. The role of mitochondrial dysfunction and neuronal nitric oxide in animal models of neurodegenerative diseases. Mol. Cell. Biochem. 174(1-2):193—97 16. Wink DA, Mitchell JB. 1998. Chemical

Gastrointest.

276(2 Pt 1):G315-21

2

Malinski T, Taha Z, Grunfeld S, Patton S, Kapturczak M, Tomboulian P. 1993.

Diffusion of nitric oxide in the aorta wall monitored in situ by prophyrinic microsensors. Biochem. Biophys. Res. Commun. 193:1076-82 Huie RE, Padmaja S. 1993. The reaction of NO with superoxide. Free Radic. Res. Commun.

Pp), Rubbo

18(4):195—99

H, Radi

R, Trujillo

leri R, Kalyanaraman

M,

B, et al.

Tel1994.

Nitric oxide regulation of superoxide and peroxynitrite-dependent lipid

226

DAVIS ET AL

peroxidation. Formation of novel nitrogencontaining oxidized lipid derivatives. J. Biol. Chem. 269(42):26066-—75

23, Lander HM, Ogiste JS, Pearce SF, Levi R, Novogrodsky

A.

1995.

Nitric oxide-

stimulated guanine nucleotide exchange on p2Iras. J. Biol. Chem. 270(13):7017—20

24. Xu L, Eu JP, Meissner G, Stamler JS. 1998. Activation of the cardiac calcium release channel (ryanodine receptor) by poly-Snitrosylation. Science 279(5348):234—37 Mah, Stamler JS, Toone EJ, Lipton SA, Sucher NJ.

1997.

(S)NO

signals:

translocation,

regulation, and a consensus motif. Neuron 18(5):691-96 26. McBain CJ, Mayer ML. 1994. N-methyID-aspartic acid receptor structure and function. Physiol. Rev. 74(3):723-60 27 Garthwaite J, Charles SL, Chess-Williams R. 1988. Endothelium-derived relaxing factor release on activation of NMDA receptors suggests role as intercellular messenger in the brain. Nature 336(6197):385—

88 28. Bredt DS, Hwang PM, Snyder SH. 1990. Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature 347(6295):768-70 2S), Manzoni O, Bockaert J. 1993. Nitric oxide synthase activity endogenously modulates NMDA receptors. J. Neurochem. 61(1):368-70 30. Choi YB, Tenneti L, Le DA, Ortiz J, Bai G, et al. 2000. Molecular basis of NUDA receptor-coupled ion channel modulation by S-nitrosylation. Nat. Neurosci. 3(1):1521 Shh Lipton SA, Rosenberg PA. 1994. Excitatory amino acids as a final common pathway for neurologic disorders [see comments]. N. Engl. J. Med. 330(9):613—22

. Lipton SA, Choi YB, Pan ZH, Lei SZ, Chen HS, et al. 1993. A redox-based mech-

anism for the neuroprotective and neurodestructive effects of nitric oxide and related nitroso-compounds [see comments]. Nature 364(6438):626—32

ia Broillet MC, Firestein S. 1996. Direct activation of the olfactory cyclic nucleotidegated channel through modification of sulfhydryl groups by NO compounds. Neuron 16(2):377-85

34. Broillet MC. 2000. A single intracellular cysteine residue is responsible for the activation of the olfactory cyclic nucleotidegated channel by NO. J. Biol. Chem. 275(20):1513541 Sp Lander HM, Hajjar DP, Hempstead BL, Mirza UA, Chait BT, et al. 1997. A molecular redox switch on p21(ras). Structural basis for the nitric oxide-p21(ras) interaction. J. Biol. Chem. 272(7):4323—26

36. Noda M, Ko M, Ogura A, Liu DG, Amano T, et al. 1985. Sarcoma viruses carrying ras oncogenes induce differentiationassociated properties in aneuronal cell line. Nature 318(6041):73-75

Sik Bar-Sagi D, Feramisco JR. 1985. Microinjection of the ras oncogene protein into PC12 cells induces morphological differentiation. Cel] 42(3):841—48

38. Borasio GD, Markus A, Wittinghofer A, Barde YA, Heumann R. 1993. Involvement of ras p21 in neurotrophin-induced response of sensory, but not sympathetic neurons. J. Cell Biol. 121(3):665—72

. Szeberenyi J, Cai H, Cooper GM. 1990. Effect of a dominant inhibitory Ha-ras mutation on neuronal differentiation of PC12 cells. Mol. Cell. Biol. 10(10):5324— 32 40. Teng KK, Esposito DK, Schwartz GD, Lander HM, Hempstead BL. 1999. Activation of c-Ha-Ras by nitric oxide modulates survival responsiveness in neuronal PC12 cells. J. Biol. Chem. 274(52):37315-20 41. Li J, Billiar TR, Talanian RV, Kim YM. 1997. Nitric oxide reversibly inhibits seven members of the caspase family via S-nitrosylation. Biochem. Biophys. Res. Commun. 240(2):419-24 42. Nicholson DW, Ali A, Thornberry NA, Vaillancourt JP, Ding CK, et al. 1995. Identification and inhibition of the ICE/CED-3

NON-cGMP MEDIATED EFFECTS OF NO

227

protease necessary for mammalian apoptosis [see comments] Nature 376(6535): 37-43

rat tissues that modifies nitrotyrosinecontaining proteins. Proc. Natl. Acad. Sci. USA 95(20):11584-89

43. Kim YM, Talanian RV, Li J, Billiar TR. 1998. Nitric oxide prevents IL-1beta and IFN-gamma-inducing factor (IL-18) release from macrophages by inhibiting caspase-1 (IL-lbeta-converting enzyme). J. Immunol. 161(8):4122-28

52. Gow AJ, Duran D, Malcolm S, Ischiropoulos H. 1996. Effects of peroxynitriteinduced protein modifications on tyrosine phosphorylation and degradation. FEBS Lett. 385(1-2):63-66 DS: Beckman JS, Ischiropoulos H, Zhu L,

44. Dimmeler S, Haendeler J, Nehls M, Zei-

van der Woerd M, Smith C, et al. 1992.

her AM. 1997. Suppression of apoptosis by nitric oxide via inhibition of interleukinlbeta-converting enzyme (ICE)-like and cysteine protease protein (CPP)-32-like proteases. J. Exp. Med. 185(4):601—7

Kinetics of superoxide dismutase- and iron-catalyzed nitration of phenolics by peroxynitrite. Arch. Biochem. Biophys. 298(2):438-45

4S. Mannick JB, Hausladen A, Liu L, Hess DT, Zeng M, et al. 1999. Fas-induced caspase

denitrosylation.

Science

284(5414):651-—

54 46. Mannick JB, Miao XQ, Stamler JS. 1997. Nitric oxide inhibits Fas-induced apoptosis. J. Biol. Chem. 272(39):24125-28 47. Molina y Vedia L, McDonald B, Reep Be

Brune

By

Dr

Silvioy

IVE

et

“al:

1992. Nitric oxide-induced S-nitrosylation of glyceraldehyde-3-phosphate dehydrogenase inhibits enzymatic activity and increases endogenous ADP-ribosylation. J. Biol. Chem. 267(35):24929-32. Erratum. 1993. J. Biol. Chem. 268(4):3016

48. Mohr S, Stamler JS, Brune B. 1996. Posttranslational modification of glyceraldehyde-3-phosphate dehydrogenase by S-nitrosylation and subsequent NADH at-

54. Ischiropoulos H, Zhu L, Beckman JS. 1992. Peroxynitrite formation from macrophage-derived nitric oxide. Arch. Biochem. Biophys. 298(2):446-51 215 Estevez AG, Spear N, Manuel SM, Radi R, Henderson CE, et al. 1998. Nitric

oxide and superoxide contribute to motor neuron apoptosis induced by trophic factor deprivation. J. Neurosci. 18(3): 923-31 56. Pfeiffer S, Mayer B. 1998. Lack of tyrosine nitration by peroxynitrite generated at physiological pH. J. Biol. Chem. 273(42):27280-85 56a.

JS. 2000.

Superoxide reacts with nitric oxide to nitrate tyrosine at physiological pH via peroxynitrite. J. Biol. Chem. In press

56b. Sawa

T, Akaike

T, Maeda

H.

2000.

Tyrosine nitration by peroxynitrite formed from nitric oxide and superoxide generated by xanthine oxidase. J. Biol. Chem. In press

tachment. J. Biol. Chem. 271(8):4209-14

49. Stamler JS. 1994. Redox signaling: nitrosylation and related target interactions of nitric oxide. Cell 78(6):931—36 50. Viner RJ, Ferrington DA, Williams TD, Bigelow DJ, Schoneich C. 1999. Protein modification during biological aging: selective tyrosine nitration of the SERCA2a isoform of the sarcoplasmic reticulum Ca?+-ATPase in skeletal muscle. Biochem. J. 340(Pt 3):657-69 ole Kamisaki Y, Wada K, Bian K, Balabanli B Davis K, et al. 1998. An activity in

Reiter CD, Teng R, Beckman

oh

Squadrito GL, Pryor WA. 1998. Oxidative

chemistry of nitric oxide: the roles of superoxide, peroxynitrite, and carbon dioxide. Free Radic. Biol. Med. 25(4-5):392403

58. Eiserich JP, Patel RP, O’Donnell VB. 1998. Pathophysiology of nitric oxide and related species: free radical reactions and modification of biomolecules. Mol. Asp. Med. 19(4-5):221-357

228

DAVIS ET AL

ae). Eiserich

JP,

Cross

CE,

Jones

AD,

Halliwell B, van der Vliet A. 1996. Forma-

tion of nitrating and chlorinating species by reaction of nitrite with hypochlorous acid. A novel mechanism for nitric oxidemediated protein modification. J. Biol. Chem. 271(32):19199—208

60. Wu W, Chen Y, Hazen SL. 1999. Eosinophil peroxidase nitrates protein tyrosyl residues. Implications for oxidative damage by nitrating intermediates in eosinophilic inflammatory disorders. J. Biol. Chem. 274(36):25933—-44 61. van Dalen CJ, Winterbourn CC, Senthilmohan

R,

Kettle

AJ.

2000.

Nitrite as a substrate and inhibitor of myeloperoxidase. Implications for nitration and hypochlorous acid production at sites of inflammation. J. Biol. Chem. 275(16):11638-44

62. Gunther MR, Hsi LC, Curtis JF, Gierse JK, Marnett LJ, et al. 1997. Nitric oxide trapping of the tyrosyl radical of prostaglandin H synthase-2 leads to tyrosine iminoxyl radical and nitrotyrosine formation. J. Biol. Chem. 272(27):17086-90 63. Reichard P, Ehrenberg A. 1983. Ribonucleotide reductase—a radical enzyme. Science 221(4610):514-19

64. Barry BA, Babcock GT. 1987. Tyrosine radicals are involved in the photosynthetic oxygen-evolving system. Proc. Natl. Acad. Sct. USA 84(20):7099-103 65. Beckman JS, Koppenol WH. 1996. Nitric oxide, superoxide, and peroxynitrite: the good, the bad, and ugly. Am. J. Physiol. Cell Physiol. 271(5 Pt 1):C1424—37

66. Buttery LD, Springall DR, Chester AH, Evans TJ, Standfield EN, et al. 1996. Inducible nitric oxide synthase is present within human atherosclerotic lesions and promotes the formation and activity of peroxynitrite. Lab. Invest. 75(1):77-85 67. MacMillan-Crow LA, Crow JP, Kerby JD, Beckman JS, Thompson JA. 1996. Nitration and inactivation of manganese superoxide dismutase in chronic rejection of

human renal allografts. Proc. Natl. Acad. Sci. USA 93(21):11853-58

68. Fukuyama

N, Takebayashi

Y, Hida

M,

Ishida H, Ichimori K, Nakazawa H. 1997.

Clinical evidence of peroxynitrite formation in chronic renal failure patients with septic shock. Free Radic. Biol. Med. 22(5):771-74 69. Singer I], Kawka DW, Scott S, Weidner JR, Mumford RA, et al. 1996. Expression of inducible nitric oxide synthase and nitrotyrosine in colonic epithelium in inflammatory bowel disease. Gastroenterology 111(4):871-85 70. Kaur H, Halliwell B. 1994. Evidence for nitric oxide-mediated oxidative damage in chronic inflammation. Nitrotyrosine in serum and synovial fluid from rheumatoid patients. FEBS Lett. 350(1):9-12 fat Myatt L, Rosenfield RB, Eis AL, Brockman DE, Greer I, Lyall F. 1996. Nitrotyrosine residues in placenta. Evidence of peroxynitrite formation and action. Hypertension 28:488—93 . Bachmaier K, Neu N, Pummerer C. Dun-

can GS, Mak TW, et al. 1997. iNOS expression and nitrotyrosine formation in the myocardium in response to inflammation is controlled by the interferon regulatory transcription factor 1. Circulation

96(2):585-91 . Ishiyama S, Hiroe M, Nishikawa T, Abe S, Shimojo T, et al. 1997. Nitric oxide contributes to the progression of myocardial damage in experimental autoimmune myocarditis in rats. Circulation 95(2):489—

96 74. Bian K, Davis K, Kuret J, Binder L, Mutad F. 1999. Nitrotyrosine formation with endotoxin-induced kidney injury detected by immunohistochemistry. Am. J. Physiol. Renal Physiol. 277(1 Pt 2):F33-40 ids Bosse HM, Bachmann S. 1997. Immunohistochemically detected protein nitration indicates sites of renal nitric oxide release in Goldblatt hypertension. Hypertension 30(4):948-52

NON-cGMP MEDIATED EFFECTS OF NO 76. Ara J, Przedborski S, Naini AB, JacksonLewis V, Trifiletti RR, et al. 1998. Inactivation of tyrosine hydroxylase by nitration following exposure to peroxynitrite and 1-methyl-4-phenyl-1,2,3,6tetrahydropyridine (MPTP). Proc. Nail. Acad. Sci. USA 95(13):7659-63 WH. Bruijn LI, Beal MF, Becher MW, Schulz JB, Wong PC, et al. 1997. Elevated free nitrotyrosine levels, but not protein-

bound nitrotyrosine or hydroxyl radicals, throughout amyotrophic lateral sclerosis (ALS)-like disease implicate tyrosine nitration as an aberrant in vivo property of one familial ALS-linked superoxide dismutase 1 mutant. Proc. Natl. Acad. Sci. USA 94(14):7606-11

78. Zou MH, Klein T, Pasquet JP, Ullrich V. 1998. Interleukin lbeta decreases prostacyclin synthase activity in rat mesangial cells via endogenous peroxynitrite formation. Biochem. J. 336(Pt 2):507—12

US. Zou MH, Leist M, Ullrich V. 1999. Selective nitration of prostacyclin synthase and defective vasorelaxation in atherosclerotic bovine coronary arteries. Am. J. Pathol. 154(5):1359-65 80. Zou M, Martin C, Ullrich V. 1997. Tyrosine nitration as a mechanism of selective inactivation of prostacyclin synthase by peroxynitrite. Biol. Chem. 378(7):707-13 81. Yamakura F, Taka H, Fujimura T, Mu-

rayama K. 1998. Inactivation of human manganese-superoxide dismutase by peroxynitrite is caused by exclusive nitration of tyrosine 34 to 3-nitrotyrosine. J. Biol. Chem. 273(23):14085—89 82. Strong MJ, Sopper MM, Crow JP, Strong WL, Beckman JS. 1998. Nitration of the

low molecular weight neurofilament is equivalent in sporadic amyotrophic lateral sclerosis and control cervical spinal cord. Biochem. Biophys. Res. Commun. 248(1):157-64 83. Chou SM, Wang HS, Taniguchi A. 1996. Role of SOD-1 and nitric oxide/cyclic GMP cascade on neurofilament

229

aggregation in ALS/MND. J. Neurol. Sci. 139(Suppl): 16-26

84. Viner RI, Huhmer AF, Bigelow DJ, Schoneich C. 1996. The oxidative inactivation of sarcoplasmic reticulum Ca(2+)ATPase by peroxynitrite. Free Radic. Res. 24(4):243-59 85. Haddad IY, Pataki G, Hu P, Galliani C, Beckman

JS, Matalon S. 1994. Quantita-

tion of nitrotyrosine levels in lung sections of patients and animals with acute lung injury. J. Clin. Invest. 94(6):2407-13

86. Zhu S, Haddad IY, Matalon S. 1996. Nitration of surfactant protein A (SP-A) tyrosine residues results in decreased mannose binding ability. Arch. Biochem. Bio-

phys. 333(1):282-90 87. Greis KD, Zhu S, Matalon S. 1996. Identification of nitration sites on surfactant protein A by tandem electrospray mass spectrometry. Arch. Biochem. Biophys.

335(2):396—402 88. Saleh D, Ernst P, Lim S, Barnes PJ, Giaid A. 1998. Increased formation of the potent oxidant peroxynitrite in the airways of asthmatic patients 1s associated with induction of nitric oxide synthase: effect of inhaled glucocorticoid.

FASEB

J.

12(11):929-

37

89. Kong SK, Yim MB, Stadtman ER, Chock PB. 1996. Peroxynitrite disables the tyrosine phosphorylation regulatory mechanism: Lymphocyte-specific tyrosine kinase fails to phosphorylate nitrated cdc2(6-20)NH2 peptide. Proc. Natl. Acad. Sci. USA 93(8):3377-82

90. MacMillan-Crow LA, Greendorfer JS, Vickers SM, Thompson JA. 2000. Tyrosine nitration of c-SRC tyrosine kinase in human pancreatic ductal adenocarcinoma. Arch.

Biochem.

Biophys.

377(2):350-

56 Dil. Di Stasi AM, Mallozzi C, Macchia G, Petrucci TC, Minetti M. 1999. Peroxynitrite induces tryosine nitration and modulates tyrosine phosphorylation of synaptic proteins. J. Neurochem. 73(2):727-35

230

DAVIS ET AL

oD Radi R. 1996. Reactions of nitric oxide with metalloproteins. Chem. Res. Toxicol. 9(5):828-35 eh. Kharitonov VG, Sharma VS, Magde D, Koesling D. 1997. Kinetics ofnitric oxide dissociation from five- and six-coordinate nitrosyl hemes and heme proteins, including soluble guanylate cyclase. Biochemistry 36(22):6814-18

94. Stone JR, Marletta MA. 1994. Soluble guanylate cyclase from bovine lung: activation with nitric oxide and carbon monoxide and spectral characterization of the ferrous and ferric states. Biochemistry

33(18):5636—40 O55 Hoenicka M, Becker EM, Apeler H, Sirichoke T, Schroder H, et al. 1999. Purified soluble guanylyl cyclase expressed in a baculovirus/Sf9 system: stimulation by

spectroscopy. Biochemistry 36(33): 10155—60 101. Ishii K, Warner TD, Sheng H, Murad F. 1991. Endothelin increases cyclic GMP levels in LLC-PK1 porcine kidney epithelial cells via formation of an endotheliumderived relaxing factor-like substance. J. Pharmacol. Exp. Ther. 259(3):1102-8

102. Schweizer M, Richter C. 1994. Nitric oxide potently and reversibly deenergizes mitochondria at low oxygen tension. Biochem. Biophys. Res. Commun. 204(1):169-75 103. Cleeter MW, Cooper JM, Darley-Usmar VM, Moncada S, Schapira AH. 1994. Reversible inhibition of cytochrome c oxidase, the terminal enzyme of the mitochondrial respiratory chain, by nitric oxide. Implications for neurode-

YC-1, nitric oxide, and carbon monoxide.

J. Mol. Med. 77(1):14-23 96. Kharitonov VG, Russwurm M, Magde D, Sharma VS, Koesling D. 1997. Dissocia-

tion of nitric oxide from soluble guanylate cyclase. Biochem. Biophys. Res. Commun. 239(1):284—-86

ie Stone JR, Marletta MA. 1996. Spectral and kinetic studies on the activation of soluble guanylate cyclase by nitric oxide. Biochemistry 35(4):1093-99

98. Cooper CE. 1999. Nitric oxide and iron proteins. Biochim. Biophys. Acta 411(2-3):290-309 ON), Burstyn JN, Yu AE, Dierks EA, Hawkins BK, Dawson JH. 1995. Studies of the heme coordination and ligand binding properties of soluble guanylyl cyclase (sGC): characterization of Fe(II)sGC and Fe(ID)sGC(CO) by electronic absorption

and magnetic circular dichroism spectroscopies and failure of CO to activate the enzyme. Biochemistry 34(17): 5896-903 100. Tomita T, Ogura T, Tsuyama S, Imai Y, Kitagawa T. 1997. Effects of GTP on bound nitric oxide of soluble guanylate cyclase probed by resonance Raman

generative

diseases.

FEBS

Lett. 345(1):

50-54 104. Brown GC, Cooper CE. 1994. Nanomolar concentrations of nitric oxide reversibly inhibit synaptosomal respiration by competing with oxygen at cytochrome oxidase. FEBS Lett. 356(2-3):295—98 5.

Blackmore

RS,

Greenwood

C. Gibson

QH. 1991. Studies of the primary oxygen intermediate in the reaction of fully reduced cytochrome oxidase. J. Biol. Chem. 266(29):19245—49 106. Giuftre A, Sarti P, D’Itri E, Buse G, Soulimane T, Brunori M. 1996. On the mechanism of inhibition of cytochrome c¢ oxidase by nitric oxide. J. Biol. Chem. 271(52):33404-8 107. Torres J, Darley-Usmar V, Wilson MT. 1995. Inhibition of cytochrome c oxidase in turnover by nitric oxide: mechanism and implications for control of respiration. Biochem. J. 312(Pt 1):169-73 108. Torres J, Cooper CE, Sharpe M, Wilson MT. 1998. Reactivity of nitric oxide with cytochrome c oxidase: interactions with the binuclear centre and mechanism of inhibition. J. Bioenerg. Biomembr. 30(1):63—69

~

231

NON-cGMP MEDIATED EFFECTS OF NO

109. Rousseau DL, Singh S, Ching YC, Sassaroli M. 1988. Nitrosyl cytochrome c oxidase. Formation and properties of mixed valence enzyme. J. Biol. Chem. 263(12):5681-85

110. Forfia PR, Hintze TH, Wolin MS, Kaley G. 1999. Role of nitric oxide in the control of mitochondrial function. Adv. Exp. Med. Biol. 471:381-88 EM Hoshino M, Ozawa K, Seki H, Ford PC. 1993. Photochemistry of the nitric

2s

Ute

ode adduct of water soluble iron(II) porphrin and ferrihemoproteins studied by nanosecond laser photolysis. J. Am. Chem. Soc. 115:9568-75 Brown GC. 1995. Reversible binding and inhibition of catalase by nitric oxide. Eur. J. Biochem. 232(1):188-91 Wink DA, Cook JA, Pacelli R, DeGraff W,

Gamson J, et ai. 1996. The effect of var-

ious nitric oxide-donor agents on hydrogen peroxide-mediated toxicity: a direct correlation between nitric oxide formation and protection. Arch. Biochem. Biophys. 331(2):241-48

114. Kim YM, Bergonia HA, Muller C, Pitt BR, Watkins WD, Lancaster JR Jr. 1995. Nitric oxide and intracellular heme. Adv. Pharmacol. 34:277-91 ie. Granger DL, Taintor RR, Cook JL, Hibbs JB Jr. 1980. Injury of neoplastic cells by murine macrophages leads to inhibition of mitochondrial respiration. J. Clin. Invest. 65(2):357-70 116. Granger DL, Lehninger AL. 1982. Sites of inhibition of mitochondrial electron transport in macrophage-injured neoplastic cells. J. Cell Biol. 95(2 Pt 1):527-35 SIW/e Drapier JC, Hibbs JB Jr. 1988. Differentiation of murine macrophages to express nonspecific cytotoxicity for tumor cells results in L-arginine-dependent inhibition of mitochondrial iron-sulfur enzymes in the macrophage effector cells. J. Immunol. 140(8):2829-38 118. Drapier JC, Hibbs JB Jr. 1986. Murine cytotoxic activated macrophages inhibit

aconitase in tumor cells. Inhibition involves the iron-sulfur prosthetic group and is reversible. J. Clin. Invest. 78(3): 790-97

TG). Gruer MJ, Artymiuk PJ, Guest JR. 1997. The aconitase family: three structural variations on a common Biochem. Sci. 22(1):3-6

theme.

Trends

. Pellat C, Henry Y, Drapier JC. 1990. IFN-gamma-activated macrophages: detection by electron paramagnetic resonance of complexes between L-argininederived nitric oxide and non-heme iron proteins. Biochem. Biophys. Res. Commun. 166(1):119-25

121. Klausner RD, Rouault TA, Harford JB. 1993. Regulating the fate of mRNA: the control of cellular iron metabolism. Cell 72(1):19-28 Wa Theil EC. 2000. Targeting mRNA to regulate iron and oxygen metabolism. Biochem. Pharmacol. 59(1):87—93

IDB: Gray NK, Hentze MW. 1994. Iron regulatory protein prevents binding of the 43S translation pre-initiation complex to ferritin and eALAS mRNAs. Embo J. 13(16):3882-91 124. Binder R, Horowitz JA, Basilion JP, Koeller

DM,

Klausner

RD,

Harford

JB. 1994. Evidence that the pathway of transferrin receptor mRNA degradation involves an endonucleolytic cleavage within the 3’ UTR and does not involve poly(A) tail shortening. Embo J. 13(8):1969-80 a),

Haile

DJ, Rouault TA, Harford JB, Kennedy MC, Blondin GA, et al. 1992.

Cellular regulation of the iron-responsive element binding protein: disassembly of the cubane iron-sulfur cluster results in high-affinity RNA binding. Proc. Natl. Acad. Sci. USA 89(24):11735-39

126. Drapier JC, Hirling H, Wietzerbin J, Kaldy P, Kuhn LC. 1993. Biosynthesis of nitric oxide activates iron regulatory factor in macrophages. Embo J. 12(9):3643— 49

232

DAVIS ET AL

127 . Weiss G, Goossen B, Doppler W, Fuchs D, Pantopoulos K, et al. 1993. Transla-

tional regulation via iron-responsive elements by the nitric oxide/NO-synthase pathway. Embo J. 12(9):3651—57

128. Bouton C. 1999. Nitrosative and oxidative modulation of iron regulatory proteins. Cell. Mol. Life Sci. 55(89):1043-53 129% Nunoshiba T, deRojas-Walker T, Wishnok JS, Tannenbaum

SR, Demple

B.

1993. Activation by nitric oxide of an oxidative-stress response that defends Escherichia coli against activated macrophages. Proc. Natl. Acad. Sci. USA 90(21):9993-97 130. Ding H, Demple B. 2000. Direct nitric oxide signal transduction via nitrosylation of iron-sulfur centers in the SoxR transcription activator. Proc. Natl. Acad. Sci. USA 97(10):5146—50 Sie Hidalgo E, Demple B. 1994. An ironsulfur center essential for transcriptional activation by the redox-sensing SoxR protein. Embo J. 13(1):138-46

13). Gaudu P, Weiss B. 1996. SoxR, a [2Fe2S] transcription factor, is active only in its oxidized form. Proc. Natl. Acad. Sci. USA 93(19):10094—98 138}, Demple B. 1999. Genetic responses against nitric oxide toxicity. Braz. J. Med. Biol. Res. 32(11):1417-27 134. Lee M, Arosio P, Cozzi A, Chasteen ND. 1994. Identification of the EPRactive iron-nitrosyl complexes in mammalian ferritins. Biochemistry 33(12): 3679-87 135: Kennedy MC, Gan T, Antholine WE, Petering DH. 1993. Metallothionein reacts with Fe2+ and NO to form products with Ag = 2.039 ESR signal. Biochem. Biophys. Res. Commun. 196(2):632-35 135a. Wink DA, Laval J. 1994. The Fpg potein, a DNA repair enzyme, is inhibited by the biomediator nitric oxide in vitro and

in vivo.

2125-29

Carcinogenesis

15(10):

136. Hoovers JM, Mannens M, John R, Bliek J, van Heyningen V, et al. 1992. Highresolution localization of 69 potential human zinc finger protein genes: a number are clustered. Genomics 12(2): 254-63 LSie Takatsuji H. 1999. Zinc-finger proteins: the classical zinc finger emerges in contemporary plant science. Plant Mol. Biol. 39(6):1073-78 138. Kroncke KD, Carlberg C. 2000. Inactivation of zinc finger transcription factors provides a mechanism for a gene regulatory role of nitric oxide. FASEB J. \4(1):166—73 139: Cho Y, Gorina S, Jeffrey PD, Pavletich NP. 1994. Crystal structure of a p53 tumor suppressor-DNA complex: understanding tumorigenic mutations [see comments]. Science 265(5170):346—S5

140. Ambs §, Merriam WG, Ogunfusika MO, Bennett WP, Ishibe N, et al. 1998. p53 and vascular endothelial growth factor regulate tumor growth of NOS2expressing human carcinoma cells. Nat. Med. 4(12):1371-76 141. Forrester K, Ambs S, Lupold SE, Kapust RB, Spillare EA, et al. 1996. Nitric oxideinduced p53 accumulation and regulation of inducible nitric oxide synthase expression by wild-type p53. Proc. Natl. Acad. Sci. USA 93(6):2442-47 142. Calmels S, Hainaut P, Ohshima H. 1997. Nitric oxide induces conformational and functional modifications of wild-type p53 tumor suppressor protein. Cancer Res. 57(16):3365-69 3. Pedersen JZ, Finazzi-Agro A. 1993. Protein-radical enzymes. FEBS Lett. 325(1-2):53-58 . Kauppi B, Nielsen BB, Ramaswamy S, Larsen IK, Thelander M, et al. 1996. The three-dimensional structure of mam-

malian ribonucleotide reductase protein R2 reveals a more-accessible iron-radical site than Escherichia coli R2. J. Mol. Biol. 262(5):706—20

233

NON-cGMP MEDIATED EFFECTS OF NO 145. Nordlund P, Sjoberg BM, Eklund H. 1990. Three-dimensional structure of the free radical protein of ribonucleotide

reductase.

Nature

345(6276):

593-98 146. Sahlin M, Petersson L, Graslund A, Ehrenberg A, Sjoberg BM, Thelander L. 1987. Magnetic interaction between the tyrosyl free radical and the antiferromagnetically coupled iron center in ribonucleotide reductase. Biochemistry 26(17):5541-48 147. Roy B, Lepoivre M, Henry Y, Fontecave M. 1995. Inhibition of ribonucleotide reductase by nitric oxide derived from thionitrites: reversible modifications of both subunits. Biochemistry

34(16):5411-18 148. Lepoivre Lemaire

M, G;

Chenais

B,.

Yapo

A,

Thelander®

1,

Tenu'

JP.

1990. Alterations of ribonucleotide reductase activity following induction of the nitrite-generating pathway in adenocarcinoma cells. J. Biol. Chem. 265(24): 1414349

149. Lepoivre M, Fieschi F, Coves J, Thelander L, Fontecave

M.

1991. Inactiva-

tion of ribonucleotide reductase by nitric oxide. Biochem. Biophys. Res. Commun. 179(1):442-48

150. Lepoivre M, Flaman JM, Bobe P, Lemaire G, Henry Y. 1994. Quenching of the tyrosyl free radical of ribonucleotide reductase by nitric oxide. Relationship to cytostasis induced in tumor cells by cytotoxic macrophages. J. Biol. Chem. 269(34):21891-97 lawl Guittet O, Ducastel B, Salem JS, Henry Y, Rubin H, et al. 1998. Differential sen-

sitivity of the tyrosy] radical of mouse ribonucleotide reductase to nitric oxide and peroxynitrite. J. Biol. Chem. 273(34): 2213644 152. Karthein R, Dietz R, Nastainczyk W, Ruf HH. 1988. Higher oxidation states

of prostaglandin H synthase. EPR study of a transient

tyrosyl

radical

in the

enzyme during the peroxidase reaction. Eur. J. Biochem. 171(1-2):313-20

IY3)3) Tsai A, Kulmacz RJ, Palmer G. 1995. Spectroscopic evidence for reaction of prostaglandin H synthase-1 tyrosyl radical with arachidonic acid. J. Biol. Chem. 270(18):10503-8 154. Tsai A, Palmer G, Xiao G, Swinney DC, Kulmacz

RJ.

1998.

Structural

characterization of arachidonyl radicals formed by prostaglandin H synthase-2 and prostaglandin H synthase-1 reconstituted with mangano protoporphyrin IX. J. Biol. Chem. 273(7):3888—94 155). Goodwin

DC,

Gunther

MR,

Hsi

LC,

Crews BC, Eling TE, et al. 1998. Nitric

oxide trapping of tyrosyl radicals generated during prostaglandin endoperoxide synthase turnover. Detection of the radical derivative of tyrosine 385. J. Biol. Chem. 273(15):8903-9 156. Maccarrone M, Putti S, Finazzi Agro A. 1997. Nitric oxide donors activate the cyclo-oxygenase and peroxidase activities of prostaglandin H synthase. FEBS Lett. 410(2-3):470-76 he Upmacis RK, Deeb RS, Hajjar DP. 1999. Regulation of prostaglandin H2 synthase activity by nitrogen oxides. Biochemistry 38(38):12505-13 158. Burney S, Caulfield JL, Niles JC, Wishnok

JS, Tannenbaum

SR.

1999.

The

chemistry of DNA damage from nitric oxide and peroxynitrite. Mutat. Res. 424(1-2):37-49 159: Halliwell B. 1999. Oxygen and _ nitrogen are pro-carcinogens. Damage to DNA by reactive oxygen, chlorine and nitrogen species: measurement, mechanism and the effects of nutrition. Mutat. Res. 443(1-2):37-52

160. Aust AE, Eveleigh JF. 1999. Mechanisms of DNA oxidation. Proc. Soc. Exp. Biol. Med. 222(3):246—-52 161. Beckman JS. 1996. The physiological and pathophysiological chemistry of nitric oxide.

In Nitric Oxide

Principles and

234

DAVIS ET AL

Actions, ed. JR Lancaster, pp. 1-82. New York: Academic

162. Marla SS, Lee J, Groves JT. 1997. Peroxynitrite rapidly permeates phospholipid

iis

membranes. Proc. Natl. Acad. Sci. USA 94(26):14243-48

mation. Free Radic. Res. Commun.

163. Kennedy

LJ, Moore K Jr, Caulfield JL, Tannenbaum SR, Dedon PC. 1997.

Quantitation of 8-oxoguanine and strand breaks produced by four oxidizing agents. Chem. Res. Toxicol. 10(4): 386-92 164. Uppu RM, Cueto R, Squadrito GL, Salgo MG, Pryor WA. 1996. Competi-

tive reactions of peroxynitrite with 2’deoxyguanosine and 7,8-dihydro-8-oxo2’-deoxyguanosine (8-oxodG): relevance to the formation of 8-oxodG in DNA exposed to peroxynitrite. Free Radic. Biol. Med. 21(3):407-11 165. Tretyakova NY, Niles JC, Burney S, Wishnok JS, Tannenbaum SR. 1999. Peroxynitrite-induced reactions of synthetic oligonucleotides containing 8oxoguanine.

Chem.

Res. Toxicol.

12(5):

459-66 166. Eiserich JP, Hristova M, Cross CE, Jones AD, Freeman BA, et al. 1998. Formation

of nitric oxide-derived inflammatory oxidants by myeloperoxidase in neutrophils. Nature 391(6665):393-97 167. Szabo C, Ohshima H. 1997. DNA damage induced by peroxynitrite: subsequent biological effects. Nitric Oxide 1(5):373— 85 168. Loeb LA, Preston BD. 1986, Mutagenesis by apurinic/apyrimidinic sites. Annu. Rev. Genet. 20:201—30 169. Domena JD, Timmer RT, Dicharry SA, Mosbaugh DW. 1988. Purification and properties of mitochondrial uracil-DNA glycosylase from rat liver. Biochemistry 27(18):6742-5 1 170. Dizdaroglu M, Rao G, Halliwell B, Gajewski E. 1991. Damage to the DNA bases in mammalian chromatin by hydrogen peroxide in the presence of ferric

and cupric ions. Arch. Biochem. Biophys. 285(2):317-24 Floyd RA, Watson JJ, Wong PK, Altmiller DH, Rickard RC. 1986. Hydroxyl free radical adduct of deoxyguanosine: sensitive detection and mechanisms of for1(3):

163-72 1990. The role of 8WZ: Floyd RA. hydroxyguanine in carcinogenesis. Carcinogenesis 11(9):1447-—50 17S: Fiala ES, Conaway CC, Mathis JE. 1989. Oxidative DNA and RNA damage in the livers of Sprague-Dawley rats treated with the hepatocarcinogen 2-nitropropane. Cancer Res. 49(20):5518—22 174. Umemura T, Sai K, Takagi A, Hasegawa R, Kurokawa Y. 1990. Formation of 8-hydroxydeoxyguanosine (8-OH-dG) in rat kidney DNA after intraperitoneal administration of ferric nitrilotriacetate

(Fe-NTA).

Carcinogenesis

11(2):

345-47 175: Jaiswal M, LaRusso NF, Burgart LJ, Gores GJ. 2000. Inflammatory cytokines induce DNA damage and inhibit DNA repair in cholangiocarcinoma cells by a nitric oxide-dependent mechanism. Cancer Res. 60(1):184-90

Ambs S, Merriam WG, Bennett WP, Felley-Bosco E, Ogunfusika MO, et al. 1998. Frequent nitric oxide synthase-2 expression in human colon adenomas: implication for tumor angiogenesis and colon cancer progression. Cancer Res. 58(2):334441 177. Yoshie Y, Ohshima H. 1998. Synergistic induction of DNA strand breakage by catechol-estrogen and nitric oxide: impli- . cations for hormonal carcinogenesis. Free Radic. Biol. Med. 24(2):341-48 178. Mannick EE, Bravo LE, Zarama G, Realpe JL, Zhang XJ, et al. 1996. Inducible nitric oxide synthase, nitrotyrosine, and apoptosis in Helicobacter pylori gasttitis: effect of antibiotics and antioxidants. Cancer Res. 56(14):3238-43

176.

NON-cGMP MEDIATED EFFECTS OF NO Wi . Souici AC, Mirkovitch J, Hausel P, Keefer LK, Felley-Bosco E. 2000. Transition mutation in codon 248 of the p53 tumor suppressor gene induced by reactive oxygen species and a nitric oxide-releasing compound. Carcinogenesis 21(2):281—-87

180. Wei YH. 1998. Oxidative stress and mitochondrial DNA mutations in human aging. Proc. Soc. Exp. Biol. Med. 217(1):53-63 181. Lee HK. 1999. Evidence that the mitochondrial genome is the thrifty genome. Diabetes Res. Clin. Pract. 45(2-3):127-—

35 182. Kobzik L Stringer B, Balligand JL, Reid MB, Stamler JS. 1995. Endothelial type nitric oxide synthase in skeletal muscle fibers: mitochondrial relationships. Biochem. Biophys. Res. Commun. 211(2):375-81

183. Bates TE, Loesch A, Burnstock G, Clark JB. 1996. Mitochondrial nitric oxide synthase: a ubiquitous regulator of oxidative phosphorylation? Biochem. Biophys.

189. O’ Donnell VB, Eiserich JP, Bloodsworth A, Chumley PH Kirk M, et al. 1999. Nitration of unsaturated fatty acids by nitric oxide-derived reactive species. Methods Enzymol 301:454-70 190. Rubbo H, Parthasarathy S, Barnes S, Kirk M, Kalyanaraman B, Freeman BA. 1995. Nitric oxide inhibition of lipoxygenasedependent liposome and low-density lipoprotein oxidation: termination of radical chain propagation reactions and formation of nitrogen-containing oxidized lipid derivatives. Arch. Biochem. Biophys. 324(1):15-25 LONE Gallon AA, Pryor WA. 1994. The reaction of low levels of nitrogen dioxide with methyl linoleate in the presence and absence of oxygen. Lipids 29(3):171—76 192. Denicola A, SouzaJM, RadiR. 1998. Diffusion of peroxynitrite across erythrocyte membranes. Proc. Natl. Acad. Sci. USA 95(7):3566-71 133 Liu X, Miller MJS, Joshi MS, Thomas DD, Lancaster JR Jr. 1998. Accelerated

reaction of nitric oxide with O2 within the hydrophobic interior of biological membranes. Proc. Natl. Acad. Sci. USA 95(5):2175-79

Res. Commun. 218(1):40-44

184. Giulivi C, Poderoso JJ, Boveris A. 1998. Production of nitric oxide by mitochondria. J. Biol. Chem. 273(18):11038-43 185. Tatoyan A, Giulivi C. 1998. Purification and characterization of a nitric-oxide synthase from rat liver mitochondria. J. Biol. Chem. 273(18):11044-48 186. Hogg N, Singh RJ, Kalyanaraman B. 1996. The role of glutathione in the transport and catabolism of nitric oxide. FEBS Lett. 382(3):223-28 187. Sharpe MA, Cooper CE. 1998. Reactions of nitric oxide with mitochondrial cytochrome c: a novel mechanism for the formation of nitroxyl anion and peroxynitrite. Biochem. J. 332(Pt 1): 9-19 188. Suter M, Richter C. 1999. Fragmented mitochondrial DNA is the predominant carrier of oxidized DNA bases. Biochemistry 38(1):459-64

235

194. Khan AU, Wilson T. 1995. Reactive oxygen species as cellular messengers. Chem. Biol. 2(7):437-45

193: Finkelstein A, Cass A. 1967. Effect of cholesterol on the water permeability of thin lipid membranes. Nature 216(5116):717-18 196. Vasquez-Vivar J, Denicola A, Radi R, Augusto O. 1997. Peroxynitrite-mediated decarboxylation of pyruvate to both carbon dioxide and carbon dioxide radical OTe

anion. Chem. Res. Toxicol. 10(7):786—-94 Podrez EA, Schmitt D, Hoff HF, Hazen

SL. 1999. Myeloperoxidase-generated reactive nitrogen species convert LDL into an atherogenic form in vitro. J. Clin. Invest. 103(11):1547-60

198. Hazen SL, Heinecke JW. 1997. Chlorotyrosine, a specific marker

3-

of

236

DAVIS ET AL

myeloperoxidase-catalyzed oxidation, is markedly elevated in low density lipoprotein isolated from human atherosclerotic intima. J. Clin. Invest. 99(9): 2075-8 1

199% Weiss SJ, Klein R, Slivka A, Wei M. 1982. Chlorination of taurine by human neutrophils. Evidence for hypochlorous acid generation.

J. Clin.

Invest.

70(3):598-

607 200. Ducrocq C, Blanchard B, Pignatelli B, Ohshima H. 1999. Peroxynitrite: an endogenous oxidizing and_nitrating agent. Cell. Mol. Life Sci. 55(8-9): 1068-77 201. Bartlett D, Church DF, Bounds PL, Koppenol WH. 1995. The kinetics of the oxidation of L-ascorbice acid by peroxynitrite. Free Radic. Biol. Med. 18(1): 85-92 202. Squadrito GL, Jin X, Pryor WA. 1995. Stopped-flow kinetic study of the reaction of ascorbic acid with peroxynitrite. Arch. Biochem. Biophys. 322(1):53-59 203. Skinner KA, White CR, Patel R, Tan S,

Barnes S, et al. 1998. Nitrosation of uric acid by peroxynitrite. formation of a vasoactive nitric oxide donor. J. Biol. Chem. 273(38):2449 1-97 204. Minetti M, Mallozzi C, Di Stasi AM, Pietraforte D. 1998. Bilirubin is an effective antioxidant of peroxynitrite-mediated protein oxidation in human blood plasma. Arch. Biochem. Biophys. 352(2): 165-74 205. Christen S, Woodall AA, Shigenaga MK, Southwell-Keely PT, Duncan MW, Ames

BN. 1997. gamma-tocopherol traps mutagenic electrophiles such as NO(X) and complements alpha-tocopherol: physiological implications. Proc. Natl. Acad. Sci. USA 94(7):3217—22 206. Daveu C, Servy C, Dendane M, Marin P, Ducrocg C. 1997. Oxidation and nitration of catecholamines by nitrogen oxides derived from nitric oxide [see comments]. Nitric Oxide 1(3):234-43

207. Haenen GR, Paquay JB, Korthouwer RE, Bast A. 1997. Peroxynitrite scavenging by flavonoids. Biochem. Biophys. Res. Commun. 236(3):591—93

208. Gilad E, Cuzzocrea S, Zingarelli B, Salzman AL, Szabo C. 1997. Melatonin is a scavenger of peroxynitrite. Life Sci. 60(10):L169—74 209. Mayer B, Pfeiffer S, Schrammel A, Koesling D, Schmidt K, Brunner F. 1998. A new pathway of nitric oxide/cyclic GMP signaling involving S-nitrosoglutathione. J. Biol. Chem. 273(6):3264-70

210. Kikugawa K, Hiramoto K, Tomiyama S, Asano Y. 1997. beta-carotene effectively scavenges toxic nitrogen oxides: nitrogen dioxide and peroxynitrous acid. FEBS Lett. 404(2-3):175-78 . Kondo H, Takahashi M, Niki E. 1997. Peroxynitrite-induced hemolysis of human erythrocytes and its inhibition by antioxidants. FEBS Lett. 413(2):236-38 to — bh

. Moro MA, Darley-Usmar VM, Lizasoain I, Su Y, Knowles RG, et al. 1995. The formation of nitric oxide donors from peroxynitrite. Br J. Pharmacol. 116(3):

1999-2004

Annu. Rey. Pharmacol. Toxicol. 2001. 41:237-60 Copyright © 2001 by Annual Reviews. All rights reserved

INTERACTIONS BETWEEN MONOAMINES,

GLUTAMATE, AND GABA IN SCHIZOPHRENIA: New Evidence Arvid Carlsson, Nicholas Waters, Susanna Holm-Waters,

Joakim Tedroff, Marie Nilsson, and Maria L. Carlsson Institute of Physiology and Pharmacology, University of Goteborg, Géteborg, Sweden; e-mail: arvid.carlsson@ pharm.gu.se, nicholas.waters @ pharm. gu.se, susanna.waters @ pharm. gu.se, joakim.tedroff@ swipnet.se, marie.nilsson@ pharm. gu.se, maria.carlsson @ pharm. gu.seé

Key Words

dopamine, noradrenaline, serotonin, movement disorders, stabilization

@ Abstract In spite of its proven heuristic value, the dopamine hypothesis of schizophrenia is now yielding to a multifactorial view, in which the other monoamines as well as glutamate and GABA are included, with a focus on neurotransmitter interactions in complex neurocircuits. The primary lesion(s) in schizophrenia does not necessarily involve any of these neurotransmitters directly but could deal with a more general defect, such as a faulty connectivity of developmental origin. Nevertheless, a precise identification of neurotransmitter aberrations in schizophrenia will probably provide clues for a better understanding of the disease and for the development of new treatment and prevention strategies.

INTRODUCTION The dopamine hypothesis of schizophrenia has guided schizophrenia research for several decades and has clearly proven its heuristic value. However, despite emerging direct evidence supporting a dopaminergic dysfunction in schizophrenia, this hypothesis is now yielding to a multifactorial view, in which the other monoamines as well as glutamate and GABA are joining up. Research in this area tends to focus on neurotransmitter interactions in complex neurocircuits. The primary lesion(s) in schizophrenia may not even involve any of these neurotransmitters directly but could deal, for example, with a more general defect in connectivity, perhaps of developmental origin (1). Nevertheless, the identification of neurotransmitter

aberrations in schizophrenia, now underway, is likely to bring clues for understanding the fundamental nature of the disease as well as for the development of new treatment and prevention strategies.

0362-1642/01/0421-0237$14.00

237

238

CARLSSON ET AL

This review starts out with the present state of dopamine research in schizophrenia. It then brings in the evidence for a role of glutamate in this disease and discusses the possibility of changes in one neurotransmitter being primary or secondary to those occurring in another. In this context serotonin and noradrenaline must also be considered. The review largely focuses on pharmacological evidence and discusses the two animal models that currently seem to be the most relevant, that is the hyperdopaminergia and the hypoglutamatergia models. Postmortem and imaging data are referred to only briefly, although much fascinating work is ongoing in this area. The integration of such data with the pharmacological evidence may well turn out to point to important directions for future research but will have to be left out of this review, because of space limitations.

DOPAMINE: Still a Cornerstone In recent years important progress has been made in basic schizophrenia research. The dopamine hypothesis of schizophrenia, which postulates a dopaminergic dysfunction in this disorder, has for a long time been supported only by indirect pharmacologic evidence (2), but has now received more direct support from two different lines of research, both using imaging techniques. First, it has been shown that the synthesis of labeled dopamine or fluorodopamine in the brain, measured by means of positron emission tomography following administration of radiolabeled L-3,4-dihydroxyphenylalanine (L-DOPA) or fluoro-L-DOPA, is increased in drug-naive schizophrenic patients, compared with age-matched controls (3-5). Second, single photon emission computed tomography and positron emission tomography studies, using a sophisticated technique to measure the release of dopamine in the basal ganglia in vivo, have shown that following an amphetamine challenge, this release is elevated in drug-naive schizophrenic patients compared with age-matched controls, and that this elevation correlates to the induction of positive psychotic symptoms (6-8). In an additional series of experiments Laruelle et al (9), using the single photon emission computed tomography technique with w-methyltyrosine as a tool, have obtained evidence that the unchallenged release of dopamine is elevated in schizophrenic patients compared with controls. Although these novel data are impressive, a number of caveats should be remembered. First, although statistically significant aberrations of dopamine synthesis and release have been demonstrated in schizophrenic patients, the data show a considerable scatter with some of the values observed within a normal range. Thus, a dysfunction of dopamine may be limited to a subpopulation of patients suffering from this probably heterogeneous disorder. (A few observations suggest that the aberration of dopamine synthesis may actually go in the opposite direction in catatonia, compared with other cases of schizophrenia.) Interestingly, the elevated dopamine release correlated with a good response to antipsychotic drugs.

NEUROTRANSMITTERS

AND SCHIZOPHRENIA

239

Secondly, it must be remembered that the abnormal values were obtained in patients who were exposed to the stress inevitably caused by the imaging procedures. Whether a dopamine dysfunction also occurs under minimal stress thus remains an Open question. Third, the patients examined were in acute episodes, and the situation may be different in chronic schizophrenic patients between episodes. In fact, recent observations by Laruelle et al (9) indicate that the amphetamine-induced release of

dopamine in schizophrenic patients in remission is within the normal range. This tallies with the clinical experience of patients in remission complaining about the side effects of antidopaminergic drugs more than they do during an exacerbation. If the level of dopaminergic function is normal in patients in remission, the practical implications are obvious. All agents used today to prevent relapse in schizophrenia are antidopaminergic and should thus induce hypodopaminergia in such patients. This is a most unpleasant and incapacitating condition that leads to extrapyramidal side effects and, perhaps more importantly, to a failure of the reward system, resulting in dysphoria and anhedonia and to a cognitive deficit. To develop drugs capable of preventing relapse without these side effects should be an urgent task. In fact, as is discussed below, such agents may already be underway. An additional point deals with the interpretation of the data obtained with labeled L-DOPA. An increased synthesis rate of labeled dopamine does not necessarily mean that the rate of endogenous dopamine synthesis is increased. The rate-limiting step in the synthesis of dopamine is generally assumed to be the hydroxylation of tyrosine rather than the decarboxylation of L-DOPA. A cautious interpretation of these interesting observations would thus be that there seems to exist in central dopaminergic neurons of schizophrenic patients a metabolic aberration involving the rate of L-DOPA decarboxylation. The functional significance of this aberration has not yet been fully clarified. However, these findings agree with the observations by Laruelle et al (9) regarding unchallenged dopamine release. Taken together, the data thus support the assumption of an elevated baseline release of dopamine in schizophrenic patients. Finally, this elevation of dopamine release in schizophrenia is not necessarily a primary phenomenon, but could be secondary to, for example, hypoglutamatergia. In support of this it has been found in rats as well as in humans that drugs that biock N-methyl-D-aspartic acid (NMDA) receptors are capable of enhancing the spontaneous and especially the amphetamine-induced release of dopamine. The latter effect could be due to blockade of a negative feedback mechanism (see below). On the whole, these recent data tend to support treatment strategies involving dopamine. However, most of the efforts made in recent years using such strategies have been discouraging. These efforts have largely focused on the D4 and partly also on the D3 and D1 subtypes. A general feeling seems to prevail that the D2 subtype is no longer an attractive target for an improved antipsychotic therapy.

240

CARLSSON ET AL

DOPAMINERGIC STABILIZERS: A Promising Novel Therapeutic Approach Is the therapeutic potential of dopaminergic agents exhausted? Several reasons support the view that this is far from the case. First, the roles of the various subtypes of dopamine receptors need to be explored further. However, perhaps even more important will be ongoing attempts to reach a deeper understanding of the function of these receptors. This may open up entirely new ways of improving this function and optimizing the receptors’ abilities to cope with aberrations in neural circuits. In support of this prediction some recent observations in our research group are briefly mentioned. We have developed a series of compounds capable of stabilizing the dopaminergic system without inducing the hypodopaminergia so ominous for the currently used antipsychotic drugs. Some of these new drugs are partial dopamine-receptor agonists, acting on the D2 family of receptors. A number of partial dopaminereceptor agonists, developed by us and by others, are now in clinical trials and seem to offer promise [for recent clinical data on (—)-3PPP, see Ref. 10; several of the partial dopamine receptor agonists studied so far are less suitable as probes because of their poor selectivity, e.g the ergolines terguride and SDZ 208-912]. Others are pure antagonists, again acting on the D2 family of receptors, and can thus readjust elevated dopamine functions, but in contrast to the currently used antipsychotic agents, they do not cause hypodopaminergia. On the contrary, they antagonize subnormal dopamine function. The reason for this aberrant pharmacological profile seems to be that their action on different subpopulations of dopamine receptors differs from that of the currently used drugs. Thus, whereas they exert a strong action on dopaminergic autoreceptors, they have a weaker effect postsynaptically and seem unable to reach a subpopulation of postsynaptic dopamine receptors (11-13). In subhuman primates, in which Parkinsonism had been induced by 1-methyl4-phenyl-1,2,3,6-tetrahydropyridine, one member of this class, named (S)-(—)3-(3-(methylsulfonyl)phenyl)- 1-propylpiperidine [(—)-OSU6162], given in single doses, could prevent L-DOPA-induced dyskinesias without interfering with the therapeutic movement response. In subsequent trials on Parkinsonian patients the same kind of response was observed (14; J Tedroff, personal communication). Subsequent trials on patients with Huntington’s disease (15) showed a marked reduction of choreatic movements, considerably outlasting the presence of the drug in the blood. These observations support the view that drugs of this class are capable of stabilizing dopaminergic function; that is, they are able to alleviate signs of hyperdopaminergia without inducing any signs of reduced dopaminergic function. If these findings can be extrapolated from neurology to psychiatry, these agents should possess antipsychotic activity without any concomitant signs of hypodopaminergia. Forthcoming trials with such agents in schizophrenia will answer this question. A preliminary study on a few schizophrenic patients, using

NEUROTRANSMITTERS

AND SCHIZOPHRENIA

241

a double-blind crossover design, has demonstrated an antipsychotic action of (S)-(—)-3-(3-(methylsulfonyl)phenyl)-1-propylpiperidine [(—)-OSU6162] (16).

BEYOND DOPAMINE In view of the close interaction between neurotransmitters in the brain, it is unlikely that dopamine is the only neurotransmitter showing dysfunction in schizophrenia. As already indicated, the change in dopaminergic function may even be secondary to aberrations elsewhere. In any event, there are good reasons to study the function of several other neurotransmitters in schizophrenia, such as noradrenaline, sero-

tonin, acetylcholine, glutamate, and GABA. These neurotransmitters are more difficult to study in the living intact brain than dopamine. Least difficult would perhaps be serotonin because it seems possible to study it using the same kind of approach as for dopamine; that is, to administer radiolabeled precursor (5-hydroxytryptophan) and measure the turnover of serotonin. Such a study has been carried out in depressed patients, and an aberration was actually demonstrated (17). For several years considerable interest has focussed on the possible role of glutamate in schizophrenia (18, 19). One reason for this is the discovery that phencyclidine (PCP, “angel dust’), which can induce a psychotic condition mimicking schizophrenia, perhaps even more faithfully than the amphetamines, is a powerful antagonist on one of the glutamate receptor subtypes, namely the NMDA receptor (20). This receptor is equipped with an ion channel that regulates the penetration of calcium and other cations into the neuron. PCP binds to a specific site in this channel, thereby blocking the function of the receptor. A number of other NMDA antagonists are available, binding to different sites of the receptor molecule, such as to the “PCP site,’ e.g. (+)-5-methyl-10,11-dihydro-5H-dibenzo-(a,d)cyclohepten-5,10-imine hydrogen maleate = dizocilpine (MK-801) and ketamine (an “uncompetitive” binding), or to the same site as glutamate, e.g. DL-2-amino5-phosphonopentanoic acid (APS), 3-(2-carboxypiperazine-4-yl)-1-propenyl-1phosphonic acid (D-CPPene), and cis-4-phosphonomethyl-2-piperidine carboxylic acid (CGS 19755) (a competitive binding), or to still another site on the NMDA receptor, where glycine functions as an additional agonist. An example of antagonists acting at the glycine site is D-cycloserine (a mixed agonist-antagonist). All these different NMDA antagonists appear to be psychostimulants, at least in rodents, and psychotogenic in humans.

GLUTAMATERGIC CONTROL OF MONOAMINE RELEASE At first the psychotogenic action of glutamate antagonists was suggested to be mediated by an increased catecholaminergic activity. Dopamine neurons, like other monoaminergic brainstem neurons, seem to be controlled by corticofugal

242

CARLSSON ET AL

glutamatergic neurons either directly or via gabaergic interneurons, acting as accelerators and brakes, respectively (Figure 1). Hypoglutamatergia may then cause an increase or a decrease in dopamine function, depending on whether the effect on the brake or the accelerator would predominate. Normally there appears to be a balance between the accelerator and the brake, perhaps with slightly more effect on the brake. Thus, a reduced glutamate function, induced e.g. by MK-801, may cause some elevation of dopamine release. However, if dopamine release is enhanced dramatically, e.g. by amphetamine, a negative feedback regulation appears to be activated, leading to a much greater effect on the brake. This can be demonstrated in experimental animals by superimposing an NMDA antagonist upon amphetamine. Then the release of dopamine is markedly enhanced (21). This phenomenon is of clinical interest because it opens up a possibility to explain the previously mentioned, enhanced amphetamine-induced release in schizophrenic patients. This enhancement could be due to a glutamate deficiency that leads to a weakened negative feedback control. In fact, cotreatment with the NMDA antagonist ketamine has been found to cause enhancement of the amphetamine-induced dopamine release in humans, as demonstrated by means of single photon emission computed tomography (9). Treatment of experimental animals with NMDA antagonists alone has given variable results. For example, in the previously cited work of Miller & Abercrombie (21) a slight, not dose-dependent release of dopamine, studied by microdialysis, was observed in rats, following treatment with MK-801. Using the same technique, other laboratories have found similar, more or less impressive effects of this agent. In fact, different portions of the dopaminergic system have been found to respond differentially to treatment with MK-801 (22). As to the competitive NNDA antagonists, the available evidence suggests that these agents, if anything, inhibit dopamine release, and this decrease is concomitant with behavioral stimulation (23). Thus, we have to look for a mechanism other than increased dopamine release

to account for at least an important part of the psychostimulant and psychotogenic action of NMDA antagonists. NMDA receptor antagonists appear to stimulate 5-HT turnover and release more consistently than dopaminergic activity (24). This is of special interest in view of the striking effect of the selective S-HT2A antagonist R(+)-a-(2,3dimethoxyphenyl)- 1-[2-(4-fluorophenylethy])]-4-piperidinemethanol (M100907) (25) on the behavioral stimulation induced by NMDA-receptor antagonism (26). This effect can be seen after doses of M100907 that are unable to influence the activity of normal mice. In fact, hyperserotonergia appears to be a prerequisite for this antagonism (26). This remarkable profile of M100907 is very different from that of neuroleptic agents and may have important therapeutic implications. The postmortem observations that suggest a presynaptic hyperserotonergia in paranoid

schizophrenic patients (27) are of interest in this context.

PCP, which is a somewhat less selective NUDA antagonist than MK-801, does

indeed cause a fairly pronounced release of dopamine, probably owing to a concomitant blockade of the dopamine transporter. However, the psychostimulation

NEUROTRANSMITTERS AND SCHIZOPHRENIA

243

Cortical Glutamate/GABA-Mediated Steering of Subcortical Systems

Cortex

SS mip dante

Glu

Glu

Brake

Accelerator

+

Figure 1 Hypothetical scheme showing the cortical regulation of the activity of the monoaminergic brainstem neurons by means of a direct glutamatergic pathway (accelerator) and an indirect glutamatergic/gabaergic pathway (brake). The outcome of a glutamatergic failure partly depends on the balance between the accelerator and the brake. If the latter predominates in the cortical regulation of a dopaminergic pathway, for example, such failure will lead to an elevated activity of this pathway. As indicated, feedback loops probably exist, at least partly mediated via the striatum and the thalamus. If, for example, the release of dopamine is enhanced by amphetamine, the feedback regulation will increase the activity of the brake, which will counteract the amphetamineinduced release. If the brake fails after treatment with an NMDA-receptor antagonist, or in the case of a hypothetic glutamatergic deficiency in schizophrenia, the amphetamine-induced release

of dopamine will be enhanced. A similar failure may occur in chronic abuse of central stimulants, leading to “sensitization” (cf. GABA: AN ACHILLES

HEEL). (From Ref. 45)

244

CARLSSON ET AL

caused by PCP does not depend so much on this release, because it can be nearly abolished by LY354740, a group-II metabotropic glutamate receptor agonist, despite the fact that this agonist leaves the enhanced dopamine release unchanged (28). Interestingly, in this study PCP was found to enhance the release of glutamate, and this effect was antagonized by LY354740. This phenomenon is commented on below. Many antipsychotic agents, especially some so-called atypical antipsychotics, possess adrenergic receptor-blocking properties, and the possible contributory role of this blockade has been recognized ever since our original report in 1963 (29).

Recent observations tend to corroborate this contention. Thus, in the hypoglutamatergia model of schizophrenia the w,-adrenergic antagonist prazosine could antagonize the behavioral stimulation and normalize the firing pattern of dopaminergic neurons in a manner somewhat similar to M100907 (22). The atypicality of clozapine and other more recently introduced antipsychotic agents may thus depend not only on antiserotonergic but also antiadrenergic properties. Other properties such as affinity for cholinergic receptors may have to be considered as well. Recently, Gessa (30) presented convincing evidence that the extracellular dopamine measured by means of microdialysis in rat prefrontal cortex is largely derived from noradrenergic neurons. In fact, we have observed for a long time in our laboratory that there exists a very close correlation between extracellular dopamine and noradrenaline in rat frontal cortex. Thus, variations in dopaminergic tone in this region should be looked at as an expression of locus ceruleus neuron activity and may then be regulated very differently from the dopaminergic neurons.

GLUTAMATE-MONOAMINE INTERACTIONS AT THE POSTSYNAPTIC (STRIATAL) LEVEL Carlsson & Carlsson (31) reported that MK-801, given systemically, is capable of inducing motor activity in mice completely depleted of dopamine and noradrenaline (by pretreatment with reserpine plus w-methyltyrosine). Subsequently, Svensson & Carlsson (32) showed that competitive NMDA-receptor antagonists were also active under these conditions, and that not only systemic but also local treatment with NMDA antagonists in the nucleus accumbens could induce movements in spite of virtually complete monoamine depletion. The local administration of NMDA receptor antagonists in the nucleus ac-— cumbens of monoamine-depleted mice induced a fairly normal motility pattern, but systemic treatment with these drugs caused a highly abnormal motility, i.e. compulsory forward locomotion with apparently total loss of the ability to switch between different behavioral patterns. Systemic treatment will inhibit NMDA receptors not only in the basal ganglia but also, for example, in the cerebral cortex, where the failure of glutamatergic association pathways could lead to loss of important functions, such as the ability to select appropriate behavioral programs. If glutamatergic deficiency is a relevant pathogenetic mechanism in schizophrenia

NEUROTRANSMITTERS AND SCHIZOPHRENIA

245

and if this includes the cerebral association pathways, it is not far-fetched to propose that this could lead to important consequences, involving cognitive disturbances, loss of flexibility, ambivalence and other behavioral aberrations, perhaps mainly

belonging to the sphere of negative schizophrenic symptomatology. Hypofrontality could also be a result of failure of cortical association pathways, and these could be especially vulnerable in so far as they engage chains of glutamatergic pathways. Thus, it may be speculated that glutamatergic failure in the cerebral cortex may lead to negative symptoms, whereas glutamatergic failure in the basal ganglia could be resposible for the positive symptoms. However, failure of the glutamatergic control of the so-called direct striatothalamic pathways may also contribute to the complex negative symptomatology (see below). Our subsequent work revealed a dramatic synergism between a variety of monoaminergic agonists and MK-801 or other NMDA receptor antagonists, all of them already in low dosage (33-35). This was true, for example, of apomorphine; a mixed D1/D2 agonist, SKF 38393; a selective D1 agonist, clonidine; an

a-adrenergic agonist; and a SHT2 agonist, LSD. A synergy between muscarinic and NMDA receptor antagonists was also demonstrated. Because these phenomena could be demonstrated in the absence of monoamines, the synergism must be

assumed to occur postsynaptically, and then presumably in the ventral! striatum. The exact mechanism of this synergism is not clear. It may occur locally or involve some kind of loop-mediated regulation. Based on these observations, we have proposed a hypothetical scheme that illustrates how the interaction between several neurotransmitters forms networks of psychotogenic pathways (Figure 2).

GABA: An Achilles Heel? Besides the glutamatergic neurons, the GABAergic neurons form by far the dominating neuronal cell population in the brain, and it is hard to imagine any neuronal circuitry that does not involve GABA. A reduction of GABAergic neurons has been observed in, for example, limbic and prefrontal cortical regions of schizophrenic brains post mortem, as well as an increase in the density of GABA, receptors (36). Moreover, GABAergic neurons have been found to be especially vulnerable to, for

example, glucocorticoid hormones, and also to glutamatergic excitotoxicity. Certain glutamatergic neurons seem to occur in increased number in, for example, the cingulate gyrus of schizophrenic brains, and this, in conjunction with a postulated role of stress in the pathogenesis of schizophrenia, would strengthen the assumption of an important role of GABA in schizophrenia, at least as a vulnerability factor. A GABAergic dysfunction might arise in the course of the disorder during adulthood or; for example, in utero or neonatally as a consequence of obstetric complications or other stressful events. In either case the activation of negative feedback circuits during psychosis or stress may well have a deleterious impact on vulnerable interneurons, resulting in longlasting and perhaps lifelong sensitivity changes.

246

CARLSSON ET AL

Psychotogenic Pathways vicious circle (left)

Muscimol (GABA A rec.)

PCP (NMDA rec.) '

1

Se)

LSD (5-HT2 rec.) 1

1

'' '

'

8

Ae

1

PCP (NMDA rec.) i

’

ae --

Atropine (M1 rec.)

PCP (NMDA rec.)

LSD (5-HT?2 rec.) - - -

PCP (NMDA rec.) (DA/NA

releaser)

Figure 2 Schematic diagram illustrating potential psychotogenic pathways and sites of action of psychotogenic and antipsychotic agents. The striatal complexes (STR, the centrally located circle) are composed of the dorsal and ventral striatum/pallidum. The striatum receives glutamatergic inputs from all parts of the cerebral cortex as well as serotonergic, dopaminergic, and noradrenergic inputs from the lower brainstem. Cholinergic interneurons, located in the striatum, seem to cooperate with glutamate to some extent. Striatopallidal chains of GABAergic neurons project to the thalamus, not shown in this Figure (but see Figure 3). Amphetamine and PCP are supposed to act psychotogenically by enhancing striatal dopamine release and blocking striatal NMDA receptors, respectively. These actions are partly located in the (limbic) striatum and partly in other sites. For example, PCP may act by blocking cortical as well as striatal NMDA receptors as well (e.g. in the hippocampus, as indicated in the figure) leading to reduced tone in corticostriatal glutamatergic pathways. The 5-HT2 agonist LSD may act by stimulating GABAergic interneurons in the limbic cortex, thereby reducing corticostriatal glutamatergic tone (46). LSD also seems to act on neurons in the striatum. (In contrast to this coupling in the limbic, piriform cortex, 5-HT2 receptors located presynaptically in prefrontal cortex on glutamatergic nerve terminals, seem to stimulate glutamate release, an effect that is counteracted by metabotropic glutamate autoreceptors (see Ref. 47)). The GABA A receptor agonist muscimol, which also appears to be psychotogenic (48), may likewise act by reducing corticostriatal glutamatergic tone. Anticholinergic agents appear to act by blocking predominantly muscarinic M1 receptors. (Modified from Ref. 2)

NEUROTRANSMITTERS

AND SCHIZOPHRENIA

247

THE THALAMIC (AND EXTRATHALAMIC) FILTER Carlsson (37) proposed that psychomotor activity and psychotogenesis depend, inter alia, on an interplay between dopamine and glutamate pathways projecting to the striatum from the lower brainstem and cortex, respectively (Figure 3). These neurotransmitters are predominantly, though not entirely, antagonistic to each other, the former being inhibitory and the latter stimulatory, when acting on striatal GABAergic projection neurons. These GABAergic projection neurons belong to so-called indirect striatothalamic pathways, which exert an inhibitory action on thalamocortical glutamatergic neurons, thereby filtering off part of the sensory input to the thalamus to protect the cortex from a sensory overload and hyperarousal. Hyperactivity of dopamine or hypofunction of the corticostriatal glutamate pathway should reduce this protective influence and could thus lead to confusion or psychosis.

Pallidum

eho

Behaviour

Figure 3 Neurocircuitries of the basal ganglia. Detail of the striatopallidothalamic pathways. Among these, the top and bottom pathways drawn with thick lines contain three GABAergic neurons and are referred to as “indirect,” i.e. inhibitory, pathways. The pathway in between contains two GABAergic neurons and is referred to as “direct” (modified from Ref. 33). SN, substantia

nigra; VTA, ventral tegmental area; STN, subthalamic nucleus; Glu, glutamate. Ach, acetylcholine; DA, dopamine.

248

CARLSSON ET AL

Carlsson’s hypothesis (37) focussed on the indirect striatothalamic pathways, which have an inhibitory influence on the thalamus. The corresponding direct pathways exert an opposite, excitatory influence. Both pathways are controlled by glutamatergic corticostriatal fibers, enabling the cortex to regulate the thalamic gating in opposite directions. In other words, they appear to serve as brakes and accelerators, respectively, in analogy to the regulation of monoaminergic brainstem neurons mentioned above. Normally, the inhibitory, indirect pathways seem to dominate over the direct pathways. Thus, NMDA receptor inhibitors are behavioral stimulants. However, the balance between the direct and indirect pathways may vary, depending on the state of the system. Failure of the direct pathway, induced e.g. by glutamatergic deficiency, might contribute to the so-called negative symptomatology of schizophrenia. It has been suggested that the activity of the direct pathways is predominantly phasic, whereas that of the indirect pathways is mainly tonic (38). This difference could have important consequences for a differential responsiveness of the direct and indirect pathways to drugs (39). Needless to say, the postulated existence of a thalamic filter would not exclude

a gating function located in other parts of the brain, e.g. the prefrontal cortex. The impressive sophistication of the gating function, enabling a focussed attention to relevance and novelty at the expense of trivial sensory inputs, would actually speak in favor of a more widely spread location. Little is known about the role of different receptor subtypes in the respective pathways. As to the glutamatergic receptors, NMDA receptor antagonists, as mentioned, are behavioral stimulants, at least in rodents, and this has been interpreted

as the result of a failure of the indirect, inhibitory pathways. AMPA receptor antagonists have been studied less intensely but have been found to act in the same direction as NMDA antagonists in some experiments, whereas they act as antagonists to NMDA antagonists in other experiments (22). As for the metabotropic receptors, the recent observations of Moghaddam & Adams (28), briefly referred to above, are most interesting. They found that the behavioral stimulation caused by PCP could be antagonized by a metabotropic receptor agonist, and at the same time, the PCP-induced elevation of glutamate release was antagonized. The question arises as to whether these data can be accommodated to the model of direct and indirect pathways outlined above. Glutamate release is, generally speaking, much more difficult to measure and interpret than the release of monoamines. For example, glutamate plays an important role in general cell metabolism, in addition to serving as a neurotransmitter. Perhaps glutamate release is predominantly indicative of the activity of the direct pathways, because they seem to be mainly phasic, and release by burst firing may be more likely to show up in microdialysis. Thus, the PCP-induced elevation of glutamate release, as measured by microdialysis, is perhaps indicative of an increased activity of the direct pathway; possibly the metabotropic receptor agonist antagonized this release by stimulating glutamatergic autoreceptors. Of course, such a mechanism is speculative. As mentioned above, the dopaminergic projections to the indirect striatothalamic pathways appear to be predominantly inhibitory at the cellular level. They

NEUROTRANSMITTERS AND SCHIZOPHRENIA

249

seem to operate largely via dopamine D2 receptors. The dopaminergic projections to the direct pathways, however, seem to be stimulating and mediated via D1 receptors.

COMPARING TWO EXPERIMENTAL SCHIZOPHRENIA MODELS: Therapeutic Implications As it looks today, the two most important animal models of psychosis are those induced by hyperfunction of dopamine and hypofunction of glutamate. However, one cannot disregard the other monoamines, among them perhaps especially serotonin. There seem to be strong interactions between glutamate and all the monoamines in psychotogenesis, and among them the link between glutamate and serotonin appears to be especially noteworthy. Presumably, several other neurotransmitters will ultimately have to be considered. To illustrate some remarkable transmitter interactions one can compare the actions of the typical antidopaminergic agent haloperidol and the specific 5-HT2A antagonist M100907 in two models of psychosis (Figure 4). As expected, haloperidol is quite powerful in alleviating the hyperdopaminergic stimulation induced by amphetamine, but it is less efficacious in the hypoglutamatergic behavioral stimulation induced by MK-801. However, M100907 is clearly more powerful in counteracting MK-801 than amphetamine-induced stimulation (40). These observations indicate that serotonin plays a more prominent role than dopamine in the behavioral stimulation induced by hypoglutamatergia. In behavior studies one must of course pay attention both to the quantitative and qualitative aspects. Using an ethovision system, we have found that MK-801 not only stimulates motility but at the same time changes the behavioral pattern in a direction suggestive of a retrogression to more primitive behavior, with an impoverishment of the behavioral repertoire (M Nilsson, S Waters, N Waters,

ML Carlsson, unpublished manuscript; Figure 5). One can speculate that this retrogression involves a cognitive deficit. Besides, it may be related to another behavioral abnormality induced by MK-801 and related drugs: the loss of habituation (42). Itis well known that habituation is also deficient in psychosis. When haloperidol counteracts the behavioral stimulation induced by hypoglutamatergia, it does so without any clear trend to normalization of the abnormal,

primitive movement pattern observed in this condition. In contrast, M100907 can concomitantly normalize the hypermotility and, albeit partially, the behavioral pattern, thus reinstating a richer behavioral repertoire (Figure 5). Given these results from animal model experiments, the question arises whether

one can distinguish between subpopulations of schizophrenic patients responding preferentially to either antidopaminergic or antiserotonergic agents; such subpopulations might then be assumed to be predominantly hyperdopaminergic and hyperserotonergic, respectively, and in the latter case perhaps also hypoglutamatergic.

250

CARLSSON ET AL

QA 25075

M100907

min 30-60

0.7 0.01 0.001

min 30-60

+

}Tr

Oso S

Figure 4 (a) Effect of various doses of M100907 (0.001, 0.01, and 0.1 mg/kg) on locomotion stimulated by d-amphetamine (Amph, 3 mg/kg) or MK-801 (0.3 mg/kg). All drugs were given immediately before the animals were placed in the activity meters. Statistics: Mann-Whitney U-test. ++-+p < 0.001 vs vehicle; *p < 0.05, **p < 0.01, ***p < 0.001 vs d-amphetamine or MK-801. RS, Spearman correlation coefficient. (b) Effect of various doses of haloperidol (0.001, 0.01, 0.1, and 1 mg/kg) on locomotion stimulated by d-amphetamine (Amph, 3 mg/kg) or MK-801 (0.3 mg/kg). All drugs were given immediately before the animals were placed in the activity meters. Statistics: Mann-Whitney U-test. +p < 0.05, ++p < 0.01, +++p < 0.001 vs vehicle; *p < 0.05, **p < 0.01, ***p < 0.001 vs d-amphetamine or MK-801. RS, Spearman correlation coefficient (from Ref. 45).

NEUROTRANSMITTERS

a) control

AND SCHIZOPHRENIA

b) MK-801

251

Cc) MK+haloperidol

W CIstg4

ASae as ees V
Asn and Met’4—> Thr (32). These mutations may represent cloning artifacts. A novel isoform of Trx in MP6 T lymphocyte cells has been reported with Lys— Arg replacements determined by protein sequencing at positions 3, 8, 21, and 96 (33). However, other researchers have been unable to detect the corresponding Arg-specific codons in Trx mRNA from the same cell line (34). Our studies with over 50 human cancer cell lines have failed to find

evidence for expression of mutant forms of thioredoxin-1 (11, 35). A second, slightly larger Trx based on electrophoretic mobility was identified in pig heart mitochondria (36). This second Trx, thioredoxin-2, was cloned from a rat heart library and found to encode a 166 amino acid, 18 kDa protein that had a

264

POWIS

® MONTFORT

conserved Trx catalytic site but that lacked the other Cys residues found in mammalian thioredoxin-1 (37). A 60-amino-acid N-terminal extension of thioredoxin-

2 exhibited characteristics consistent with a mitochondrial translocation signal, and Western blotting has confirmed the mitochondrial localization of thioredoxin-2. A putative cleavage site gives a 12.2 kDa protein, and recombinant truncated thioredoxin-2 that contains the catalytic site has been shown to reduce insulin and to be reduced by thioredoxin reductase-1 and NADPH (37). Human thioredoxin-2 has also been cloned (38).

A 32-kDa thioredoxin-like cytosolic protein (p32'™') has been cloned from a human testis cDNA library (39). The predicted protein sequence is 289 amino acids with an N-terminal Trx domain of 105 amino acids, a conserved Trx active site (-Cys-Gly-Pro-Cys-), and a high degree of homology to human thioredoxin-1 (39) (Figure 1). The sequence of the remaining 184 C-terminal amino acids showed no homology to other proteins in the database. p32!" is ubiquitously expressed in human tissues, with the highest levels appearing in stomach, testis, and bone

marrow. Neither the full length p32! protein nor the N-terminal 105 or 107 amino acid Trx-like fragments were reduced by thioredoxin reductase and NADPH (39, 40). However, when reduced by dithioethreitol, the N-terminal 107 amino acid fragment of p32

was able to reduce insulin (40).

A 435-amino-acid redox protein with similarity to Trx but with a -Trp-Cys-ProPro-Cys- catalytic site (instead of -Trp-Cys-Gly-Pro-Cys-) has been cloned from a mouse YAC library and localized to the nucleus (41). The protein has been called nucleoredoxin and localized to mouse chromosome

11 (41).

Processed Forms of Trx A shorter, 10-kDa, form of thioredoxin-1 has been reported to be secreted and bound to the outer plasma membrane of human U937 cells (42) and MP6 cells (33). The shorter form has the same N-terminal amino acid sequence as thioredoxin-1 and is immunoreactive with human anti-thioredoxin-1 polyclonal antibody but has

enhanced eosinophil cytotoxic activity (42). Although they still have the conserved active site, truncated mutant forms of human thioredoxin-1 lacking the C-terminal 16 or 24 amino acids similarly showed increased eosinophil cytotoxic activity and, surprisingly, are without insulin disulfide reductase activity (43). Alternatively spliced forms of thioredoxin-1 mRNA have been isolated, but with in-frame stop codons (34). It appears, therefore, that the truncated forms of Trx are formed byproteolytic cleavage.

STRUCTURE OF TRX Extensive structural data exist for human thioredoxin-1. The solution structures of

oxidized and reduced mutant (Cys°*—> Ala, Cys®—> Ala, Cys”3—> Ala) thioredoxin-

1 (44) and crystal structures of oxidized and reduced wild-type thioredoxin-1, as well as solution and crystal structures of various mutant forms of thioredoxi n-1

THIOREDOXIN

Ws aia

265

BAT

Figure 2 Structure of the thioredoxin-1 dimer showing the position of the Cys** and Cys*> catalytic site residues and the cross-linked Cys’? residues.

(44, 45), have been reported. These studies and earlier structure studies with E.

coli Trx (46-48) show that Trx is a compact globular protein with a five-stranded beta sheet forming a hydrophobic core surrounded by four alpha helices on the external surface. The conserved active site amino acids, -Trp-Cys-Gly-Pro-Cys-,

link the second beta strand to the second alpha helix and form the first turn of the second helix. This stable tertiary structure is known as the Trx fold (49). The

mechanism for the reducing action of Trx is that substrate X-S, binds to a conserved hydrophobic surface and in the hydrophobic environment of the complex, the thiolate of Cys*? acting as a nucleophile, combines with the protein substrate to form a covalently linked mixed disulfide (-Cys**-S-S- protein). Finally, the now

deprotonated Cys*> attacks the -Cys*?-S-S-protein disulfide bond, releasing the reduced protein substrate and forming Trx-Cys**-Cys*?-disulfide, which is then reduced by thioredoxin reductase (50). X-ray crystallography has shown small redox-dependent conformational change in the active site and an unusual thiolthiol hydrogen bond that may provide an explanation for the putative depressed

pKa of the active site Cys*” (45). However, the true pKa of Cys*” remains in dispute On):

Homo- and Heterodimer Formation by Trx Human thioredoxin- 1 forms covalently linked homodimers in solution (45, 52) es-

pecially in the presence of a strong oxidant or when stored at high concentrations

266

POWIS

® MONTFORT

Trx(ox)

thioredoxin

|

growth factor

reductase Trx(red)

peroxidase a

ee

ae

transcription factor ||protein binding

ribonucleotide reductase

ee ribonucleotides a) riponucieo

cell proliferation

DNA binding

Trx (homodimer)

AP-1

ASKI1

NE-xB

GR

PEBP2/CBF GR ER

vit D; upreg protein p40phox lipocalin

p53 es

Ref-1

PKCo.5,€,2

AP2 |

siaes

anti-apoptosis

Figure 3. Actions of Trx in the cell.

(26). The crystal structure of human thioredoxin-1 revealed a dimeric structure

in which monomers are disulfide-linked through Cys’* from each subunit, and active site residues are reduced and buried in the dimer interface (45) (Figure 3). The crystal structure of the Cys’*— Ser mutant revealed a dimeric form of the protein nearly identical to that of the wild-type protein, only without the disulfide linkage, suggesting the dimer may represent a naturally occurring form of the protein. The dimer contains an 1100 A? largely hydrophobic interface with five hydrogen bonds in addition to the disulfide bond. In the dimer interface are 12 amino acids from each monomer, 10 of which are invariant in higher animals, including Cys’? (45,49). The apparent dissociation constant for non-covalent dimer forma-

tion under reducing conditions varied between 6 and 166 4M at pHs of 3.8 to 8.0, respectively, using a chemical modification assay and gel filtration (53). A study using analytical ultracentrifugation and NMR spectroscopy failed to find evidence of dimer formation in solution (54). However those experiments were

done under conditions where active site residues Cys** and Cys*> were oxidized (disulfide linked), which interferes with dimer formation (DAR Sanders & WR

Montfort, unpublished data), and dimer formation of the reduced protein can be detected by ultracentrifugation (RR Sotelo-Mundo & WR Montfort, unpublished

data). It is not yet clear if dimer formation occurs under physiological conditions, since the association is relatively weak, nor is it clear what role thioredoxin-1 dimers might play, since the active site becomes buried on dimer formation, is

THIOREDOXIN

267

not a substrate for reduction by thioredoxin reductase, and does not stimulate cell

growth (26, 52). The local concentration of human thioredoxin-1 in tissues has not yet been accurately determined but overall is about 2 to 12 uM in bovine tissues (55) and possibly higher in tumor tissues that over express thioredoxin-1, suggesting the protein may reach concentrations where dimer formation of the reduced protein is appreciable. Human thioredoxin-1 also forms heterodimers with numerous protein redox partners, as is described in more detail later. Two such interactions have been studied by NMR (56, 57) witha Cys —> Ala mutant form of human thioredoxin-1 and

peptides from p50 NFKB and Ref-1. Both p50 NFKB and Ref-1 have redox sensitive Cys residues that must be in the reduced state for the proteins to participate in transcriptional regulation. Stable complexes were achieved through disulfide

bond formation between the thioredoxin-1 Cys** and Cys™ of p50 NFkB or Cys® of Ref-1. Both peptides formed specific complexes with human thioredoxin-1 through a boot-shaped cleft on the thioredoxin-1 surface that coincides with part of the homodimer interface. In addition to the disulfide bond, both complexes were found to be stabilized by numerous hydrogen-bonding, electrostatic and hydrophobic interactions, however, the orientation of the peptides in the binding cleft differs between the two complexes in that they run in opposite directions (NC vs C-—N). The numerous interactions between the peptides and thioredoxin-1 suggest there is specificity to the interactions, but dissociation constants for the full heterodimers have not been reported.

REGULATION OF TRX EXPRESSION The coding region of the human thioredoxin-1 gene spans 1.3 kb and is organized into 5 exons (28). The promoter region contains many possible regulatory binding motifs compatible with constitutive expression, including GCF, SP1, and WT-ZFP (28); with inducible expression, including AP-1, AP-2, NF-«B, Oct-1, PEA-3, and Myb (28); and with an oxidative stress response element (58, 59). A variety of

stress stimuli increase thioredoxin-1 expression in cells including hypoxia (60, 61), Staph. aureus protein (62), lipopolysaccharide (63), O (64,65), H,0, (66-68), phorbol ester, viral infection, photochemical oxidative stress (69), X-radiation, and UV irradiation (67, 70-75). Thioredoxin-1 is transcriptionally induced by activation of heat shock factor 2 during hemin-induced differentiation of K562 erythroleukemia cells (76) and by retinol (vitamin A) in monkey tracheobronchial epithelial cells (77). Estradiol increases thioredoxin-1 expression in primary cultures of human endometrial stromal cells (78), whereas estradiol, although not progesterone and testosterone, increases the expression of thioredoxin-1 in bovine artery endothelial cells (79). Estradiol and testosterone increase thioredoxin-1 ex-

pression in uterus of ovarectomized rats, whereas there is no effect on the levels of hepatic thioredoxin-1 (80). The expression of thioredoxin-1 also shows cell cycle dependency (60, 81).

268

POWIS

® MONTFORT

SUBCELLULAR LOCALIZATION Thioredoxin-1 is predominantly a cytosolic protein, and although it has no known nuclear localization sequence, it has been detected in the nucleus of normal endometrial stromal cells (78), tumor cells (82-84), and primary solid tumors (73, 82, 84, 85). Treatment of cells with H,O, (86), PMA (83), UV irradiation (59), hypoxia (61, 87), and the cancer drug cisplatin (88) has been reported to cause the translocation of thioredoxin-1 from the cytoplasm to the nucleus. The mechanism for this translocation is not known, but it may be the consequence of thioredoxin-1 being carried along bound to another protein with a nuclear import sequence (89). Thioredoxin-1 is also found associated with plasma and cell membranes, presumably on the outside surface of cells (90,91). As previously noted thioredoxin-2 is found in the mitochondria (37) while p32!" is a cytosolic protein (39).

BIOLOGICAL ACTIVITIES Trx has many biological activities. discussed below.

These are summarized in Figure 3 and are

Growth Factor Thioredoxin-1 acts as a growth factor, and it is produced by a variety of cells including EBV-transformed B cells (7,92), MP6 T-cell hybridoma (93), and hepatoma cells (25,72). Thioredoxin-1 is secreted by lymphocytes, hepatocytes, fibroblasts, and a variety of cancer cells (60, 62, 94-96). The mechanism by which secretion occurs is not known, but since thioredoxin-1 has no leader sequence, the mechanism appears to involve a pathway independent of the endoplasmic-Golgi secretory pathway (94,97). Thioredoxin-1 stimulates the growth of lymphocytes (92), normal fibroblasts (98), and a variety of tumor cell lines (25,72). Neither

the catalytic site mutant Cys**—> Ser/Cys*>-> Ser thioredoxin-1 nor E. coli Trx at

concentrations 50-fold higher than human thioredoxin-1 are able to stimulate cell growth (98). The mechanism responsible for growth stimulation by thioredoxin-1 is atypical for a growth factor. No evidence has been found of saturable binding of '°°I-labeled human thioredoxin-1 to the surface of MCE-7 breast cancer cells that would indicate the presence of a thioredoxin-1 receptor, and thioredoxin-1 appears to sensitize cells to growth factors produced by the cell itself (25). Not all growth factors are enhanced by thioredoxin-1. For example, in MCF-7 breast cancer cells, the effects of insulin and epidermal growth factor are not potenti-

ated (25), while IL-2 is enhanced up to a 1000-fold and basic fibroblast growth factor (bFGF) up to a 100-fold (15). The mechanism that stimulates the effects

of IL-2 may include an increase in the alpha subunits of the IL-2 receptor (8, 99). Thioredoxin-1 also increases the cell expression of a variety of cytokines, includ-

ing IL-1, IL-2, IL-6, IL-8, and tumor necrosis factor-a (TNF-a) (100), as well as

of the TNF-a induced expression of IL-6 and IL-8 by human rheumatoid arthritis

THIOREDOXIN

269

fibroblasts (101). However, thioredoxin-1 transfection of murine L929 cells down-

regulates the expression of IL-6, which is associated with inhibition of p38 MAP kinase activity in the cells (102). Thioredoxin-1 prevents the inhibition of cdc25 phosphatase activity that leads to impairment of p34°°* dephosphorylation during the second cell cycle of mouse embryonic development in vitro (103).

Antioxidant The direct antioxidant properties of thioredoxin-1 include removal of hydrogen peroxide (104). Thioredoxin-1

has also been reported to be an efficient electron

donor to human plasma glutathione peroxidase (105). Thioredoxin-1 is present in plasma at concentrations up to 6 nM (see below) and has been suggested to have an antioxidant role in plasma (3). However, plasma contains considerably higher levels of reduced glutathione, around | 4M (106, 107), and reduced plasma proteins provide additional thiol buffering capacity (108). It seems unlikely that the relatively low concentration of thioredoxin-1 in plasma would contribute much antioxidant activity. Thioredoxin-1 appears to exert most of its antioxidant properties in cells through thioredoxin peroxidase. The thioredoxin peroxidases belong to a conserved family of antioxidant proteins, the peroxiredoxins (PRDXs), which use thyl groups as reducing equivalents to scavenge oxidants (109). The reduced form of thioredoxin peroxidase scavenges oxidant species such as H,O, and alky peroxides and, in the process, homo or heterodimerize with other family members through disulfide bonds formed between conserved Cys residues (109, 110). Trx reduces the

oxidized thioredoxin peroxidase to the monomeric form. Several human thioredoxin peroxidases have been cloned: thioredoxin peroxidase-1 (22 kDa monomer), also known as natural killer enhancing factor-B (111); thioredoxin peroxidase-2 (22-kDa monomer), also known as natural killer enhancing factor-A (112), pro-

liferation associated gene (113), or heme-binding protein 23 (114); thioredoxin peroxidase-3 (21.6 kDa monomer), also known as human murine erythroleukemiaassociated 5, AOP1, or SP-22, which is primarily mitochondrial (115, 116); and thioredoxin peroxidase-4 (31 kDa monomer), or human AOE372, identified as a

cytosolic and secreted protein (110,117). A factor purified from bovine adrenal cortex, identified as mitochondrial thioredoxin-2, is necessary for the peroxidase activity of thioredoxin peroxidase-4 (116). Thioredoxin peroxidase-4 has also been identified as TRANK (thioredoxin peroxidase-related activator of NF-«B and c-Jun N-terminal kinase), which when added to U-937 human myeloid cells in-

creases the binding of the trancription factor NF-« Bto DNA and increases NF-« B transactivation (118). However, transfection of cells with thioredoxin peroxidase-2 (110, 115) or thioredoxin peroxidase-4 (110) inhibits NF-« B activation measured

by gel shift mobility assays, reporter constructs, and nuclear immunohostochemical localization. The inhibition of NF-«B activation by thioredoxin peroxidase-4 may be caused by decreasing the phosphorylation of the cytosolic NF-«B inhibitor, IxB, which leads to its decreased degradation (110). A fifth member of the PRDX family, PRDX-5, contains only one of the essential Cys residues and reduces H,0,

270

POWIS

® MONTFORT

in the presence of dithioethreitol, but not in the presence of thioredoxin-1 (119). PRDX-6 has been cloned from a human liver library (120) but appears to be a nonselenium containing glutathione peroxidase (121, 122). Thioredoxin peroxidases are found in human red blood cells as antioxidants protecting red blood cells from oxidant injury (123, 124). Transfection of human endothelial cells with thioredoxin peroxidase-| protects the cells from H,O,-induced cytotoxicity and inflammation-induced monocyte adhesion (124). Thioredoxin peroxidase-2 is found abundantly expressed in rapidly growing and transformed cells but is present at low levels in quiescent cells (113,125). Transfection of Molt-4 leukemia cells with thioredoxin peroxidase-2 protects the cells from apoptosis induced by serum deprivation, ceramide, and etoposide and inhibits the release of cytochrome c from mitochondria during apoptosis (124, 126). Thioredoxin peroxidase-2 binds to the SH3-binding domain of c-Abl tyrosine kinase and inhibits kinase activity (127). Thioredoxin peroxidase-2 protein is increased as cells enter S-phase and by agents that induce oxidative stress (128). The crystal structure of rat thioredoxin peroxidase-2 has been reported (114). Another mechanism for removing HO, in the cell is through selenocysteinecontaining glutathione peroxidases that use reduced glutathione as a source of reducing equivalents (129). As far as is known the glutathione and Trx redox systems are not coupled in the cell. Selenium has different effects on the Trx and glutathione peroxidase systems, increasing the activity of thioredoxin reductase (130) but not of thioredoxin peroxidase, and increasing the activity of glutathione peroxidase but not glutathione reductase (129; Figure 4). Thus, in the presence of an excess of peroxide, an increase in available selenium would predictably increase the levels of reduced Trx relative to reduced glutathione.

Cofactor One of the earliest functions ascribed to bacterial Trx was as a source of reducing equivalents for ribonucleotide reductase (1), which catalyzes the conversion of nucleotides to deoxynucleotides, the first unique step of DNA synthesis and an essential step for cellular proliferation (131). The importance of thioredoxin-1 for eukaryotic ribonucleotide reductase is less understood, and there may be other sources of reducing equivalents (132). However, irreversible inhibition of thioredoxin reductase by some antitumor quinones has been associated with a decrease in cellular ribonucleotide reductase activity (133). Thioredoxin-1

has also been

suggested to be a source of reducing equivalents for vitamin K epoxide reductase, which is necessary for the biosynthesis of plasma clotting factors (134, 135), although other studies have shown that inhibitors of the Trx system and antibodies against thioredoxin-1 do not inhibit vitamin K epoxide reduction (136).

Transcription Factor Regulation Thioredoxin-1 selectively activates the DNA-binding of a number of transcription factors. This includes the transcription factor NF-«B that is important for

THIOREDOXIN

271

H,0,

ie (ree)

GSH

Tx.

Trx(0x) Figure4

H,O

GSSG

Effects of selenium on thioredoxin peroxidase (TrxP) and glutathione peroxidase cycles.

Trx is thioredoxin, reduced (red) and oxidized (ox), GSH

is reduced glutathione and GSSG

is

oxidized glutathione. The enzymes are the selenoproteins (shown shaded) thioredoxin reductase (TrxR) and glutathione peroxidase, and the non-selenoproteines are thioredoxin peroxidase and glutathiore reductase.

the cell response to oxidative stress, apoptosis, and tumorigenesis (137, 138). The binding of NF-«B to DNA measured by gel shift assays is inhibited under oxidizing conditions (139, 140) and by oxidized thioredoxin-1 (141). Thioredoxin-1 increases the DNA binding of NF-«B to DNA and is more active than L-cysteine, reduced glutathione, and nonphysiological reducing agents such as N-acetyl cysteine, 2-mercaptoethanol, or dithiothreitol (140, 142). The binding of NF-«B to DNA requires that the Cys°’ of the NF-«B p50 subunit be reduced (143). If

the Cys® of one subunit is linked in a disulfide bridge with the Cys of the other in uncomplexed NF-«B, the DNA can no longer gain access to the bind-

ing surface of the p50 homodimer. If the p50 Cys is disulfide-linked to DNA, it can no longer be released from the complex (144, 145). NMR has been used to show a disulfide-bonded complex between a catalytically inactive mutant human thioredoxin-1 (Cys*°—> Ala,Cys°’— Ala, Cys®—> Ala,Cys’?— Ala) and a 13residue peptide comprising residues 56—68 of the p50 subunit of NF-«B that encompasses the critical Cys®’ residue (56). Both stable and transient transfection of cells with human thioredoxin-1 have been found to increase NF-«B transactivation measured with a reporter construct (143, 146). However, other studies have found that transient transfection with thioredoxin-1 or addition of thioredoxin-1 to the

272

POWIS

® MONTFORT

medium results in a dose-dependent inhibition of constitutive as well as phorbol ester—-stimulated NF-«B DNA binding and transactivation (147). This difference may be due to the use of reporter constructs that contain different NF-« B promoter elements or to the differing roles thioredoxin-1 plays in the cytosol (inhibition of I «B degradation) and in the nucleus (NF-«B activation).

A second transcription factor whose activity is regulated by Trx is the glucocorticoid receptor, which is subject to redox modulation through critical cysteine residues, whose oxidation leads to decreased ligand binding activity (148) and decreased DNA

binding (74). Transfection of cells with thioredoxin-1

has been

shown to increase the expression of glucocorticoid receptor mediated genes and glucocorticoid receptor reporter activity in oxidant treated cells, whereas transfection with antisense thioredoxin-1 decreases the expression of the genes and glucocorticoid receptor reporter activity (86, 149-151). Thioredoxin-1 associates with the DNA binding domain of the glucocorticoid receptor in the nucleus, which may account for the increased glucocorticoid receptor DNA binding activity caused by thioredoxin-1 (86). The transcription factor AP-1 (Fos and Jun homo- and heterodimers), whose activation is closely correlated with increased cell growth, is redox regulated (147, 152). DNA binding of AP-1 is increased by the reduction of a single conserved Cys residue in the DNA binding domain of each of the homodimers (131). Mutant Fos and Jun proteins in which this Cys residue is replaced by Ser show constitutive DNA binding (131). Cells transfected with human thioredoxin-1 show an increase in AP-1 activity measured with a reporter construct (146). Thioredoxin-1

does not reduce AP-1 directly but does so through another nuclear redox protein Ref-1, (153). Ref-1 is a 37-kDa protein that also has an apurine/apyrimidine endonuclease repair activity and a core domain that is highly conserved in a family of prokaryotic and eukaryotic DNA repair enzymes (154). Sequences in the N-terminal domain of Ref-1 are required for redox activity while C-terminal sequences are necessary for DNA repair activity (154). Other transcription factors whose binding to DNA is increased by thioredoxin-1 are AP-2 (155), the estrogen receptor (156), and transcription factor PEBP2/CBF (157). Cotransfection experiments have shown that thioredoxin-1 and Ref-1 increase the transactivating activity of the C-terminal activation domain of hypoxiainducible factor la (HIFla@), which contains a specific cysteine residue essential for the hypoxia inducible interaction with the CREB binding protein that leads to increased expression of genes such as erythropoietin (87). p53 is a tumor suppressor protein and transcription factor found to be deleted in a large number of human cancers (158). p53 is often referred to as the gatekeeper of the genome, and it induces G, arrest to allow cells time to repair DNA damage or, if the damage is too great, to induce apoptosis. The binding of p53 to DNA is enhanced by phosphorylation by specific kinases in response to DNA damage leading to the regulation of cell cycle-related genes (159, 160). p53 binding to DNA is also redox sensitive and is inhibited by oxidizing conditions (161). The redox regulation of p53 DNA binding occurs through critical cysteine residues in

THIOREDOXIN

273

the DNA binding domain of p53, the mutation of which markedly decreases DNA binding (127). When wild-type but not mutant forms of human p53 are expressed in the fission yeast Saccharomyces pombe, strong growth inhibition occurs (162). This was used as a model system to screen for genes whose function is required for normal activity of p53, and a mutant yeast strain partially resistant to the effects of human p53 expression was found that contained a recessive mutation in a novel gene, trrl, which has strong homology to thioredoxin reductase (163). The levels and localization of the p53 protein were unchanged in the mutant yeast strain, suggesting that it was not p53 expression that was altered. Loss of trr/ function resulted in yeast with an increased sensitivity to the toxic effects of H,O, and a 100% oxygen atmosphere. Studies in the budding yeast Saccharomyces cerevisiae have shown that deletion of the trr/ gene inhibits the ability of human p53 to stimulate reporter gene expression (164, 165). The effect of the trr] gene on p53 transactivation, but not DNA binding, was exerted through the negative regulatory domain rather than the DNA binding or oligomerization domains of p53 (165). Human thioredoxin-1 at concentrations as low as 10 nM has been shown to increase the sequence-specific binding of p53 to DNA measured by a gel shift assay and to increase p53 transactivating activity measured with a p21 reporter construct and p21 protein levels in cells transiently transfected with thioredoxin-1 (88). The redox inactive mutant thioredoxin-1 C32S/C35S partly blocks the increase in p53 transactivation caused by cisplatin treatment of cells. The DNA binding of p53 and transactivating activity in cells is also increased by the nuclear redox protein Ref-1 (88, 166). Thioredoxin-1, which as previously noted, associates with and

reduces Ref-1, increases the effects of p53 on DNA binding and cotransfection of cells with the redox inactive mutant thioredoxin-1. C32S/C35S decrease the Ref-1 mediated p53 transactivation, suggesting a role for both thioredoxin-1s in the Ref-1-mediated transactivation of p53 (88). Thioredoxin peroxidase 5 inhibits p53-induced apoptosis, raising the possibility that part of the effect of thioredoxin-] on p53 activity is mediated through scavenging of H,O, (69).

Protein Binding Thioredoxin-1 binds to a variety of cellular proteins (Table 1). Protein binding occurs only with the reduced but not the oxidized or mutant redox-inactive C32S/C35S forms of thioredoxin-1 (167-170). The mechansim of thioredoxin-1

binding has not been elucidated but may involve mixed disulfide formation between a catalytic site cysteine residue and a cysteine on the other protein. A protein to which thioredoxin-1 binds and that has received attention is apoptosis signal-regulating kinase 1 (ASK1). ASK1 is an activator of the c-Jun N-terminal

kinase (JNK) and p36 MAP kinase pathways and is required for TNF-a-induced apoptosis (171, 172). Reduced but not oxidized or redox inactive C32S/C35S mu-

tant thioredoxin-1 binds to the N-terminal portion of ASK1 to inhibit its activity, and thioredoxin-1 also inhibits ASK1-dependent apoptosis when transiently transfected into MvILu containing an inducible ASK1 (167). Deletion of specific

POWIS

274

® MONTFORT

TABLE 1. Thioredoxin-1 binding to proteins

Protein

Comments

Method

Ref

Trx-1

Homodimerization

X-ray

(26, 45, 228, 2249)

Ref-1

Trx reduces Ref-1 which then reduces AP-1 and p53 Ref-1 peptides bind to a boat shaped cleft in Trx-1

Mammaliantwo hybrid assay NMR

(88)

(Apoptosis signal-regulated kinase) binds reduced but not oxidized or redox inactive mutant Trx and is inhibited. TNF-a causes ROS mediated dissociation of Trx and ASK1 activation

Yeast two hybrid assay

(167)

Trx binding inhibits autophosphorylation and histone phosphorylation by PKC

Phage display panning

(230)

NF-«B

NF-«B peptides bind to a boat shaped cleft in Trx-1

NMR

(56)

vit D(3) up regulated protein-1

Binds reduced but not oxidized or redox inactive mutant Trx, in cells vit D(3) upregulated protein-1 expression inhibits Trx expression

Yeast two hybrid assay

(231)

Glucocorticoid

Trx associates with the GR receptor

—

(86)

ASKI1

PKCa,é,¢

receptor

and ¢

(57)

(168)

in the nucleus under oxidizing conditions

p40 phox

Cytosolic component of phagocyte oxidase, does not bind mutant redox Trx

Yeast two hybrid assay

(170)

Lipocalin

Lipid binding protein and cystein proteinase inhibitor from human

Phage display panning

(232)

tears and other tissues

N-terminal residues renders ASK1 constitutively active and no longer influenced by thioredoxin-1 (167). Oxidation of thioredoxin-1 by the TNF-a or stress- induced generation of reactive oxygen species leads to dissociation of thioredoxin-1 and the activation of ASK1, which may contribute to TNF-a induced apoptosis. A role for the release of thioredoxin-1 bound to ASK-1 in mediating TNF-a induced apoptosis has been proposed (167, 168). However, whether Trx binding to other proteins has physiological significance remains to be determined. The Trx-related

protein p32'™" has also been found to bind to the catalytic fragment of mammalian

THIOREDOXIN

275

STE-20-like (MST) kinase (40). MST is proteolytically activated by caspases during CD95 (FAS, Apo-1) induced apoptosis and has been suggested to generate apoptotic signals downstream of caspase activation (173).

Inhibition of Apoptosis Thioredoxin-1 added to culture medium prevents apoptosis of lymphoid cells induced by L-cysteine and glutathione depletion (174) and also protects B-type chronic lymphocytic leukemia cells against apoptosis associated with an increase in TNF-a

secretion, which is an autocrine growth factor for these cells (175).

We have shown that stable transfection of mouse WEHI7.2 lymphoid cells with human thioredoxin-1 inhibits apoptosis induced by a variety of agents, including dexamethasone, staurosporine, thapsigargin, and etoposide (176). The inhibition of apoptosis caused by transfection with thioredoxin-1 is similar to the pattern of inhibition of apoptosis caused by transfection of the cells with the anti-apoptosis oncogene bcl-2. When innoculated into scid mice, the thioredoxin-1-transfected cells form tumors that grow more rapidly and show less spontaneous apoptosis than do vector-alone or Bcl-2 transfected cells. The tumors are also resistant to growth inhibition by dexamethasone. Thus, thioredoxin-1 offers a survival as well as a growth advantage to tumors in vivo, unlike Bcl-2, which offers only a survival advantage and requires other genetic changes to stimulate tumor growth (177). The mechanism for the anti-apoptotic effects of thioredoxin-1 is unknown. As noted above a role for thioredoxin-1 binding to to ASK1 in mediating TNF-a induced apoptosis has been proposed (167, 168).

Noncatalytic Site Activities of Trx While most attention has focused on the redox-dependent biological activities of Trx that require the catalytic site Cys residues, there may be activities that are independent of this redox activity, although they could still require other reduced cysteine residues. Precedents exist with E. coli Trx where catalytic site redox inactive mutants are as active as the wild-type protein in supporting in vivo F1 and M13 phage assembly (178) and the activation of phage T7 DNA polymerase (179). A catalytic site mutant human thioredoxin-1 in which both catalytic site

Cys residues are converted to Ser (Cys**—> Ser/Cys*°— Ser) retains its ability to cause an increase in the rosette inhibition titre of serum (180). Changing Cys” but not Cys® to Ser completely abolished the activity of thioredoxin-1 in the

rosette inhibition assay. Because Cys’? is needed for stable dimer formation by thioredoxin-1 (45), it is possible that dimerization or association of thioredoxin- |

with another protein is necessary for the rosette modifying activity. Truncated forms of human thioredoxin-1 retaining the conserved active site but lacking the C-terminal 16 or 24 amino acids are without insulin disulfide reductase activity and show increased eosinophil cytotoxic activity compared to wild-type thioredoxin-1 (43).

276

POWIS

® MONTFORT

Cancer Drug Resistance Several lines of evidence suggest that Trxs may play a role in resistance to the cellkilling effects of anticancer drugs. The sensitivity of adult T-cell leukemia cell lines to doxorubicin is lowest in those cell lines with the highest levels of thioredoxin-1 (181). In contrast, leukemia cell lines, which overall have lower thioredoxin-1

levels than do epithelial cancer cell lines, together with colon and renal cancer cell lines show the greatest sensitivity to thioredoxin-1 inhibiting disulfide drugs with antitumor activity and other inhibitors of thioredoxin-1 reductase (182). Human hepatoma cells with increased thioredoxin-1 show a decreased sensitivity to cell killing by cisplatin but not by doxorubicin or mitomycin C (85). Ithas been reported that bladder and prostate cancer cell lines made resistant to cisplatin have four- to sixfold increases in levels of thioredoxin-1 mRNA and thioredoxin-1 protein (183). This resistance to cisplatin could be reversed by lowering thioredoxin-1 levels with a thioredoxin-1 antisense expression plasmid, which also increases the sensitivity of the cells to doxorubicin, mitomycin C, etoposide, H,O,, and UV irradiation but not to vincristine and colchicine. Gastric and colon cancer cell lines resistant to cisplatin have also been reported to have up to 2.5-fold increases in levels of thioredoxin- | protein and up to twofold increases in Trx reductase activity, whereas a positive correlation between cisplatin resistance and Trx-1 levels was found in a panel of 11 ovarian cancer cell lines (184). In a follow-up study by the same group, stable transfection of ovarian and colon cancer cell lines with thioredoxin-1 cDNA giving a 2- to 2.5-fold increase in thioredoxin-1 levels failed to increase the resistance to cisplatin, doxorubicin, and mitomycin C (185). However, others have found that stable transfection of fibrosarcoma cells with thioredoxin-1 giving a 2fold increase in thioredoxin-1 levels gave a 3-fold increase in resistance to cisplatin (186). In the cisplatin resistant cell lines, a relatively small increase in thioredoxin1 expression of two- to threefold was associated with a cisplatin resistance of 4to 20-fold (85, 184, 187). Thioredoxin-1 levels are also increased in mitoxantrone

resistant cells (188). Thus, thioredoxin-1 appears to be necessary but not sufficient for the anticancer drug resistance.

Reperfusion Injury Recombinant human thioredoxin-1 has been shown to exert a protective effect against reperfusion-induced arrhythmias in an isolated rat heart model, with a concentration of 0.1 44M thioredoxin-1 being the most effective (189). In a rat in

vivo model of lung ischemia reperfusion injury, recombinant human thioredoxin-1 improved lung function and lessened the histological signs of damage (190). Ina rabbit in vivo model of lung reperfusion injury, human recombinant thioredoxin-1 had no effect on lung function or on lipid peroxides in the lung but attenuated the signs of lung damage including intra-alveolar exudation, interstitial thickening, and cellular infiltration (191). In the same rabbit in vivo model, N-acetyl cysteine did not provide any protection against lung damage. In a dog lung transplantation model, human recombinant thioredoxin-1 as well as N-acetyl cysteine protected against loss of pulmonary function and prevented histological signs of damage

THIOREDOXIN

277

(192). Thioredoxin-1 expression is increased in rat brain after transient and permanent cerebral occlusion (193). Thioredoxin-1 transgenic mice showed smaller

infarct sizes after middle cerebral artery occlusion (194). Thioredoxin-1 protects endothelial cells against reperfusion injury (169) and protects rats against retinal ischemia reperfusion injury (195).

Transgenic and Knockout Animals The many biological properties of Trx make it of considerable interest to have animals showing increased or decreased expression of Trx. Mice with targeted disruption of the thioredoxin-1 gene show that homozygous animals die shortly after implantation whereas heterozygous animals are viable, fertile, and in other respects apparently normal (196). The lethal effects of thioredoxin-1 in early development is a finding consistent with the one that thioredoxin-1 is widely distributed in different organs and tissues during fetal development (197). Transgenic mice with human thioredoxin-1 under the control of a B-actin promoter have been reported (194). Heart showed the highest level of increased expression of human—relative

to mouse—thioredoxin-1, with kidney, brain, lung, skin, and liver showing lower levels. The mice were functionally normal, and no histological abnormalities were

observed. There was no change in the expression of Cu Zn superoxide dismutase, Mn superoxide dismutase, or glutathione peroxidase. The thioredoxin-1| transgenic mice showed smaller infarct sizes after middle cerebral artery occlusion and a subsequent, larger increase in C-fos expression, observations that lead to the suggestion that thioredoxin-1 may have a neuroprotective function through the activation of AP-1. Selective expression of human thioredoxin-1| under the control of a human insulin promoter in the f-islet cells of the pancreas protected the mice against spontaneous diabetes and against streptozotocin-induced apoptosis (198).

Other Activities Thioredoxin-1 can act as a catalyst of protein folding because of relatively weak protein disulfide bond isomerizing activity (199). Thioredoxin-1 directly reactivates proteins that have been inactivated by oxidative stress, including glyceraldehyde-3-phosphate dehydrogenase (200), iodothyronine S’-deiodinase (201), and ornithine decarboxylase (202). Protein disulfide isomerase, which has two regions homologous to the Trx catalytic site with -Trp-Cys-Gly-His-Cys-Lys- (203), is several-fold more efficient than thioredoxin-1 at catalyzing the isomerization, and because it is abundant in the endoplasmic reticulum lumen, where most protein disulfides are formed, it is the major physiological catalyst of protein disulfide isomerization (199). It is interesting that the gonadotrophic hormone subunit B-lutropin has a Trx- like -Cys-Gly-Pro-Cys- sequence and a f-follitropin CysGly-Lys-Cys sequence, and both are several-fold more active as protein disulfide isomerases than Trx (204).

Thioredoxin-1 was identified by a random screening of antisense expressed genes as a component of the pathway signaling the interferon-gamma growth arrest of HeLa cells (10). Thioredoxin-1 is also a chemotaxic factor for cells of

278

POWIS

® MONTFORT

leukocyte lineage (205). Chemotaxis is not exhibited by redox inactive mutant thioredoxin-1, and the effects are seen in the nM range, which is comparable to other known chemokines.

TRX IN PLASMA Plasma levels of thioredoxin-1 in normal individuals are between 10 and 80 ng/ml (0.8-6.6 nM) (188, 206-209). Studies have been conducted to determine whether plasma thioredoxin-1 is elevated in human disease. Serum thioredoxin-1 has been reported to be elevated almost twofold in patients with hepatocellular carcinoma, and after surgery to remove the tumor, the levels drop (188). Serum thioredoxin-1 is not elevated in patients with chronic hepatitis or liver cirrhosis (188). Plasma thioredoxin-1 is also elevated more than twofold in patients with pancreatic ductal carcinoma compared to normal individuals (84). Serum thioredoxin-1 has been reported to be elevated in patients with HIV (206), rheumatoid arthritis, and Sjogren’s syndrome (101, 210).

TRX IN HUMAN DISEASE

Skin Damage Normal skin shows thioredoxin-1 immunohistochemical staining in the sebaceous glands, secretory components of sweat glands, and the outer root sheath of the hair follicle, but not in the interfollicular epidermis (211). Enhanced expression of thioredoxin-1 has been demonstrated by immunohistochemical staining in the epidermal cells of sun-exposed skin (67) and may be a protective response to oxidative damage.

Atherosclerosis NO and peroxynitrite contribute to the damage to smooth muscle and endothelial cells seen during atherosclerotic plaque formation (212). Thioredoxin-1 prevents the NO-dependent inhibition of purified NO-synthase (213), and when human thioredoxin-1 was transiently transfected into porcine pulmonary artery endothelial cells, it protected against a loss of NO-synthase activity (214). Thioredoxin1-transfected L929 murine fibrosarcoma cells show resistance to peroxynitrit einduced cytotoxicity, although the effect is small (215). Thioredoxin-1 mRNA and thioredoxin reductase mRNA are increased in endothelial cells and macrophages of human atherosclerotic plaques, leading to the suggestion that thioredoxin-1 plays a role in the pathogenesis of atherosclerosis (gD):

Immune Function Thioredoxin-1 is a component of early pregnancy factor in serum that provides immune protection of the developing embryo (29). Thioredoxin-1 is initially

THIOREDOXIN

279

produced by maternal tissues, but the fetus begins thioredoxin-1 production after implantation (216). Eosinophils are inflammatory cells associated with inflammatory responses, allergic diseases, and tumor cytotoxicities. Thioredoxin-1

in-

creases eosinophil migration, cytotoxicity, and the release of major basic protein from eosinophils (217-219). Thioredoxin-1 expression is increased in infiltrating B cells and epithelial cells of patients with Sjogren’s syndrome, an EB V-associated autoimmune disease (220).

Alzheimers Disease The brains of subjects with Alzheimer’s disease show decreased levels of thioredoxin-1, particularly in the amygdala and hippocampus, while at the same time the thioredoxin reductase activity is increased (221). It was suggested that these changes may contribute to the increased oxidative stress and subsequent neurodegeneration observed in Alzheimer’s disease.

HIV Trx-expressing cells are absent from the lymph nodes of patients with AIDS and AIDS related complex (222). However, plasma levels of thioredoxin-1 are significantly elevated in patients with AIDS (206). Approximately 25% of AIDS patients had thioredoxin-1 levels higher than the highest level found in plasma of control subjects and tended to have lower overall CD4 counts. Furthermore,

increased

plasma thioredoxin-1 corresponded with decreased cellular thiols and altered cell surface antigen expression (CD62L, CD38 and CD20) that occurs in the later stages of HIV infection. Added thioredoxin-1 has been reported to inhibit the expression of HIV in human macrophages as measured by p24 antigen production and integration of the provirus as well as expression of the integrated virus in chronically infected cells (223). Thioredoxin-1 was considerably more potent than N-acetyl cysteine in inhibiting HIV production. Surprisingly a truncated form of thioredoxin-1, human eosinophil cytotoxicity factor, potentiated HIV production.

Cancer Studies with a variety of human primary tumors have shown that thioredoxin-| is overexpressed in the tumor compared to levels in the corresponding normal tissue (Table 2). We have shown by immunohistochemical studies using paraffin embedded tissue sections that thioredoxin-1 expression is increased in more than half of human primary gastric cancers. The thioredoxin-1 levels showed a highly significant positive correlation (p < 0.001) with cell proliferation measured by nuclear proliferation antigen and a highly significant negative correlation (p < 0.001) with apoptosis measured by the terminal deoxynucleotidyl transferase (Tunel) assay. In a recent immunohistochemical study of thioredoxin-1 levels in human colon cancer, thioredoxin-1 protein was not increased compared to nor-

mal colonic mucosa in precancerous colon polyps but was increased sixfold in

280

POWIS

TABLE 2

® MONTFORT

Thioredoxin overexpression by human primary cancers

Tumor

Type

Number of subjects

Percent overexpressed

Reference

Lung

mRNA

10

50"

(11)

Colon

mRNA protein

10 18

604 a5

(60) (224)

Cervix

protein

9

bss

(73)

Hepatoma

mRNA protein

20 25

85 52

(85) (72)

Gastric

protein

10

50

(82)

Pancreatic

protein

58

4]

(84)

29

7 (differentiated)

Squamous cell carcinoma

protein

6 metastatic

6

(211)

Myeloma

protein

10

9

unpublished

Non—Hodgkins lymphoma

protein

20

7 (advanced)

unpublished

Acute lymphocytic leukemia

protein

33

15

unpublished

_ SSSSSSSSSSSS —— FSFSMMMmmmsmsmeffseseF “Compared to corresponding normal tissue from the same subject. >Immunohistochemistry and comparison to normal tissue from the same subject.

primary colon cancers and almost ninefold in metastatic colon tumors in adjacent lymph glands (224). Furthermore, high levels of thioredoxin-1 in the tumor appear to be associated with decreased patient survival (J Raffel, AK Bhattacharyya & G Powis unpublished observations).

DRUGS THAT INHIBIT TRX The growth-stimulating effects of thioredoxin-1, together with the finding that it is overexpressed by a number of human primary tumors, raise the intrigui ng possibility that thioredoxin-1 contributes to aggressive tumor growth and poor patient prognosis (25). Furthermore, because thioredoxin-1 also inhibits apoptosis caused by a number of anticancer drugs and is a cause of resistance to the cytotoxic effects of some anticancer drugs, it is possible that increased thioredoxin-1 could be a cause of resistance to chemotherapy. These findings make thioredoxin-1 an attractive target for the development of drugs to inhibit cancer cell growth. Several such compounds have been identified. They include 1-methylhydroxypropyl 2-imidazoloyl disulfide (PX-12, Figure 5), which was identified as an inhibitor of thioredoxin-1 by binding to the Cys’? residue (225). The median ICs» for growth inhibition of a variety of cell lines by PX-12 is 8.1 wM (177). PX-12 has been shown to have in Vivo antitumor activity against human tumor xenograf ts in scid mice (226). The growth inhibition by compound PX-12 ina 60 human tumor panel was significantly

THIOREDOXIN

N

O

L Sane CHCH,CH, Pherae

{H,CNHC

PX-12 Figure 5

281

CHN(CH,), NSC 131233

Structures of PX-12 and NSC 131233.

correlated with the expression of thioredoxin-1 mRNA (182). Several other inhibitors of thioredoxin-1 have been identified by the COMPARE program, from over 50,000 compounds tested by the Naional Cancer Institute in the 60 human tumor cell line panel, as having a pattern of cell-killing activity similar to PX-12 (227). One of these compounds, 2,5-bis{dimethylamino)methyl]cyclopentanone (NSC 131233), is an irreversible inhibitor of thioredoxin-1 with a Ki of 1.0 uM.

SUMMARY AND CONCLUSIONS The mammalian Trxs have been known for many years. However, only recently has the multiplicity of biological functions of Trx become apparent. The most studied Trx, thioredoxin-1, is found in the cytosol and the nucleus and acts as a cofactor that provides reducing equivalents to other redox enzymes. It can directly reduce cysteine groups on proteins to alter protein binding activity, for example of transcription factors, and it can bind to proteins to alter their enzymatic activity. Thioredoxin-1 is a secreted protein, and it acts outside the cell to stimu-

late cell growth and inside the cell to stimulate cell growth and inhibit apoptosis. Whether these extracellular and intracellular activities of Trx-1 are related is not clear at this time; however, redox activity is necessary for both. The ability of thioredoxin-1 to homodimerize, leading to a loss of redox activity, is conserved among mammalian species. Whether homodimerization occurs in cells and is of biological significance remains to be established. A new mitochondrial Trx family member, thioredoxin-2, has recently been identified, but its function remains

unknown. Thioredoxin-1 has multiple biological activities including antioxidant, growth stimulation, and inhibition of apoptosis. However, the mechanisms responsible for most of these activities have not been established, and their identification

provides a challenge for future work. Thioredoxin-1 may play a role in a number of human diseases but particularly cancer, in which increased levels of thioredoxin-1 are found in many tumors and are associated with aggressive tumor growth. Thus, thioredoxin-1 presents a target for drugs to inhibit cancer cell growth. The first generation of drugs that inhibit thioredoxin-1 are currently in development. The challenges for the future are to identify a unifying mechanism for the various biological

282

POWIS

® MONTFORT

function activities of the Trxs and to determine exactly how redox control of protein cell. the of ent environm lar intracellu reducing highly the in by Trx can occur ACKNOWLEDGMENTS

Supported by NIH grants CA48725, CA78277 and CA77204 (GP) and Arizona Disease Control and Research Commision Grant 2005

(WM and GP).

Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED

of thioredoxin, the hydrogen donor from coli B. J. Biol. Chem. 239: 3436-44 . Holmgren A. 1985. Thioredoxin. Annu. Rev. Biochem. 54:237-71

10. Deiss LP, Kimchi A. 1991. A genetic tool used to identify thioredoxin as a mediator of a growth inhibitory signal. Science 252:117-20 . Gasdaska PY, Oblong JE, Cotgreave IA, Powis G. 1994. The predicted amino acid sequence of human thioredoxin is identical to that of the autocrine growth factor human

. Nakamura H, Nakamura K, Yodoi J. 1997.

adult T-cell derived factor (ADF): Thiore-

iI, Laurent TC, Moore EC, Reichard P. 1964. Enzymatic synthesis of deoxyribonucleotides VI. Isolation and characterization Escherichia

Redox regulation of cellular activation. Annu. Rev. Immunol. 15:351—69 . Follmann H, Haberlein I. 1995. Thioredox-

ins: Universal, yet specific thiol-disulfide redox cofactors. BioFactors 5:147-S6 . Moore EC. 1967. A thioredoxin-thioredoxin

.

.

.

.

reductase system from rat tumor. Biochem. Biophys. Res. Commun. 29:264—68 Herrmann EC, Moore EC. 1973. Purification of thioredoxin from rat Novikoff ascites hepatoma. J. Biol. Chem. 248:1219—23 Wakasugi H, Rimsky L, Mahe Y, Kamel AM, Fradelizi D, et al. 1987. Epstein-Barr virus-containing B-cell line produces an interleukin | that it uses as a growth factor. Proc. Natl. Acad. Sci. USA 84:804—-808 Tagaya Y, Okada M, Sugie K, Kasahara T, Kondo N, et al. 1988. IL-2 receptor(p55)/ TAC-inducing factor purification and characterization of adult T cell leukemia-derived factor. J. Immunol. 140:2614—20 Clarke FM, Orozco C, Perkins AV, Cock I,

Tonissen KF, et al. 1991. Identification of molecules involved in the ‘early pregnancy factor’ phenomenon. J. Reprod. Fert. 93: 525-39

doxin mRNA is elevated in some human tumors. Biochim. Biophys. Acta 1218:29296 . Holmgren A. 1989. Thioredoxin and glutaredoxin systems. J. Biol. Chem. 264: 13963-66 13: Mustacich D, Powis G. 2000. Thioredoxin reductase. Biochem. J. 346:1-8 14. Gladyshevy VN, Jeang K, Stadtman TC. 1996. Selenocysteine, identified as the penultimate C-terminal residue in human T-cell thioredoxin reductase, corresponds to TGA in the human placental gene. Proc. Natl. Acad. Sci. USA 93:6146-51 —_nN.

Powis

G,

Gasdaska

JR,

Gasdaska

PY,

Berggren M, Kirkpatrick DL, et al. 1997. Selenium and the thioredoxin redox system: effects on cell growth and death. Oncol. Res. 9:303-12 16. Gasdaska JR, Harney JW, Gasdaska PY, Powis G, Berry MJ. 1999. Regulation of human thioredoxin reductase expression and activity by 3/-untranslated region selenocysteine insertion sequence and mRNA instability elements. J. Biol. Chem. 274:25379-85

THIOREDOXIN Lie Zhong L, Arner ESJ, Ljung J, Aslund F, Holmgren A. 1998. Rat and calf thioredoxin reductase are homologous to glutathione reductase with a carboxylterminal elongation containing a conserved catalytically active penultimate selenocysteine residue. J. Biol. Chem. 273:858191 18. Gasdaska PY, Gasdaska JR, Cochran S, Powis G. 1995. Cloning and sequencing of human thioredoxin reductase. FEBS Lett. 373:5-9 I) Gasdaska PY, Berggren MM, Berry MJ, Powis G. 1999. Cloning, sequencing and functional expression of a novel human thioredoxin reductase. FEBS Lett. 442:105-11 20. IECCUnS eeieee Ky Olan Ke aOON EL, Levine RL, et al. 1999. Molecular cloning and characterization of a mitochondrial selenocysteine-containing thioredoxin reductase from rat liver. J. Biol. Chem. 274:4722-34 Ply Miranda-Vizuete A, Damdimopoulos AE, Pedrajas JR, Gustafsson J-A, Spyrou G. 1999. Human mitochondrial thioredoxin reductase: cDNA cloning, expression and genomic organization. Eur J. Biochem. 261:405-11 2, Miranda-Vizuete A. 1999. TrxR3 a novel human thioredoxin reductase. GenBank Accession AF133519:bp 1—2165 US. Gotoh S, Terashima H, Yagi K, Mizoguchi T. 1998. Cloning and sequence of bovine thioredoxin cDNA. GenBank Accession AF104105:bp 1-550 24. Matsui M, Taniguchi Y, Hirota K, Taketo

283

Oxidative inactivation of thioredoxin as a cellular growth factor and protection by a Cys!?> Ser mutation. Biochem. Pharmacol. 52:1741-47

Padi Heppell-Parton A, Cahn A, Bench A, Lowe N, Lehrach

H, et al. 1995. Thioredoxin,

a mediator of growth inhibition, maps to 9q31. Genomics 26:379-81 28. Kaghad M, Dessarps F, Jacquemin-Sablon H,

Caput

D,

et

al.

1994.

Genomic

cloning of human thioredoxin-encoding gene: mapping of the transcription start point and analysis of the promoter. Gene 140:273-78

29. Tonissen KF, Wells JRE. 1991. Isolation and characterization of human thioredoxin-encoding genes. Gene 102: 221-28 30. Miranda-Vizuete A. 1999. Human thioredoxin | pseudogene 2. GenBank Accession AF146024:bp 1-1210 Sl. Taketo M, Matsui M, Rochelle JM, YodoiJ, Seldin MF. 1994. Mouse thioredoxin

gene maps on chromosome 4, whereas its pseudogene maps on chromosome |. Genomics 21:251-53

3, Wollman EE, d’ Auriol L, Rimsky L, Shaw A, Jacquot J, et al. 1988. Cloning and expression of cDNA for human thioredoxin. J. Biol. Chem. 263:15506-12 33) Rosen A, Lundman P, Carlsson M, Bhavani K, Srinivasa BR, et al. 1995. A CD4T T cell line-secreted factor, growth promoting for normal and leukemic B cells, identi-

fied as thioredoxin. Inter. Immunol. 7:625— 33

34, Hariharan J, Hebbar P, Ranie J, Philom-

M, Yodoi J. 1995. Structure of the mouse

ena Sinha AM, Datta S. 1996. Alternative

thioredoxin-encoding gene and its processed pseudogene. Gene 152:165—71 JR, Berggren M, PowisG. 1995. 25) Gasdaska Cell growth stimulation by the redox protein thioredoxin occurs by a novel helper mechanism. Cell Growth Differ. 6:164350

forms of the human thioredoxin mRNA: identification and characterization. Gene 173:265-70 SBE Berggren M, Mustacich D, Powis G. 2000. Absence of mutant forms of thioredoxin-| (TRX-1) in human tumor cell lines and evidence for an alternatively spliced TRX-1 MRNA. Amer. Assoc. Cancer Res. 41:151 (Conf.)

26. Gasdaska JR, Montfort

W,

Hill SR, Weichsel

Kirkpatrick L, A,

et al. 1996.

284

POWIS

® MONTFORT thioredoxins: evidence for a regulatory homodimer. Structure 4:735-51

36. Bodenstein J, Follmann H. 1991. Characterization of two thioredoxins in pig heart including a new mitochondrial protein. Z. Naturforsch. 46c:270—79 (In German)

Sik Spyrou G, Enmark E, Miranda-Vizuete A, Gustafsson J. 1997. Cloning and expression of a novel mammalian thioredoxin. J. Biol. Chem. 272:2936-41 38. Miranda—Vizuete A, Gustafsson J—A, Spyrou G. 1999, Human mitochondrial thioredoxin. GenBank Accession HSU78678.1:bp 1-756 Sy) Miranda-Vizuete A, Gustafsson J, Spyrou G. 1998. Molecular cloning and expression of acDNA encoding a human thioredoxinlike protein. Biochem. Biophys. Res. Commun. 243:284-88 40. Lee K, Murakawa M, Takahashi S, Tsubuki S, Kawashima S, et al. 1998. Purification, molecular cloning, and characterization of TRP32, a novel thioredoxin-related mam-

malian protein of 32 kDa. J. Biol. Chem. 273:19160-66 41. Kurooka H, Kato K, Minoguchi S, Takahashi Y, Ikeda J, et al. 1997. Cloning and

42.

.

44.

45.

characterization of the nucleoredoxin gene that encodes a novel nuclear protein related to thioredoxin. Genomics 39:331-39 Balcewicz-Sablinska MK, Wollman EE, Gort R, Silberstein DS. 1991. Human eosinophil cytotoxicity-enhancing factor II. Multiple forms synthesized by U937 cells and their relationship to thioredoxin/adult T cell leukemia-derived factor. J. Immunol. 147:2170-74 Silberstein DS, McDonough S, Minkoff MS, Balcewicz-Sablinska MK. 1993. Human eosinophil cytotoxicity-enhancing factor. J. Biol. Chem. 268:9138-42 Qin J, Clore G, Gronenborn A. 1994. The high-resolution three-dimensional solution structures of the oxidized and reduced state of human thioredoxin. Structure 2: 502-22 Weichsel A, Gasdaska JR, Powis G, Montfort WR. 1996. Crystal structures of reduced, oxidized, and mutated human

46. Katti SK, LeMaster DM, Eklund H. 1990. Crystal structure of thioredoxin from Es-

cherichia coli at 1.68 A resolution. J. Mol. Biol. 212:167-84 47. Holmgren A, Soderberg B-O, Eklund H, Branden C-I. 1975. Three-dimensional structure of Escherichia coli thioredoxin-

§2 to 2.8 A resolution. Proc. Natl. Acad. Sci. USA 72:2305-309 48. Jeng MF, Campbell A, Begley T, Holmgren A, Case D, et al. 1994. High-resolution solution structures of oxidized and reduced Escherichia coli thioredoxin. Structure 2:853-68 49. Eklund H, Gleason F, Holmgren A. 1991. Structural and functional relations among thioredoxins of different species. Proteins 11:13-28 50. Holmgren A. 1995. Thioredoxin structure mechanism: conformational changes on oxidation of the active-site sulfhydryls to a disulfide. Structure 3:239-43 51. Takahashi N, Creighton TE. 1996. On the reactivity and ionization of the active site cysteine residues of Escherichia coli thioredoxin. Biochem. 35:8342—53 . Ren X, Bjornstedt M, Shen B, Ericson ML, Holmgren A. 1993. Mutagenesis of structural half-cystine residues in human thioredoxin and effects on the regulation of activity by selenodiglutathione. Biochem. 32:9701-705 nNww.

Andersen JF, Sanders DAR, Gasdaska JR,

Weichsel A, Powis G, Montfort WR. 1997. Human thioredoxin homodimers: regulation of pH, role of Aspartate 60, and crystal structure of the Aspartate 60 > Asparagine mutant. Biochem. 36:13979-88 54. Gronenborn AM, Clore GM, Louis JM, Wingfield PT. 1999. Is human thioredoxin monomeric or dimeric? Protein Sci. 8: 426-29 SBE Holmgren A, Luthman M. 1978. Tissue distribution and subcellular localization of bovine thioredoxin determined by

THIOREDOXIN

radioimmunoassay. el

Biochem.

17:4071—

56. Qin J, Clore GM, Kennedy WP, Hutch JR, Gronenborn AM. 1994. Solution structure of human thioredoxin in a mixed disulfide intermediate complex with its target peptide from the transcription factor NF— kappaB. Structure 3:289-97 Sik. Oin J, Clore GM, Kennedy WP, Kuszewski J, Gronenborn

AM.

1996.

The

solution

structure of human thioredoxin complexed with its target from Ref-1 reveals peptide chain reversal. Structure 4:613-—20 58. Taniguchi Y, Taniguchi-Ueda Y, Mori K, Yodi J. 1996. A novel promoter sequence is involved in the oxidative stress-induced expression of the adult T-cell leukemia derived

factor (ADF)/human.

thioredoxin

(Trx) gene. Nucl. Acids Res. 24:2748-S2

Deh. Masutani H, Hirota K, Sasada T, UedaTaniguchi Y, Taniguchi Y, et al. 1996. Transactivation of an inducible antioxidative stress protein, human thioredoxin by HTLV-I tax. Immunol. Lett. 54:67-71 60. Berggren M, Gallegos A, Gasdaska JR, Gasdaska PY, Warneke J, Powis G. 1996.

Thioredoxin and thioredoxin reductase gene expression in human tumors and ceil lines, and the effects of serum stimulation and hypoxia. Anticancer Res. 16:345966

61. Denko N, Schindler C, Koong A, Laderoute K, Green

C, and Giaccia A. 2000.

Epigenetic regulation of gene expression in cervical cancer cells by the tumor microenvironment. Clin. Cancer Res. 6:480-87 62. Ericson ML, Horling J, Wendel-Hansen V, Holmgren A, and Rosen A. 1992. Secretion

of thioredoxin after in vitro activation of human B cells. Lymphokine Cytokine Res. 11:201-207 63. Ejima K, Koji T, Nanri H, Kashimura M, and Ikeda M. 1999. Expression of thioredoxin and thioredoxin reductase in placentae of pregnant mice exposed to lipopolysaccharide. Placenta 20:561—66

285

64. Das KC, Guo XL, White CW. 1999. Hyperoxia induces thioredoxin and thioredoxin reductase gene expression in lungs of premature baboons with respiratory distress and bronchopulmonary dysplasia. Chest 116:101s

65. Das KC, Guo X, White CW. 1999. Induction of thioredoxin and thioredoxin reductase gene expression in lungs of newborn primates by oxygen. Amer. J. Physiol. 276:L530-39 66. Hayashi T, Ueno Y, Okamoto T. 1993. Oxidoreductive regulation of nuclear factor kappa B. J. Biol. Chem. 268:11380-88 67. Sachi Y, Hirota K, Masutani H Toda K, Takahashi

O, et al. 1995.

Induction

of ADF/TRX by oxidative stress in keratinocytes and lymphoid cells. /mmunol. Lett. 44:189-93

68. Higashikubo A, Tanaka N, Noda N, Maeda I, Yagi K, et al. 1999. Increase in thiore-

doxin activity of intestinal epithelial cells mediated by oxidative stress. Biol. Pharm. Bull. 22:900—903

69. Zhou Y, Kok KH, Chun AC, Wong CM, Wu HW, et al. 2000. Mouse peroxiredoxin

V is a thioredoxin peroxidase that inhibits p53-induced apoptosis. Biochem. Biophys. Res. Commun. 268:921—27

70. Nakamura H, Matsuda M, Furuke K, Kitaoka

Y, Iwata

S, et al. 1994.

Adult T

cell leukemia-derived factor/human thioredoxin protects endothelial F-2 cell injury caused by activated neutrophils or hydrogen peroxide. Immunol. Lett. 42:75—80

7p, Tomimoto H, Akiguchi I, Wakita H, Kimura J, Hori K Yodoi J. 1993. Astroglial expression of ATL—derived factor, a human

thioredoxin homologue, in the gerbil brain after transient global ischemia. Brain Res. 625:1-8

1, Nakamura H, Masutani H, Tagaya Y, Yamauchi A, Inamoto Tet al. 1992. Expression and growth-promoting effect of adult T-cell leukemia-derived factor. A human thioredoxin homologue in hepatocellular carcinoma. Cancer 69:2091—97

286

POWIS

® MONTFORT

. Fujii S, Nanbu Y, Nonogaki H, Konishi I, Mori T, et al. 1991. Coexpression of

adult T-cell leukemia-derived factor, a human thioredoxin homologue, and human papillomavirus DNA in neoplastic cervical squamous epithelium. Cancer 68:1583—91

82.

74. Hutchison

KA, Matic G, Meshinchi S, Bresnick EH, Pratt WB. 1991. Redox ma-

nipulation of DNA binding activity and BuGR epitope reactivity of the glucocorticoid receptor. J. Biol. Chem. 266:10505— 509

12s Hoshi Y, Tanooka H, Miyazaki K, Wakasugi H. 1997. Induction of thioredoxin in human lymphocytes with low-dose ionizing radiation. Biochim. Biophys. Acta 1359:65-70 76. Leppa S, Pirkkala L, Chow SC, Eriks-

83.

84.

son JE, Sistonen L. 1997. Thioredoxin is

transcriptionally induced upon activation of heat shock factor 2. J. Biol. Chem. 272:30400-404

ke An G, Wu R. 1992. Thioredoxin gene expression is transcriptionally up-regulated by retinol in monkey conducting airway epithelial cells. Biochem. Biophys. Res. Commun. 183:170-75 78. Maruyama T, Sachi Y, Furuke K, Kitaoka Y, Kanzaki

H, et al. 1999. Induction

86.

of

thioredoxin, a redox-active protein, by ovarian steroid hormones during growth and differentiation of endometrial stromal cells in vitro. Endocrinol. 140:365—72 78). Ejima K, Nanri H, Araki M, Uchida K, Kashimura M, Ikeda M. 1999. 17betaestradiol induces protein thiol/disulfide oxidoreductases and protects cultured bovine aortic endothelial cells from oxidative stress. Eur. J. Endocrinol. 140:608—13 80. Sahlin L, Wang H, Masironi B, Holmgren A, Eriksson H. 1999, Regulation of thiore-

doxin mRNA in the rat uterus by gonadal steroids. J. Steroid Biochem. Mol. Biol. 68:203-209 81. U-Taniguchi

Y, Furuke K, Masutani H, Nakamura H, Yodoi J. 1995. Cell cycle in-

hibition of HTLV-1

85.

transformed T cell lines

87.

88.

by retinoic acid: the possible therapeutic use of thioredoxin reductase inhibitors. Oncol. Res. 7:183-189 Grogan TM, Fenoglio-Prieser C, Zeheb R, Bellamy W, Frutiger Y, et al. 2000. Thioredoxin a putative oncogene product is overexpressed in gastric carcinoma and associated with increased proliferation and increased cell survival. Hum. Pathol. 31:475— 81 Hirota K, Matsui M, Iwata S, Nishiyama A, Mori K, Yodoi J. 1997. AP—1 transcriptional activity is regulated by a direct association between thioredoxin and Ref-1. Proc. Natl. Acad. Sci. USA 94:3633-38 Nakamura H, Bai J, Nishinaka Y, Ueda S, Sasada T, et al. 2000. Expression of thioredoxin and glutaredoxin, redox-regulating proteins, in pancreatic cancer. Cancer Detect. Prev. 24:53-60 Kawahara N, Tanaka T, Yokomizo A, Nanri H, Ono M, et al. 1996. Enhanced coexpression of thioredixin and high mobility group protein | genes in human hepatocellular carcinoma and the possible association with decreased sensitivity to cisplatin. Cancer Res. 56:5330-33 Makino Y, Yoshikawa N, Okamoto K, Hirota K Yodoi J, et al. 1999. Direct association with thioredoxin allows redox regulation of glucocorticoid receptor function. J. Biol. Chem. 274:3182-88 Ema M, Hirota K, Mimura J, Abe H, Yodoi J, et al. 1999. Molecular mechanisms of transcription activation by HLF and HIFlalpha in response to hypoxia: their stabilization and redox signal-induced interaction with CBP/p300. EMBO J. 18:1905-14 Ueno M, Masutani H, Jun Arai R,

Yamauchi A, Hirota K, et al. 2000. Thioredoxin-dependent redox regulation of p53-mediated p21 activation. J. Biol. Chem. 274:35809-15 89. Imamoto N, Kamei Y, Yoneda Y. 1998. Nuclear transport factors: function, behavior and interaction. Eur. J. Histochem. 42:9-20

THIOREDOXIN 90. Martin H, Dean M. 1991. Identification of a thioredoxin-related protein associated with plasma membranes. Biochem. Biophys. Res. Commun. 175:123-28 Oe Wollman EE, Kahan A, Fradelizi D. 1997. Detection of membrane associated thioredoxin on human cell lines. Biochem. Biophys. Res. Commun. 230:602-606 Oye, Wakasugi N, Tagaya Y, Wakasugi A, Mitsui M, Maeda

M, et al. 1990.

Adult T-

cell leukemia—derived factor/Thioredoxin produced by both human T-lymphotropic virus type 1 and Epstein-Barr virustransformed lymphocytes, acts as an autocrine growth factor and synergized with interleukin-1 and interleukin-2. Proc. Natl. Acad. Sci. USA 87:8282-86

93) Ericson ML, Horling J, Wendel-Hansen V, Holmgren A, Rosen A. 1992. Secretion of thioredoxin after in vitro activation of human B cells. Lymphokine Cytokine Res. 11:201—207

94. Rubartelli A, Bajetto A, Allavena G, Wollman E, Sitia R. 1992. Secretion of thiore-

doxin by normal and neoplastic cells through a leaderless secretory pathway. J. Biol. Chem. 267:24161-64

., Hori

K,

Katayama

M,

Sato

N,

Ishii

K, Waga S, Yodoi J. 1994. Neuroprotection by glial cells through adult T cell leukemia-derived factor/human thioredoxin (ADF/TRX). Brain Res. 652:304-10

96. Rubartelli A, Bonifaci N, Sitia R. 1995. High rates of thioredoxin secretion correlate with growth arrest in hepatoma cells. Cancer Res. 55:675-80

ile Rubartelli A, Sitia R. 1991. Interleukin IB and thioredoxin are secreted through a novel pathway of secretion. Biochem. Soc. Trans. 19:255-59 98. Oblong JE, Berggren M, Gasdaska PY, Powis G. 1994. Site-directed mutagenesis of. active site cysteines in human thioredoxin produces competitive inhibitors of human thioredoxin reductase and elimination of mitogenic properties of thioredoxin. J. Biol. Chem. 269:11714—20

287

99 . Tagaya Y, Maeda Y, Mitsui A, Kondo N, Matsui H, et al. 1989. ATL-derived fac-

tor (ADF) an IL-2 receptor/Tac inducer homologous to thioredoxin: possible involvement of dithiol-reduction in the IL-2 receptor induction. EMBO J. 8:757-64

100. Schenk H, Vogt M, Droge W, SchulzeOsthoff K. 1996. Thioredoxin as a potent costimulus of cytokine expression. J. Immunol. 156:765—71 101. Yoshida S, Katoh T, Tetsuka T, Uno K, Matsui

N, Okamoto

T. 1999.

Involve-

ment of thioredoxin in rheumatoid arthritis: its costimulatory roles in the TNFalpha-induced production of IL-6 and IL8 from cultured synovial fibroblasts. J. Immunol. 163:351—58

OZ

Hashimoto S$, Matsumoto K, Gon Y, Furuichi S Maruoka S, et al. 1999. Thiore-

doxin negatively regulates p38 MAP kinase activation and IL-6 production by tumor necrosis factor-alpha. Biochem. Biophys. Res. Commun. 258:443-47

103. Natsuyama S, Noda Y, Yamashita M, Nagahama Y, Mori T. 1993. Superoxide dismutase and thioredoxin restore defective p34°4? kinase activation in mouse two-cell block. Biochim.

Biophys. Acta

1176:90-94 104. Spector A, Yan G-Z, Huang R-RC, McDermott MJ, Gascoyne PRC, Pigiet V. 1988. The effect of H,O, upon thioredoxin-enriched lens epithelial cells. J. Biol. Chem. 263:4984—-90 105. Bjornstedt M, Xue J, Huang W, Akesson B, Holmgren A. 1994. The thioredoxin and glutaredoxin systems are efficient electron donors to human plasma glutathione peroxidase. J. Biol. Chem. 269:29382-84 106. Hogarth L, English M, Price L, Wyllie R, Pearson ADJ, Hall AG. 1996. The effect of treatment with high dose melphalan, cisplatin or carboplatin on levels of glutathione in plasma, erythrocytes, mononuclear cells and urine. Cancer Chemother. Pharmacol. 37:479-85

288

POWIS

® MONTFORT

107. Berntsen H, Lonning PE, Ekse D, Netteland B, Johannessen DC, et al. 1998. Influence of treatment with aminoglutethimide on plasma and red-bloodcell glutathione status in breast cancer patients. Cancer Chemother. Pharmacol. 42:46-52 108. Cotgreave IA, Gerdes RG. 1998. Recent trends in glutathione biochemistry— glutathione-protein interactions: a molecular link between oxidative stress and cell proliferation? Biochem. Biophys. Res. Commun. 242:1-9 109. Chae HZ, Chung SJ, Rhee SG. 1994. Thioredoxin-dependent peroxide reductase from yeast. J. Biol. Chem. 269:27670-78 110. Jin D, Zoon HJ, Rhee SG, Jeang K. 1997. Regulatory role for a novel human thioredoxin peroxidase in NF-kappaB activation. J. Biol. Chem. 272:30952-61 Nay Shau H, Butterfield LH, Chiu R, Kim A. 1994. Cloning and sequence analysis of candidate human nautral killer-enhancing factor genes. Immunogenetics 40:129-—34 12s Pahl P, Berger R, Hart I, Chae HZ, Rhee SG, Patterson D. 1995. Localization of TDPX1, a human homologue of the yeast thioredoxin-dependent perox-

116.

WAGE

118.

119.

120.

.

ide reductase gene (TPX) to chromosome

13q12. Genomics 26:602-606 Prosperi M, Ferbus D, Karezinski I, Goubin G, 1993. A human cDNA corresponding to a gene overexpressed during cell proliferation encodes a product sharing homology with amoebic and bacterial proteins. J. Biol. Chem. 268:11050—56 114. Hirotsu S, Abe Y, Okada K, Nagahara

MS

N, Hori

H, et al. 1999.

Crystal

.

123%

struc-

ture of a multifunctional 2-Cys peroxiredoxin heme-binding protein 23 kDa/ proliferation-associated gene product. Proc. Natl. Acad. Sci. USA 96:12333-38 Illy. Kang SW, Chae HZ, Seo MS, Kim K, Baines IC, Rhee SG. 1998. Mammalian peroxidoxin isoforms can reduce hydrogen peroxide generated in response to

124.

growth factors and tumor necrosis factora. J. Biol. Chem. 273:6297-—302 Watabe S, Kohno H, Kouyama H, Hiroi T, Yago N, Nakazawa T. 1994. Purification and characterization of a substrate protein for mitochondrial ATP-dependent protease in bove adrenal cortex. J. Biochem. 115:648-54 Matsumoto A, Okado A, Fujii T, Fujii J, Egashira M, et al. 1999. Cloning of the peroxiredoxin gene family in rats and characterization of the fourth member. FEBS Lett. 443:246—50 Haridas V, Ni J, Meager A, Su J, Yu G-L, etal. 1998. Cutting edge: TRANK, a novel cytokine that activates NF-kappaB and c-Jun N-terminal kinase. J. Jmmunol. 161:1-6 Kang SW, Baines IC, Rhee SG. 1998. Charaterization of a mammaliam peroxiredoxin that contains one conserved cysteine. J. Biol. Chem. 273:6303-11 Lyu MC, Rhee SG, Chae HZ, Lee TH, Adamson MC, et al. 1999. Genetic mapping of six mouse peroxiredoxin genes and fourteen peroxiredoxin related sequences. Mamm. Genome 10:1017-19 Yanagawa T, Ishikawa T, Ishii T, Tabuchi K, Iwasa S, et al. 1999. Peroxiredoxin I expression in human thyroid tumors. Cancer Lett. 145:127-—32 Ichimiya S, Davis JG, O’Rourke DM, Katsumata M, Greene MI. 1997. Murine thioredoxin peroxidase delays neuronal apoptosis and is expressed in areas of the brain most susceptible to hypoxic and ischemic injury. DNA Cell Biol. 16:311-21 Shau H, Gupta RK, Golub SH. 1993. Identification of a natural killer enhancing factor (NKEF) from human erythroid cells. Cell. Immunol. 147:1-11 Shau H, Kim AT, Hedrick CC, Lusis AJ, Tompkins C, et al. 1997. Endogenous natural killer enhancing factor-B_ increases cellular resistance to oxidative stresses. Free Rad. Biol. Med. 22:497— 507

THIOREDOXIN 125. Prosperi M, Apiou F, Dutrillaux B, Goubin G. 1994. Organization and chromosomal assignment of two human PAG gene loci: PAGA encoding a functional gene and PAGB a processed pseudogene. Genomics 19:236-41 AX). Zhang P, Liu B, Kang SW, Seo MS, Rhee SG, Obeid LM. 1997. Thioredoxin peroxidase is a novel inhibitor of apoptosis with a mechanism distinct from that of Bcl-2. J. Biol. Chem. 272:30615-18 127e Ruppitsch W, Meisslitzer C, HirschKauffmann M, Schweiger M. 1998. Over-

expression of thioredoxin in fanconi anemia fibroblasts prevents the cytotoxic and DNA damaging effect of mitomycin C and diepoxybutane. FEBS Lett. 422:99102

128. Prosperi M, Ferbus D, Rouillard D, Goubin G. 1998. The pag gene product, a physiological inhibitor of c-abl tyrosine kinase is overexpressed in cells entering S phase and by contact with agents inducing oxidative stress. FEBS Lett. 423:39-44 129} Stadtman TC 1996 Selenocysteine. Annu. Rev. Biochem. 65:83—-100 130. Berggren MM, Mangin JF, Gasdaska JR, Powis G. 1999. Effect of selenium on rat thioredoxin reductase activity: increase by supranutritional selenium and decrease by selenium deficiency. Biochem. Pharmacol. 57:187—93 U3ak Jordan A, Reichard P. 1998. Ribonucleotide reductases. Annu. Rev. Biochem. 67:71-98 13%, Hopper S, Iurlano D. 1983. Properties of a thioredoxin purified from rabbit bone marrow which fails to serve as a hydrogen donor for the homologous ribonucleotide reductase. J. Biol. Chem. 258:13453-57 133% Mau B, Powis G. 1992. Inhibition of cellular thioredoxin reductase by diaziquone and doxorubicin. Biochem. Pharmacol. 43:1621-26 134. Silverman RB, Nandi DL. 1988. Reduced thioredoxin: a possible physiological cofactor for vitamin K epoxide reductase.

289

Further support for an active site disulfide. Biochem. Biophys. Res. Commun. 155:1248-54

1335). Soute BAM, Groenen-van Dooren MCL, Holmgren

A,

Lundstrom

J, Vermeer

C. 1992. Stimulation of the dithioldependent reductases in the vitamin K cycle by the thioredoxin system: strong synergistic effects with protein disulphideisomerase. Biochem. J. 281:255—59 136. Preusch PC. 1992. Is thioredoxin the physiological vitamin K epoxide reducing agent? FEBS Lett. 305:257-S9 137e Gilmore TD, Koedood M, Piffat KA, White DW 1997. Rel/NF-kappaB/ IkappaB proteins and cancer. Oncogene 11: 1367-78 138. Moore BE, Bose HRJ. 1988. Expression of the v-rel oncogene in reticuloendotheliosis virus-transformed fibroblasts. Virology 162:377-87 i399). Tolendano MB, Leonard WJ. 1991. Modulation of transcription factor NFkappaB binding activity by oxidationreduction in vitro. Proc. Natl. Acad. Sci. USA 88:4328-32 140. Hayashi T, Ueno Y, Okamoto T. 1993. Oxidoreductive regulation of nuclear factor kappaB: Involvement of a cellular reducing catalyst thioredoxin. J. Biol. Chem. 268:11380-88 141. Galter D, Mihm S, Droge W. 1994. Distinct effects of glutathione disulfide on the nuclear transcription factors kappaB and the activator protein-1. Eur. J. Biochem. 221:639-48 142. Sorachi K, Sugie K, MaekawaN, Takami M, Kawabe T, et al. 1992. Induction and

function of FcERII on YT cells: possible role of ADF/thioredoxin in FcERII expression. Immunobiology 185:193—206

143. Matthews

JR,

Wakasugi

N,

Virelizier

J-L, Yodoi J, Hay RT. 1992. Thioredoxin

regulates the DNA binding activity of NFkappaB by reduction of a disulphide bond involving cysteine 62. Nucl. Acids Res. 20:3821-—30

290

144.

POWIS

® MONTFORT

Ghosh G, Van Duyne G, Ghosh S, Sigler

153.

145.

Muller GW, Rey FA, Sodeoka M, Verdine

154.

147.

148.

149.

150.

151.

Freemerman

Xanthoudakis

S, Miao

GG,

Curran

T.

1994. The redox and DNA-repair activities of ref-1 are encoded by nonoverlapping domains. Proc. Natl. Acad. Sci. USA

GL, Harrison SC. 1995. Structure of the NF-kappaB homodimer bound to DNA. Nature 373:311-17 146.

Abate C, Patel L, Rauscher III FJ, Cur-

ran T. 1990. Redox regulation of fos and jun DNA-binding activity in vitro. Science 249:1157-61

PB. 1995. The structure of NFkappaB p50 homodimer bound to a kappaB site. Nature 373:303-10

AJ, Gallegos A, Powis G.

91:23-27

1999, NF-kappaB transactivation is increased but not involved in the proliferative effects of thioredoxin overexpression in MCF-7 breast cancer cells. Cancer Res. 59:4090-94

155. Huan Y, Domann FE. 1998. Redox modulation of AP-2 DNA binding activity in vitro. Biochem. Biophys. Res. Commun. 249:307-12 156. Hayashi S, Hajiro-Nakanishi K, Makino

Schenk H,Klein M,ErdbriiggerW, Droge

Y, Eguchi H, Yodoi J, Tanaka H. 1997.

W, Schulze-Osthoff K. 1994. Distinct effects of thioredoxin and antioxidants on the activation of transcription factors NFkappaB and AP-1. Proc. Natl. Acad. Sci. USA 91:1672-76 Chakraborti PK, Garabedian MJ, Yamamoto KR, Simons SS Jr. 1992. Role of cysteines 640, 656, and 661 in steroid binding to rat glucocorticoid receptors. J. Biol. Chem. 267:11366—73 Makino Y, Okamoto K, Yoshikawa N, Aoshima M, Hirota K, et al. 1996. Thioredoxin: a redox-regulating cellular cofactor for glucocorticoid hormone action. Cross talk between endocrine control of stress response and cellular antioxidant defense system. J. Clin. Invest. 98:2469— 77 Grippo JF, Tienrungroj W, Dahmer MK, Housley PR, Pratt WB. 1983. Evidence that the endogenous heat-stable glucocorticoid receptor-activating factor is thioredoxin. J. Biol. Chem. 258:13658—64 Tanaka H, Makino Y, Okamoto K. 1999, Thioredoxin in the endocrine response to

Functional modulation of estrogen receptor by redox state with reference to thioredoxin as a mediator. Nucl. Acids Res. 25:4035—40 Akamatsu Y, Ohno T, Hirota K, Kagoshima H, Yodoi J, Shigesada K. 1997. Redox regulation of the DNA binding activity in transcription factor PEBP2. The roles of two conserved cysteine residues. J. Biol. Chem. 272:14497-500 Lane DP. 1994. p53 and human cancers. Brit. Med. Bull. 50:582—99 Banin S, Moyal L, Shieh S, Taya Y, Anderson CW, et al. 1998. Enhanced phosphorylation of p53 by ATM in response to DNA damage. Science 281:1674—77 Canman CE, Lim DS, Cimprich KA, Taya Y, Tamai K, et al. 1998. Activation of the ATM kinase by ionizing radiation and phosphorylation of p53. Science 281:1677-79 Hainaut P, Milner J. 1993. Redox modulation of p53 conformation and sequencespecific DNA binding in vitro. Cancer Res. 53:4469-73

stress. Vitam. Horm. 57:153-75 152. Meyer M, Schreck R, Baeuerle PA. 1993. HO, and antioxidants have opposite effects on activation of NF-kappa B and AP-1 in intact cells: AP-1 as secondary antioxidant-responsive factor. EMBO J.

12:2005-15

157.

158.

159.

160.

161.

162. Bischoff JR, Casso D, Beach D. 1999, Human pS3 inhibits growth in Schizosaccharomyces pombe. Mol. Cell. Biol. 12:1405-11

163. Casso D, Beach D. 1996. A mutation in a thioredoxin reductase homolog suppresses p53-induced growth inhibition

THIOREDOXIN in the fission yeast Schizosaccharomyces pombe. Mol. Gen. Genet. 252:518-29

164. Pearson GD, Merrill GF. 1998. Deletion of the Saccharomyces cerevisiae TRR1 gene encoding thioredoxin reductase inhibits p53-dependent reporter gene expression. J. Biol. Chem. 273:5431-34 165. Mermill GF, Dowell P, Pearson GD. 1999. The human 53 negative regulatory domain mediates inhibition of reporter gene transactivation in yeast lacking thioredoxin reductase. Cancer Res. 59:3175-79 166. Jayaraman L, Murthy KG, Zhu C, Curran

Molecular cloning and characterization of the mouse apoptosis signal-regulating kinase 1. Biochem. Biophys. Res. Commun. 239:905-10

7S,

174. Iwata S, Hori T, Sato N, Hirota K, Sasada T, et al. 1997. Adult T cell leukemia (ATL)-derived factor/human thioredoxin prevents apoptosis of lymphoid cells induced by L-cystine and glutathione depletion: possible involvement of thiolmediated redox regulation in apoptosis caused by pro-oxidant state. J. Immunol. 158:3108-17

tification of redox/repair protein Ref-1 as a potent activator of p53. Genes Develop. 11:558-70 167. Saitoh M, Nishitoh H, Fujii M, Takeda K, et al. 1998. Mammalian

Lee K, Murakawa M, Nishida E, Tsubuki S, Kawashima S, et al. 1998. Proteolytic activation of MST/Krs, STE20-related

protein kinase, by caspase during apoptosis. Oncogene 16:3029-37

T, Xanthoudakis S, Prives C. 1997. Iden-

K, Tobiume

291

NS.

Nilsson J, Soderberg O, Nilsson K, Rosen

thioredoxin is a direct inhibitor of apoptosis signal-regulating kinase (ASK) 1. The EMBO 17:2596-606

A. 2000. Thioredoxin prolongs survival of B-type chronic lymphocytic leukemia cells. Blood 95:1420-26

168. Liu H, Nishitoh H, Ichijo H, Kyriakis JM. 2000. Activation of apoptosis signalregulating kinase 1 (ASK1) by tumor necrosis factor receptor-associated factor 2 requires prior dissociation of the ASK1 inhibitor thioredoxin. Mol. Cell. Biol. 20:2198—208 169. Isowa N, Yoshimura T, Kosaka S, Liu M,

176. Baker A, Payne CM, Briehl MM, Powis

Hitomi S, et al. 2000. Human thioredoxin

attenuates hypoxia-reoxygenation injury of murine endothelial cells in a thiol-free condition. J. Cell. Physiol. 182:33—40

G.

1997.

Thioredoxin,

a gene

found

overexpressed in human cancer, inhibitis

apoptosis in vitro and in vivo. Cancer Res.

57:5162-67 . McDonnell TJ, Korsmeyer SJ. 1991. Progression from lymphoid hyperplasia to high-grade malignant lymphoma in mice transgenic for the t(14:18). Nature 349:254-56 178. Russel M, Model P. 1986. The role of thioredoxin in filamentous phage assem-

170. Nishiyama A, Ohno T, Iwata S, Mat-

bly. Construction, isolation, and charac-

sui M, Hirota K, et al. 1999. Demon-

terization of mutant thioredoxins. J. Biol. Chem. 261:14997-5005

Eke

stration of the interaction of thioredoxin with p40phox, a phagocyte oxidase component, using a yeast two-hybrid system. Immunol. Lett. 68:155-59 Ichijo H, Nishida E, Irie K, ten Dijke P, Saitoh M, et al. 1997. Induction of apoptosis by ASK1,

a mammalian MAPKKK

that activates SAPK/JNK and p38 signaling pathways. Science 275:90-94

LD:

TAS). Huber HE, Russel M, Model P, Richardson CC. 1986. Interaction of mutant thioredoxin of Escherichia coli with the gene 5 protein of phage T7: The redox capacity of thioredoxin is not required for stimulation of DNA polymerase activity. J. Biol. Chem. 261:15006—5012 180. Tonissen K, Wells J, Cock I, Perkins A,

Tobiume K, Inage T, Takeda K, Enomoto

Orozco C, Clarke F. 1993. Site-directed

S,

mutagenesis

Miyazono

K,

Ichijo

H.

1997.

of

human _ thioredoxin.

292

POWIS

® MONTFORT Y, Miyata M, et al. 1998. Elevated serum

Identification of cysteine 74 as critical to its function in the “early pregnancy factor” system. J. Biol. Chem. 268:22485— 89

181. Wang J, Kobayashi M, Sakurada K, Ima-

189.

mura M, Moriuchi T, Hosokawa M. 1997.

Possible roles of an adult T-cell leukemia (ATL)-derived factor/thioredoxin in the drug resistance of ATL to adriamycin. Blood 89:2480-87

190.

182. Kunkel M, Kirkpatrick DL, Johnson G, Johnson JI, Paul KT, et al. 1997. Cell line-

directed screening assay for inhibitors of thioredoxin reductase signalling as potential anticancer drugs. Anti-Cancer Drug Des. 12:659-70 183. Yokomizo A, Ono M, Nanri H, Makino

Lot:

Y, Ohga T, et al. 1995. Cellular levels of

thioredoxin associated with drug sensitivity to cisplatin, mitomycin C, doxorubicin and etoposide. Cancer Res. 55:4293-96

184. Yamada M, Tomida A, Yoshikawa H, Taketani Y, Tsuruo T. 1996. Increased expression of thioredoxin/adult T-cell leukemia-derived factor in cisplatinresistant human cancer cell lines. Cancer Res. Clin. 2:427-32 185. Yamada M, Tomida A, Yoshikawa H,

192:

193:

Taketani Y, Tsuruo T. 1997. Overexpres-

sion of thioredoxin does not confer resistance to cisplatin in transfected human ovarian and colon cancer cell lines. Cancer Chemother. Pharmacol. 40:3137 186. Sasada T, Iwata S, Sato N, Kitaoka Y, Hirota K, et al. 1996. Redox control of resistance to cis-diamminedichloroplatinum (iI) (CDDP): protective effect of human thioredoxin against CDDP-induced cytotoxicity. J. Clin. Invest. 97:2268-76 187. Sasada T, Nakamura H, Ueda S, Sato N, Kitaoka Y, et al. 1999. Possible involvement of thioredoxin reductase as well as thioredoxin in cellular sensitivity to cis-diamminedichloroplatinum (II). Free Rad. Biol. Med. 27:504-14 188. Miyazaki K, Noda N, Okada S, Hagiwara

194.

195;

196.

level of thioredoxin in patients with hepatocellular carcinoma. Biotherapy 11:277— 88 Aota M, Matsuda K, Isowa N, Wada H, Yodoi J, Ban T. 1996. Protection against reperfusion-induced arrhythmias by human thioredoxin. J. Cardiovasc. Pharmacol. 27:727-32 Yokomise H, Fukuse T, Hirata T, Olikubo K, Go T, et al. 1994. Effect of recombinant human adult T cell leukemia-derived factor on rat lung reperfusion injury. Respiration 61:99-104 Okubo K, Kosaka S, Isowa N, Hirata T, Hitomi S, et al. 1997. Amelioration of ischemia-reperfusion injury by human thioredoxin in rabbit lung. J. Thor. Cardiovasc. Surg. 113:1-9 Yagi K, Liu C, Bando T, Yokomise H, Inui K, et al. 1994. Inhibition of reperfusion injury by human thioredoxin (adult T-cell leukemia-derived factor) in canine lung transplantation. J. Thor. Cardiovasc. Surg. 108:913-21 Takagi Y, Horikawa F, Nozaki K, Sugino T, Hashimoto N, Yodoi J. 1998. Expression and distribution of redox regulatory protein, thioredoxin during transient focal brain ischemia in the rat. Neurosci. Lett. 251:25-28 Takagi Y, Mitsui A, Nishiyama A, Nozaki K, Sono H, et al. 1999. Overexpression of thioredoxin in transgenic mice attenuates focal ischemic brain damage. Proc. Natl. Acad. Sci. USA 96:4131—36 Shibuki H, Katai N, Kuroiwa_ S, Kurokawa T, Yodoi J, Yoshimura N. 1998. Protective effect of adult T-cell leukemia-derived factor on_ retinal ischemica-reperfusion injury in the rat. Investig. Opthalmol. Vis. Sci. 39:1470-77 Matsui M, Oshima M, Oshima H, Takaku K, Maruyama T, et al. 1996. Early embryonic lethality caused by targeted disruption of the mouse thioredoxin gene. Devel. Biol. 178:179-85

THIOREDOXIN

197: Fujii S, Nanbu

Y, Konishi

I, Mori

T,

Masutani H, Yodoi J. 1991. Immunohistochemical localization of adult T-cell leukaemia-derived factor, a human thiore-

doxin homologue, in human sues. Virchows Arch. Pathol. Histopathol. 419:317-26

fetal tisAnat. &

198. Hotta M, Tashiro F, Ikegami H, Niwa H, Ogihara T, et al. 1998. Pancreati beta

cell-specific expression of thioredoxin, an antioxidative and antiapoptotic protein, prevents autoimmune and streptozotocininduced diabetes. J. Exp. Med. 188:1445— Sil

199: Hawkins HC, Blackburn EC, Freedman RB. 1991. Comparison of the activities of protein disulphide-isomerase and thioredoxin in catalyxing disulphide isomerization in a protein substrate. Biochem. J. 275:349-53

200. Fernando

MR,

Nanri

H, Yoshitake

S,

Nagata-Kuno K, Minakami S. 1992. Thioredoxin regenerates proteins inactivated by oxidative stress in endothelial cells. Eur. J. Biochem. 209:917-22

201. Bhat GB, Iwase K, Hummel BCW, Walfish PG. 1989. Kinetic characteristics of a thioredoxin-activated rate hepatic and renal low-K,,, iodothyronine 5’-deiodinase. Biochem. J. 258:785—92 202. Flamigni F, Marmiroli S, Caldarera CM, Guarnieri C. 1989. Involvement of thiol transferase- and thioredoxin-dependent systems in the protection of ‘essential’ thiol groups of ornithine decarboxylase. Biochem. J. 259:111-15 203. Freedman RB, Hirst TR, Tuite MF. 1994. Protein disulphide isomerase: building bridges in protein folding. Trends Biochem. Sci. 19:331-36 204. Boniface JJ, Reichert LE Jr. 1990. Evidence for a novel thioredoxin-like catalytic property of gonadotropic hormones. Science 247:61—64 205. Bertini R, Howard OM, Dong HF, Oppenheim JJ, Bizzarri C, et al. 1999. Thioredoxin, a redox enzyme released in

infection and inflammation,

293

is a unique

chemoattractant for neutrophils, monocytes, andT cells. J. Exp. Med. 189:178389

206. Nakamura H, DeRosa S, Roedere M, Anderson

MT,

Dubs

JG, et al. 1996. EI-

evation of plasma thioredoxin levels in HIV infected individuals. Int. Immunol. 8:603-11

207. Nakamura H, Vaage J, Valen G, Padilla CA, Bjornstedt M, Holmgren A. 1998.

Measuremnts of plasma glutaredoxin and thioredoxin in healthy volunteers and during open-heart surgery. Free Rad. Biol. Med. 24:1176-86

208. Das KC, White CW. 1998. Detection of thioredoxin in human serum and biological samples using a sensitive sandwich ELISA with digoxigenin-labeled antibody. J. Immunol. Methods 211:9-20 209. Kogaki H, Fujiwara Y, Yoshiki A, Kitajima S, Tanimoto T, et al. 1996. Sensi-

tive enyme-linked immunosorbent assay for adult T-cell leukemia-derived factor and normal value measurement. J. Clin. Lab. Anal. 10:257-61 210. Kumagi S. 1998. ADF/thioredoxin as an indicator of oxidative stress. Jpn. J. Clin. Pathol. 46:574-80 ZA. Wakita H, Yodoi J, Masutani H, Toda K, Takigawa M. 1992. Immunohistochemical distribution of aduit T-cell leukemiaderived factor/thioredoxin in epithelial components of normal and pathologic human skin conditions. J. Invest. Dermatol. 99:101-7 Zs White CR, Brock TA, Chang LY, Crapo J, Briscoe P, et al. 1994. Superoxide and per-

oxynitrite in atherosclerosis. Proc. Natl. Acad. Sci. USA 91:1044—48

Zils Patel JM, Zhang J, Block ER. 1996. Nitric oxide-induced inhibition of lung endothelial cell nitric oxide synthase via interaction with allosteric thiols: role of thioredoxin in regulation of catalytic activity. Amer. J. Respir. Cell Mol. Biol. 15:41019

294

POWIS

® MONTFORT

214. Zhang J, Li YD, Jawaharlal MP, Block ER. 1998. Thioredoxin overexpression prevents NO-induced reduction of NO synthase activity in lung endothelial cells. Amer. J. Physiol. 275:L288-93 215. Takagi Y, Gon Y, Todaka T, Nozaki K, Nishiyama A, et al. 1998. Expression of thioredoxin is enhanced in atherosclerotic plaques and during neointima formation in rat arteries. Lab. Invest. 78:957—66 216. Morton H, Rolfe BE, Cavanagh AC. 1982. Early pregnancy factor: biology and clinical significance. In Pregnancy Proteins, ed. JG Grudzinskas, B Teisner, M Seppala, pp. 391-405, Sydney: Academic PANT Hori K, Hirashima M, Ueno M, Matsuda M Wage S, et al. 1993. Regulation of eosinophil migration by adult T cell leukemia-derived factor. J. Immunol. 151:5624-30 218. Tagaya Y, Wakasugi H, Masutani H,

223:

224.

225.

226.

Nakamura H, Iwata S, et al. 1990. Role of

ATL-derived factor (ADF) in the normal

and abnormal cellular activation: involvement of dithiol related reduction. Mol. Immunol. 27:1279-89 FANE), Koyanagi H, Wakasugi N, Yoshimatsu K, Takashima Y, Yodoi J, et al. 1995. Role of ADF/TRX and its inhibitor on the release of major basic protein from human eosinophils. Biochem. Biophys. Res. Commun. 213:1140-47 220. Saito I, Shimuta M, Terauchi K, Tsubota K, Yodoi J, Miyasaka N. 1996. Increased expression of human thioredoxin/adult T

cell leukemia-derived factor in Sjogren’s syndrome. Arthrit. Rheum. 39:773-82 22M

PEN

Lovell MA, Chengsong X, Gabbita SP. Markesbery WR. 2000. Decreased thioredoxin and increased thioredoxin reductase levels in alzheimer’s disease brain. Free Rad. Biol. Med. 28:418—27 Masutani H, Naito M, Takahashi K, Hattori T, Koito A, et al. 1992. Dysregulation of adult T-cell leukemia-derived factor (ADF)/thioredoxin in HIV infection: loss of ADF high-producer cells in

227.

.

lymphoid tissues of AIDS patients. AIDS Res. Hum. Retrovir. 8:1707-—15 Newman GW, Balcewicz-Sablinska MK, Guarnaccia JR, Remold HG, Silberstein DS. 1994. Opposing regulatory effects of thioredoxin and eosinophil cytotoxicityenhancing factor on the development of human immunodeficiency virus 1. J. Exp. Med. 180:359-63 Gallegos A, Raffel J, Bhattacharyya AK, Powis G. 2000. Increased expression of thioredoxin in human pnmary and metastatic colon cancer. Amer. Assoc. Cancer Res. 41:189 Oblong JE, Chantler EL, Gallegos A, Kirkpatrick DL, Chen T, et al. 1994. Reversible inhibition of human thioredoxin reductase activity by cytotoxic alkyl 2imidazolyl disulfide analogues. Cancer Chemother. Pharmacol. 34:434-8 Powis G, Kirkpatrick DL, Angulo M, Gasdaska JR, Powell MB. et al. 1996. Thioredoxin redox signalling: A novel target for anticancer drug development. Anticancer Drugs 7(Suppl. 3):121-6 Bootman MD, Pay GF, Rick CE, Tones MA. 1990. Two sulphonated dye compounds which compete for inosital 1,4,5trisphosphate binding to rat liver microsomes: effects on 5-phosphatase activity. Biochem. Biophys. Res. Commun. 166: 1334-39 Guerri C, Godfrey W, Grisolia S. 1976. Protection against toxic effects of formaldehyde in vitro, and of methanol or formaldehyde in vivo, by subsequent administration of SH reagents. Physiol. Chem. Phys. 8(6):543-—550

229% Schlager JJ, Hoerl BJ, Riebow J, Scott DP, Gasdaska P, et al. 1993. Increased NADPH: (quinone-acceptor) oxidoreductase (DT-diaphorase) activity is asso-

ciated with density dependent growth inhibition of normal but not transformed cells. Cancer Res. 53:1338-42 230. Watson JA, Rumsby MG, Wolowacz RG. 1999. Phage display identifies

THIOREDOXIN

thioredoxin and superoxide dismutase as novel protein kinase C-interacting proteins: Thioredoxin inhibits protein kinase C-mediated phosphorylation of histone. Biochem. J. 343:301-S

2ei

Nishiyama A, Matsui M, Iwata S, Hirota K, Masutani H, et al. 1999. Identification

of thioredoxin-binding protein-2/vitamin

295

D (3) up-regulated protein | as a negative regulator of thioredoxin function and expression. J. Biol. Chem. 274:21645—50

DSoe Redl B, Merschak P, Abt B, Wojnar P. 1999. Phage display reveals a novel interaction of human tear lipocalin and thioredoxin which is relevant for ligand binding. FEBS Lett. 460:182-86

ee ee _

_

ae

me

7

-

cat:

; ser

tT iW

>

.

ee®

i» ==gls

ine

ma

-theet

Pull

OP)

&

a

ai ®

ae

am as

_

7

4

&

ies

a

=

_ -

a

a

ee

7

}

———

+ -_

_

-

-

-

i

:

_

—

:

ae Caer pee So

aA

i

hd a

_

‘

=etSe

mee

a

»

&

gy)

as

ee

Annu. Rey. Pharmacol. Toxicol. 2001. 41:297-316

Copyright © 2001 by Annual Reviews. All rights reserved

REGULATION, FUNCTION, AND TISSUE-SPECIFIC EXPRESSION OF CYTOCHROME P450 CYP1B1 Graeme I Murray,! William T Melvin, William F Greenlee,’ and M Danny Burke? ' Department of Pathology, and * Department of Molecular and Cell Biology, University of Aberdeen, Aberdeen, AB25 2ZD, United Kingdom; e-mail: g.i.murray @ abdn.ac.uk; w.t.melvin @ abdn.ac.uk 3 Chemical Industries Institute of Toxicology, Research Triangle Park, North Carolina 27709-2137; e-mail: Wgreenlee @ciit.org 4 School of Pharmacy and Pharmaceutical Sciences, De Montfort University, Leicester, LEI 9BH, United Kingdom; e-mail: mdburke @ dmu.ac.uk

Key Words

Ah receptor, drug metabolism, liver, tumor

@ Abstract Cytochrome P450 CYP1B1 is a relatively recently identified member of the CYP] gene family. The purpose of this commentary is to review the regulatory mechanisms, metabolic specificity, and tissue-specific expression of this cytochrome P450 and to highlight its unique properties. The regulation of CYP1B1 involves a variety of both transcriptional and post-transcriptional mechanisms. CYP1B1 can metabolize a range of toxic and carcinogenic chemicals in vitro but in some cases with a unique stereoselectivity. Estradiol 4-hydroxylation appears to be a characteristic reaction catalyzed by human CYP1B1. However, there are considerable species differences regarding the regulation, metabolic specificity, and tissue-specific expression of this P450. In humans CYP1B1 is overexpressed in tumor cells, and this has important implications for tumor development and progression and the development of anticancer drugs specifically activated by CYP1B1.

INTRODUCTION Until fairly recently, the cytochrome P450 CYP/ gene family was thought to consist of a single subfamily containing two very well characterized members, CYP/A1 and CYPIA2.

However, in 1994 a new member of the human CYP/ gene family

was cloned from a tetrachloro-dibenzo-p-dioxin-treated human keratinocyte cell line (1). This formed part of a series of investigations to identify the differential expression of genes that results from exposure to dioxin (2). Nucleic acid and amino

acid sequence analysis indicated that CYP1B1 showed ~40% homology with both CYPIAI1 and CYP1A2. Despite this low degree of similarity with CYPIA1 and CYP1A2, this new P450 was assigned by the P450 nomenclature committee (3) to

0362-1642/01/0421-0297$14.00

297

298

MURRAY ET AL

a new CYP1 subfamily, CYP1B, which currently contains only the single member CYP1B1 (GenBank accession nos. U56438 and U03688). DNA hybridization

studies suggest that there is in fact only one member of the CYP/B subfamily (1,4). Although assigned to the CYP/ gene family, the CYP/B/ gene has several distinct properties that clearly separate this P450 from the other well-established members of the CYP1 family. The human CYP/B/ gene is located on chromosome 2p22-21 spanning approximately 12 kilobases (kb) of DNA and is composed of three exons and two introns (4). [It is interesting that an aryl hydrocarbon hydroxylase gene locus was assigned to human chromosome 2 as early as 1976 (5).] The mRNA is 5.2 kb and the open reading frame begins at the 5’ end of the second exon, in contrast to other P450s, which all begin in exon 1. The predicted protein sequence is 543 amino acids (Figure 1). The long 3’-untranslated region contains multiple polyadenylation sites (1). Analysis of the predicted amino acid sequence indicates that there are three potential furin cleavage sites commencing at amino acid 38, which suggests that there could be post-translational N-terminal processing. CYP1B1 is the largest known human P450, both in terms of mRNA size and number of amino acids, but it is paradoxically the simplest in gene structure.

371 gene 5° —

1044

exon 1 ~ exon 2

390 -346

3707 _

exon3

3032 +1

protein

ae 3

|

+1629

+4756

CYP1B1 1 MGTSL

543 KETCQ

Figure 1 The structure of the human CYP/B/ gene. The CYP/BI gene is localized to chromosome 2p21 and encompasses ~12 kb of genomic DNA. The gene is composed of three exons and two introns and gives rise to a 5.2-kb mRNA. Translation begins close to the 5’ end of exon 2 and continues into exon 3 to produce a predicted protein composed of 543 amino acids. There is a large 3-kb untranslated 3’ region that contains multiple polyadenylation sites. The first five N-terminal and last five C-terminal amino acids are indicated.

CYTOCHROME P450 CYP1B1

299

For comparison, both the human CYPJA/ and CYP]A2 genes are located on chro-

mosome 15 (CYPJA/ on chromosome 15q22-q24 and CYP/A2 on chromosome 15q22-qter), and each gene is composed of seven exons and six introns with mRNAs of 2.8 and 3.2 kb, respectively, and proteins of 512 and 515 amino acids, respectively. Nucleic acid sequences of orthologous forms of CYP1B1 have also been isolated from both mouse and rat cells. Mouse CYP1B1

(GenBank accession nos.

U03283 and X78445) has been cloned from a mouse embryo fibroblast-derived cell line (MCA-10T1/2) induced with dioxin (6), and a similar cell line (C3H 10T1/2) treated with dimethylbenzanthracene (7). The orthologous rat CYP1B1 (GenBank accession nos. X83867 and U09540) has been cloned from the adrenal

cortex of adult rats, which had been pretreated with either adrenocorticotrophic hormone [ACTH (8)] or dioxin (9). Each of these P450s has an mRNA of 5.2 kb

and a predicted protein of 543 amino acids. Although there is a high degree (>80%) of both nucleic acid and amino acid sequence similarity between human, mouse, and rat CYP1B1,

several studies (see below) indicate that there are con-

siderable species differences regarding the regulation, metabolic specificity, and tissue-specific expression of this P450 (1, 6, 8, 10). Both the mouse and rat CYP1B1 P450s were cloned in an extensive series of experiments by Jefcoate’s laboratory. These researchers initially identified products of polycyclic aromatic hydrocarbon metabolism in a mouse embryo fibroblastderived cell line [C3H 10T1/2 (11)] and in rat adrenal gland cells (12, 13) that could not be attributed to any of the then known P450s. This led to the purification and partial immunological and metabolic characterization of P450-like proteins from both mouse embryo fibroblasts (initially designated P450EF) and adult rat adrenal gland cells (initially designated P4SORAP) and finally to the identification of these P450s as the murine and rat forms of CYP1B1, respectively. Neither the mouse Cyp/b/ gene nor the rat CYP/B/ gene have as yet been assigned to a specific chromosome. The identification of this new member of the CYP/ gene family has already stimulated considerable research activity. There has been particular interest in the field of cancer research because studies have shown that CYP1B1 is capable of metabolizing a variety of putative human carcinogens and also that CYP1B1 shows increased expression in a wide range of human tumors (14). Thus, CYP1B1

appears to have potentially important roles in tumor development and progression, as a potential target for anticancer drugs and as a tumor biomarker. CYP1B1 is also important in toxicology, given its inducibility by dioxin and other Ah receptor agonists. This commentary reviews the pharmacology and toxicology of CYPIB1 as well as its regulation, metabolic specificity, and tissue-specific expression. The potential biological roles of CYP1B1 are also reviewed. The opportunity is taken to indicate areas of controversy and also outline possible directions for future research.

300

MURRAY ET AL

MECHANISMS OF REGULATION OF CYP1B1

Regulation of Human CYP1B1 The CYP/BI gene is transcriptionally activated by polycyclic aromatic hydrocarbons, which act via the Ah receptor complex. The most potent of these Ah receptor agonists for activating transcription of the CYP/B/ gene appears to be dioxin. Initial analysis of the upstream regulatory region of the human CYP/B/ gene has identified several positive and negative regulatory elements within the promoter region (15) that are structurally distinct from the promoters of the related CYP/A/ and CYP/A2 genes. Both basal regulatory sequences and dioxin-responsive elements have been identified in the 5’ regulatory region of the CYP/B/ gene (15). Maximum expression in a reporter construct required the presence of transcriptional enhancer elements in the regions from nucleotide (nt)-1022 to nt-835 and from nt-2300 to nt-1356. The region from nt-1022 to nt-835 also contains several dioxin-responsive elements (4). Four other regulatory regions have also been identified that are required for maximal activity of the CYP1B1 promoter (15). Constitutive expression of the human CYP/B/ gene also appears to involve at least one of the dioxin or xenobiotic-response elements (16). Multiple xenobiotic responsive elements have also been identified in the upstream regulatory region of the mouse Cyp/b/ gene (17). Studies examining both the basal and inducible expression of CYP1B 1 in human tumor-derived cell lines indicate that there is distinct cell-type-specific expression of CYPIB1. For example, in HepG2 hepatoma-derived cells, CYP1A1 (mRNA and protein) but not CYP1B1 was highly induced by dioxin, whereas in ACHN cells (a kidney tumor-derived cell line) the converse was true, with only CYP1B1 (mRNA and protein) being inducible by dioxin. In the ACHN cell line, there was also weak constitutive expression of CYP1B1 mRNA (18), but there was no detectable constitutive expression of CYP1B1 protein. In HepG2 cells, there was no constitutive expression of either CYP1A1 or CYP1B1 mRNA. Since both cell lines contained a functional Ah receptor and structurally intact CYPJA] and CYP1B1 genes, it was suggested that transcriptional repression was occurring due to the presence of an as-yet-uncharacterized repressor protein (18). The inducibility of

CYPIB1 and not CYPIA1 in ACHN cells has also been demonstrated by Spink

et al (19), who showed that

CYP1B1 induced by dioxin in ACHN

cells catalyzed

the 4-hydroxylation of estradiol. Expression of both immunoreactive CYP1A1 and CYP1B1 proteins has been

identified in dioxin-treated MCF-7 (a breast cancer—derived cell line) cells, whereas

in noninduced MCF-7 cells, there is a very low constitutive level of CYP1B1 but not CYPIA1

(20). In different breast cancer cell lines, there is a similar variable

expression of CYP1A1

and CYP1B1 mRNAs, both constitutive and induced in

response to dioxin (21). However, the expression of CYP1B1

mRNA, either con-

stitutive or induced, does not correlate with the expression of either Ah receptor mRNA or estrogen receptor mRNA, although “cross-talk” between Ah receptor and

CYTOCHROME P450 CYP1B1

301

estrogen receptor (21) signaling pathways has been shown to be important in the context of endocrine disrupters. Treatment of MCF-7 cells with tetradecanoylphorbol-13-acetate also results in differential changes in the induction of CYPIA1 mRNA and CYP1B1 mRNA by dioxin (22). There is enhanced expression of CYP1B1 mRNA as a result of treatment with tetradecanoylphorbol-13-acetate prior to induction with dioxin, perhaps related to the observed up-regulation of the Ah receptor and suppression of estrogen receptor mRNA and reduced expression of CYPIA1 mRNA. The mechanisms that enable distinct patterns of inducibility of CYP1A1 and CYP1B1 in different cell types by Ah receptor agonists remain to be fully elucidated. In addition, inducible expression of CYP1B1 may also involve non-Ah receptor-mediated mechanisms. In human breast cancer samples we have shown that there is also differential expression of CYP1B1

mRNA and CYP1A1

mRNA, with consistent constitutive

expression of CYP1B1 mRNA, but expression of CYP1A1 mRNA has been identified in only 25% of samples (23). We have found no expression of CYP1A2 mRNA in any of the samples examined. More recently we have also found similar differential expression of CYP1B1 mRNA and CYPIA1 mRNA in kidney tumors (24). CYP1B1 mRNA is expressed in all tumor samples, whereas CYPIA1 mRNA is present in 81% of kidney tumors. Increased constitutive expression of CYP1B1 protein has been shown in a variety of human tumors (14). The mechanisms controlling the expression of CYP1B1 in actual human tumor cells (as opposed to cultured cancer-derived cell lines) have not been elucidated, although transcriptional processing and translational stability could potentially contribute. The possibility of transcriptional up-regulation via the Ah receptor by an endogenous ligand is suggested by a report of activation of the Ah receptor complex in the absence of exogenous ligand; this mechanism could conceivably contribute to the expression of CYP1B1 (25,26). The potential for interaction with other receptor pathways exists, particularly in breast cancer with the estrogen receptor as indicated above. Although inducible expression of CYP1B1 by Ah receptor agonists, especially dioxin, has been well documented, other mechanisms possibly contribute to the regulation of CYP1B1, including non-Ah receptor-mediated pathways of transcriptional regulation and also post-transcriptional mechanisms, but these mechanisms have received much less study. The human CYP1B1 mRNA contains multiple polyadenylation sites, and it has been suggested that there is cell-type-specific alternative processing of the CYP1B1 mRNA, which may regulate the amount and/or the ability of the final transcript to be translated (16). The use of alternative processing may also influence the stability of the mRNA (16).

Regulation of Rodent CYP1B1 Both mouse and rat CYP1B1s are inducible by dioxin and are also regulated by cyclic AMP (cAMP)-mediated pathways (17,27). Steroidogenic transcription factor-1 motifs associated with cAMP-dependent transcriptional activation

302

MURRAY ET AL

of genes have recently been identified in the 5’ upstream regulatory sequences of the mouse Cyp/b/ gene (17). However, it is not known whether cAMP-dependent transcriptional regulation is involved in regulating human CYP1B1. Constitutive expression of CYP1B1 was absent in Ah-receptor-deficient mouse embryo fibroblasts (17), suggesting a key role for the Ah receptor in both constitutive and inducible expression of mouse CYP1B1. Recently, however, it has been suggested that aryl hydrocarbon nuclear translocator (ARNT) may function as a repressor of basal CYP1B1 expression independently of its interaction with the Ah receptor (28). In a mouse hepatoma-derived cell line (hepa-1), cells with a defective ARNT gene and no detectable immunoreactive ARNT protein had constitutive expression of CYP1B1 at a much higher level compared with wildtype cells with a normal ARNT gene and detectable immunoreactive ARNT protein. Moreover, those cells with a defective ARNT gene were also unresponsive to dioxin (28). A recent study has shown

that overexpression

of the 1

iso-

form of protein kinase C in C3H10T1/2 cells resulted in decreased expression of both CYP1B1 mRNA and protein (29), thus providing further evidence for the involvement of multiple cellular pathways in regulation of this form of P4S0. Inducible expression of CYP1B1 may also involve non-Ah receptor-mediated mechanisms. For example, in Ah receptor knockout mice, CYP1B1 mRNA (but not CYPIAI mRNA) is weakly inducible in liver by piperonyl butoxide (30). This finding appears to contrast, but not necessarily conflict, with the results of Zhang et al (17), who found that no induction by dioxin of a Cyp1B1 promoter construct occurred when the construct was transiently transfected into Ah receptor-deficient mouse embryo fibroblasts. These cells had been derived from Ah receptor knockout mice. In cultured cells derived from mouse embryo fibroblasts, it has been shown that CYP1B1 is a labile protein in the absence of substrate (benzanthracene) and that protein stabilization by substrate may contribute to the regulation of CYP1B1 (31). Once again it is not known whether a similar mechanism regulates human CYPIB1. If stabilization of human CYP1B1 protein does occur, then this could in part explain the apparent discordant results regarding CYP1B1 mRNA and protein levels in human tissues and tumors (see below). The transcriptional regulation of rat CYP1B1

has received less attention and,

in contrast to both human and murine CYP1B1, there have been no studies to map the regulatory regions of the rat CYP/B/ gene. Rat CYP1B1 is inducible by Ah receptor agonists and is also inducible by ACTH (27). ACTH treatment of cultured cells derived from either the zona fasciculata or zona glomerulosa of rat adrenal cortex resulted in an approximately fourfold increase in CYP1B1 expression as assessed by the increases in CYPIB1 mRNA, immunoreactive CYP1B1 protein, and stereoselective metabolism of 7, 12-dimethlybenz[aJanthracene. However, the response towards ACTH was greater than the corresponding response elicited by treatment with dioxin, which resulted in a twofold increase in CYP1B1 protein

but a sixfold increase in CYP1B1 mRNA. In contrast, angioten sin had no effect on

CYTOCHROME P450 CYP1B1

303

CYPI1B1 expression (27). Inhibition of CYP1B1 activity using l-ethynylpyrene, a suicide inhibitor of CYP1B1, had no effect on corticosteroid hormone synthesis.

It is interesting that, in contrast to P450s in other types of cultured cells, especially hepatocytes, high basal expression of CYP1B1 protein appears to be maintained in cultured rat adrenocortical cells, thus further highlighting the unique aspects of this P450. The effect of other steroid hormones, including estradiol and progesterone, on CYP1B1 expression has also been studied in cultured stromal fibroblasts isolated from rat mammary gland cells (32). A low level of immunoreactive CYP1B1 protein was detected in isolated rat mammary stromal fibroblasts cultured for 1—2 days without any pretreatment, whereas there was no significant CYP1B1 identified in isolated mammary epithelial cells prepared at the same time as the stromal fibroblasts (32). The expression of CYP1B1 was increased by exposure to benzanthracene and estradiol but not to cortisol, which suppressed the induction of CYP1B1 (32). The suppression of CYP1B 1 expression by cortisol in rat mammary fibroblasts appears to be caused by glucocorticoid receptor suppression of the Ah receptor complex (27, 33).

METABOLIC SPECIFICITY OF CYP1B1

Metabolic Reactions Catalyzed by Human CYP1B1 Most of the studies examining the metabolic specificity of human CYP1B1 have been performed using recombinant CYP1B1 expressed in either yeast (Saccharomyces cerevisiae) (34,35) or human (36) lymphoblastoid cells, because native

human CYP1B1 protein has as yet not been purified. Recently, human CYP1B1 has also been expressed at a high level in Escherichia coli (37) and insect cells (Gentest Corp., Woburn, Mass.). Initial experiments (34) to express CYP1B1 in S. cerevisiae resulted in a much higher expression of CYP1B1 when, instead of using the full coding sequence, the construct contained a deletion corresponding to an absence of the N-terminal amino acid residues 2-4 (34). Subsequent studies

have been performed with this 5’-modified CYP1B1 cDNA to achieve satisfactory expression of CYP1B1 protein. Highly purified, functionally active expressed human CYP1B1 has been obtained from E. coli using diethylaminoethy] cellulose and carboxymethyl-sepharose chromatography (37). Once again the highest level of expression was obtained from a construct that had a deletion of N-terminal amino acids 2-4 (37). Sufficient purified, expressed protein was obtained to determine carbon monoxide-binding spectra (37). - CYPIB1 is capable of activating a variety of putative human carcinogens as assessed by mutagenicity assays (35, 36). The Ames test was used to measure the ability of CYP1B1 expressed in S. cerevisiae to activate a wide range of polycyclic aromatic hydrocarbons and both aromatic and heterocyclic amines (35). CYP1B1 had higher activity than either CYP1A1 or CYP1A2 for the activation of various

304

MURRAY ET AL

promutagens and was particularly efficient in the activation of dibenzo [11, 12][a,]

pyrene 11,12 diol (35). The preferential metabolism of this latter compound (which is one of the most potent known procarcinogens) by CYP1B1 has been studied more extensively using V79 Chinese hamster cells containing expressed human CYP1B1

(38). This P450 showed distinct stereoselectivity towards this substrate,

preferentially forming dibenzo[a,l]pyrene 11,12-dihydrodiol-13,14 epoxides with a significantly higher turnover number than CYP1A1 (38). CYP1B1 is one of the most effective P450s in activating aflatoxin B1.to a mutagenic metabolite (36). The metabolism by human CYP1B1 expressed in S. cerevisiae of a dietary heterocyclic amine, 2-amino-1 methyl-6-phenylimidazo[4,5-b] pyridine, which has been implicated in the aetiology of breast and colon cancer, has been studied in detail (39). CYP1B1 metabolized this compound to several products including its mutagenic N2-hydroxylated derivative (39). CYP1B1 activity has also been characterized by using several xenobiotics that are considered model

substrates for individual P450s

(36,40). The activity of

human CYP1B1 in a number of standard probe reactions for clinically important P450 forms has been tested using CYP1B1 expressed in either human lymphoblastoid cells (36) or S. cerevisiae (40). CYP1B1

was active in the hydroxylation of

benzpyrene and bufuralol, the oxidation of caffeine and theophylline, and the Odealkylation of ethoxycoumarin, ethoxy-trifluoromethyl-coumarin, and ethoxyresorufin. It was less active in the metabolism of ethoxyresorufin than CYPIAI. However, CYP1B1 was inactive in the hydroxylation of chlorzoxazone, coumarin, diclofenac, lauric acid, nifedipine, p-nitrophenol, $-mephenytoin, and tolbutamide (40), which are all metabolic probes for other P450s. Interestingly, CYP1B1 hydroxylated testosterone when expressed either in lymphoblastoid cells (36) or E. coli (41) but not when expressed in S. cerevisiae (40), although only a low level of testosterone hydroxylation was observed. Dioxin has previously been observed to induce a novel pathway of estradiol metabolism in MCF-7 cells (42) involving the induction of the 4-hydroxylation of 17£-estradiol. This activity was not attributable to CYP1A1, which catalyzes the 2-hydroxylation of estradiol and was inhibited by an antibody raised to mouse CYPIB1. Subsequent studies with human CYP1B1 expressed in S. cerevisiae (34) or E. coli (37) have confirmed CYP1B1 as a highly selective estradiol 4hydroxylase that is inefficient in catalyzing the 2-hydroxylation of estradiol. Indeed, 4-hydroxylation of estradiol currently represents the most specific assay available to investigate the activity of human CYP1B1.

Metabolism by Murine CYP1B1 In contrast to human CYP1B1, recombinant mouse CYP1B1 expressed in E. coli does not act as an estradiol hydroxylase (10). The recombinant mouse CYP1B1 was produced by creating an expression construct in which the 20 N-terminal amino acids of mouse CYP1B1 were replaced with 8 amino acids of human GYPL7, because this facilitated expression of P450 in E. coli. The recombinant mouse

CYTOCHROME P450 CYPIB1

305

CYP1B1 also had a modified C terminus with an additional 12 amino acids, including 6 histidine residues to aid purification. Compared with the expression of human CYP1B1 in E. coli (37), only a relatively low level of expression of mouse CYP1B1 was achieved (10). Mouse CYP1B1 expressed in E. coli shows a stereoselective metabolism of 7,12-dimethylbenz[a]anthracene, which is similar to this same compound’s metabolism by CYP1B1 constitutively or inducibly present in murine C3H10T1/2 cells. The major metabolites produced by both native and expressed mouse CYP1B1

are the 3,4- and 10,1 1-dihydrodiols of 7,12-

dimethylbenz[a]anthracene. Expressed murine CYP1B1 also binds polycyclic aromatic hydrocarbons in a manner similar to that observed for native mouse CYP1B1 induced in C3H10T1/2 cells (10).

Rat CYP1B1 Metabolism Rat CYPIB1

present in adrenal cortex, testis, or ovary cells shows a stereos-

electivity towards the metabolism of 7,12-dimethylbenz[a]Janthracene, which is identical to that of mouse CYP1B1

isolated from embryo fibroblasts (12, 13). Rat

CYP1B1 expressed in S. cerevisiae shows a stereoselectivity in the metabolism of benzo[a]pyrene and benzo[a]pyrene-7,8-diol (43) that is similar to that shown by human CYP1B1 expressed either in yeast or baculovirus (43) cells, although rat CYP1B1

is more active than human CYP1B1

(43).

Metabolic Probes for CYP1B1 Activity The investigation of the metabolic specificity of CYP1B1 and the identification of functionally active CYP1B1 in biological tissues would be considerably enhanced by the identification of further specific metabolic probes with straightforward assay methods, especially single-step or direct fluorescent methods. The alkoxyresorufin group of substrates, especially ethoxyresorufin, has been used extensively to study and characterize other P450-associated monooxygenase activity. Recently in our laboratories, using a series of homologous alkoxyresorufin substrates, we compared the O-dealkylation activities of recombinant human CYPIB1, CYPIAI, and CYP1A2 expressed in lymphoblastoid cells. We found that CYP1B1 was less active overall than CYP1A1 and CYP1A2 but had a different substrate selectivity (Table 1). In particular, CYP1B1 was relatively inefficient in catalyzing the de-ethylation of ethoxyresorufin in comparison with CYP1A1. CYPIB1 expressed in E. coli had slightly higher activity towards ethoxyresorufin compared with expressed CYP1A2, but in both cases activity was significantly less than that of expressed CYP1A1 (37). The low overall rate of metabolism by CYP1B1 towards all of the alkoxyresorufin substrates tested means that none of them can be recommended as specific probes for CYP1B1. The study of the metabolic specificity of CYP1B1 will also be made much easier by the further development of selective chemical inhibitors (44,45) and the production of specific inhibitory antibodies.

306

MURRAY ET AL

TABLE 1 Alkoxyresorufin O-dealkylase activities of recombinant human CYP1A1, CYP1A2, or CYP1B1 expressed in human lymphoblastoid cells towards a series of alkoxyresorufin substrate P450

Methoxyresorufin

Ethoxyresorufin

Propoxyresorufin

Butoxyresorufin

Benzyloxyresorufin

CYPIAI

Me

9.0

CYPIA2

5.56

1.10

5.8

3.6

0.2

0.15

0.11

CYPIBI

0.01

0.23

0.18

0.014

0.04

0.14

“Values are nanomoles of resorufin formation per minute per nanomole of P450.

The P450s were supplied by Gentest

Corporation, Woburn, Mass.

CYP1B1 EXPRESSION IN NORMAL TISSUES AND TUMORS In reviewing the literature regarding the expression of CYP1B1 in individual tissues from different species, it became clear that a number of potentially misleading conclusions regarding the presence of CYP1B1 have been published. This is especially true for those studies that have used the reverse transcriptase polymerase chain reaction (RT-PCR) to detect CYP1B1 mRNA in human tissues (46-49). In all of the studies documenting the presence of CYP1B1 in cells or tissues, it is therefore important to distinguish between the identification of mRNAs, especially in RT-PCR studies, which may detect low amounts of CYP1B1 mRNA of uncertain biological significance and the presence of either immunoreactive or catalytically active protein. It is also important to distinguish between constitutive expression of CYP1B1 and the presence of CYP1B1 induced by different agents. Furthermore, inferences about the presence of CYP1B1 in human cells and tissues based on data from corresponding isolated cells or cell lines should not be made. Ideally the presence of CYP1B1 should be confirmed by demonstrating functionally active protein. However, this is difficult given the current paucity of specific substrates for CYP1B1 as described above.

CYP1B1 Expression in Normal Adult Human Tissues A low level of CYPIB1 mRNA has been detected by Northern blotting in several normal human tissues including kidney, liver, intestine, eye tissue, and brain (1, 35, 50,51). However, the results obtained from the studies by Sutter et al (1),

Shimada et al (35), and Rieder et al (51) cannot be considered definitive because, in each case, only a single set of commercially obtained tissue samples was investigated. No information was provided for these tissues regarding potential prior exposure in vivo to chemicals, drugs, or dietary factors capable of inducing the

expression of CYP1B1 nor whether RNA was obtained from single individuals or was pooled from several. Furthermore, no clinicopathological information was provided regarding the samples. CYP1B1 mRNA has also been identified by RT-PCR in several normal tissues and cell types, including liver (46), lymphocytes (46), endometrium (46, 47),

CYTOCHROME P450 CYP1B1

307

normal breast tissue (48), and lung epithelial cells obtained by bronchoalveolar lavage from both current smokers and nonsmokers (49). Quantitative RT-PCR

showed that CYP1B1 mRNA was present at a very low level in lung epithelial cells obtained from nonsmokers, whereas there was a higher level of CYP1B1 mRNA in

cells obtained from current smokers (49). However, in a similar study, CYP1B1 mRNA was not detected in cells (a mixture of lung epithelial cells and macrophages) obtained by bronchoalveolar lavage from nonsmokers, whereas CYP1B1 mRNA was detected in current smokers (52). CYP1B1 mRNA has also been detected by semiquantitative RT-PCR in nontumorigenic breast epithelial cell lines (22). However, the level of expression of CYP1B1 protein (as opposed to mRNA) in normal human tissues is another matter. It has been proposed, based particularly on the Northern blot results described above, that CYP1B1 has widespread expression in extra-hepatic tissues and hence is potentially important in tumor development in these tissues. Native CYP1B1

protein has not, however, been isolated

and purified from any normal human tissue, suggesting that constitutive expression of CYP1B1 protein in normal human tissues is at best at a very low level. (The failure to detect human CYP1B1 at all until 1994 also suggests that CYP1B1 protein is not significantly present in normal human tissues; otherwise, it would seem reasonable, given the extensive research into P450s over the previous 15—20

years, to expect that CYP1B1 would have been identified previously.) In contrast, nonhuman CYP1B1 has been partially purified in a form that is metabolically active from both mouse embryo fibroblasts (11) and rat adrenal gland (12). In both

cases protein purification, which required an extensive amount of cells or tissue, preceded cloning of the gene in the respective species. The first direct demonstration of CYP1B1 protein in human tissues was by our group, using polyclonal antibodies to CYP1B1 to identify it in breast cancer using immunoblotting (24) and in a variety of cancers using immunohistochemistry (14). However, we were unable to detect CYP1B1 protein in normal liver by either immunohistochemistry or immunoblotting of microsomal fractions, even though CYP1B1 mRNA is detectable in normal liver by Northern blotting (1,35) and RT-PCR (46; Gi Murray, MCE McFadyen, WT Melvin, unpublished observa-

tion). The absence of CYP1B1 protein in normal adult human liver microsomes has subsequently been confirmed by Edwards et al (53), who used immunoblotting with a polyclonal antibody raised against a peptide corresponding to the five C-terminal amino acid residues of the CYP1B1 protein. We have also failed to detect CYP1B1 in human liver by immunoblotting with a monoclonal antibody specific for CYP1B1 (54) and a highly sensitive chemiluminescence detection system. This antibody was produced using, as the immunogen, a synthetic peptide corresponding to amino acids 437 to 451 of the human CYP1B1 amino acid sequence (54). A very low and variable level of CYP1B1 protein has also recently been detected by immunoblotting of the microsomal fraction of cultured breast epithelial cells (55) and cultured breast stromal cells (56), which were prepared from non-

neoplastic breast tissue. However, it is not clear whether the presence of CYP1B1 in either cultured epithelial or stromal cells was representative of the cells in vivo or whether expression of CYP1B1 was induced either as a result of the isolation

308

MURRAY ET AL

of the cells or their culture over several days. No information regarding exposure of donors to potential inducing agents was presented. During the isolation and primary culturing of cells, there is a phase of proliferation, and it has been shown that CYP1B1 may be up-regulated during cellular proliferation in mouse embryo-derived fibroblasts (57). Our results showing the highly selective expression of CYP1B1 protein in tumors do not exclude the possibility of a low level of CYP1B1 in normal tissues under certain physiological conditions, although to date the normal tissues we have examined by using both monoclonal and polyclonal antibodies to CYP1B1 have not shown CYP1B1 expression. A low level of CYP1B1 in normal tissues would not preclude the utility of the overexpression of CYP1B1 in tumors for cancer cell growth, detection, and treatment.

The Presence of CYP1B1 in Human Fetal Tissues CYP1B1 mRNA has been identified in several human fetal tissues by Northern blotting, using a commercially available Northern blot (35). CYP1Bl1 mRNA was identified in all the fetal tissue samples stucied with the strongest signal present in fetal kidney, which suggests that CYP1B 1 is adevelopmentally regulated gene. For the reasons stated above for the results of Northern blotting of adult tissues, the findings obtained from this fetal study must be regarded as nondefinitive because only a single set of tissue samples was used, and the gestational ages of the fetuses were not provided. Consistent expression of CYP1B1 mRNA has also been identified by RT-PCR in several human fetal tissues (kidney, lung, adrenal gland, and brain) obtained from normal first and mid-trimester fetuses (46). In the

same study, CYP1B1 mRNA was also detected in fetal liver but only in half of the samples investigated, and it was not stated whether the expression of CYP1B1 mRNA in liver varied with gestational age. CYP1B1 mRNA was also consistently present in both first-trimester and full-term placental tissues (46). The presence of CYP1B1 mRNA in a variety of fetal tissues has led to the suggestion that CYP1B1 may have an important role in normal fetal development. This hypothesis has received recent support as a result of mutations in the CYP1B1 gene being linked with the development of a form of primary congenital glaucoma, which has an autosomal recessive pattern of inheritance (50, 58). Several mutations of the CYP1B1 gene were identified, all of which are present in the coding region of the gene (50, 59). The mutations included missense mutations, nonsense mutations,

frame-shift mutations, and deletions, and some of the mutations were predicted to lead to the formation of a truncated protein. The mechanism by which abnormal CYPIBI plays a role in the development of this form of glaucoma has not been established. CYP1B1-null

mice have recently been developed and, from initial reports,

are apparently normal with no pathological defect (60). However, it is not known whether the eyes were specifically examined to exclude glaucoma or whether mice are even susceptible to this type of abnormality.

CYTOCHROME P450 CYP1B1

309

Several polymorphisms of the CYP1B1 gene have also been identified (59, 61, 62), and these show ethnic variation in their allelic frequency (62). The CYPIB1 polymorphisms result in amino acid substitutions at codons 48 (Arg— Gly), 119 (Ala— Ser), 432 (Val—Leu), and 453 (Asn— Ser), respectively (61). The polymorphism at codon 432 has been correlated with both estrogen and progesterone receptor status in breast cancer, with a stronger association with progesterone receptor positivity (61). Polymorphic variants of CYP1B1 with amino acid substitutions at codons 432 and 453 have been expressed in E. coli, along with NADPH cytochrome P450 reductase, and the metabolic activity of the different variants have been assessed using a panel of procarcinogens and several steroid hormones, including estradiol, testosterone, and progesterone (41, 63). The CYP1B1

variant

with the valine-to-leucine substitution at amino acid 432 resulted in a fourfold increase in the K,,, of the 4-hydroxylation of estradiol, suggesting that amino acid substitution at codon 432 causes alteration in CYP1B1 binding of estradiol (63). The other CYP1B1 variants did not show any altered catalytic properties towards the 4-hydroxylation of estradiol (63). Interestingly, none of the variants including the codon 432 (Val—Leu) variant displayed altered activity towards ethoxyresorufin (63).

CYP1B1 Expression and Activity in Tumors Possibly the most significant finding regarding CYP1B1 expression is the high frequency of expression of CYP1B1 protein in various types of malignant tumors (14; Figure 2, see color insert) without expression in various corresponding types of normal tissues. The majority of these observations were made using immunohistochemistry, which showed that CYP1B1 was specifically localized to tumor cells. The presence of CYP1B1 protein in breast cancer was confirmed by immunoblotting (14). Significant expression of CYP1B1 in breast cancer is also consistent with the findings of Liehr & Ricci (64), who demonstrated increased estradiol 4-hydroxylase activity in microsomes prepared from breast cancer. This activity was not detected in normal breast tissue, where 2-hydroxylation of estradiol occurred. They also found increased 4-estradiol hydroxylation in fibroadenoma, a common type of benign breast tumor. The same group has also found estradiol 4-hydroxylation that was inhibited by an antibody to mouse CYP1B1 in uterine leiomyomas (65), a benign tumor derived from smooth muscle of the uterine wall. The studies by Liehr’s group of estradiol 4-hydroxylation in individual types of tumor currently represent the only demonstration of CYP1B1 activity in human tissues or tumors (as opposed to tumor cell lines). These results of increased CYP1B1 activity in tumors are consistent with the concept of overexpression of CYPIB1 in tumors. Further studies are required to examine the expression of CYP1B1 in different types of benign tumors, because it will be important to investigate the stage during tumor development and progression at which alterations of CYP1B1 expression occurs. These investigations are also important in determining the utility of CYP1B1 as an early-stage tumor marker.

310

MURRAY ET AL

CYP1B1 in Murine Cells and Tissues CYP1B1 has been purified from a C3H10T 1/2 mouse embryo fibroblast cell line after benzanthracene induction, with only a low level of constitutive CYP1B1 protein in these cells (11). Expression of CYP1B1 in C3H10T1/2 cells has been suggested to be associated with enhanced transformation of these cells (66). Cultured mouse

bone marrow stromal cells also contain a detectable constitutive level of CYP1B1 protein, which is active in the metabolism of 7,12-dimethylbenz|a]anthracene (67).

The level of constitutive expression of CYP1B1 is approximately fourfold higher in cultured bone marrow stromal cells than in C3H10T1/2 cells. A very low amount of immunoreactive CYP1B1 has also been detected in . several normal adult murine tissues including kidney and uterus, whereas no immunoreactive CYP1B1 was detected in lung, liver, and brain. However, CYP1B1 was induced in vivo by £-naphthoflavone in all of those tissues, with the highest level of induction observed in lung tissue (6). CYP1B1

mRNA

is inducible in

mouse liver by dioxin, whereas there is a very low level of constitutive expression of CYP1B1

mRNA in murine liver (68, 69).

CYP1B1 mRNA has a higher level of induction than CYP1A1 mRNA in two different strains of mice (CS7BL/6J

and DBA/2J), which have different allelic

forms of the Ah receptor (69). Both CYP1A1 and CYP1B1 showed a higher level of inducible expression in the mouse strain (C57BL/6J) with the higher affinity form of the Ah receptor, indicating that inducible expression of CYP1B1 is linked to the Ah receptor complex. In mice, a lower frequency of lymphomas induced by high-dose 7,12-dimethlybenz[aJanthracene has been found in CYP1B1 knockout mice compared with wildtype mice (60). This finding emphasizes the importance of CYP1B1 in tumor development and progression.

CYP1B1 in Rat Tissues CYPI1B1 is present constitutively in the cortex of rat adrenal gland, ovary, and testis (8,12, 13). It is present in testis and ovary at 10-fold- and 25-fold-lower concentrations, respectively, than in adrenal gland. Within the adrenal cortex, CYPIB1 appears to be present in cells of both the zona fasciculata and the zona glomerulosa (27). However, it has not been established if there is preferential localization of CYP1B1 to a particular zone of the adrenal cortex. Similarly, the cell type or types expressing CYP1B1 in Ovary and testis have not been identified although it is likely that CYP1B1 is present in cells involved in steroid hormone metabolism, that is, Leydig cells in the testis and luteal cells in the ovary. CYPIBI1 is not constitutively present in rat liver (8), but it is inducible in liver by dioxin (70,71) or -naphthoflavone (8). Induction of hepatic CYP1B1 occurs only after high doses of tetrachloro-dibenzo-p-dioxin (71). After chronic exposure to dioxin, there is a marked zonal distribution of CYPI1B1, with this

enzyme present mainly in hepatocytes of zone 3 (perivenular) of the liver acinus

MURRAY ET AL C-1

Figure 2 The immunohistochemical localization of CYP1B1 in individual types of human malignant tumors. The tumors have been immunostained with a monoclonal antibody to human CYPIB1. Sites of antibody binding were detected using a tyramide amplification system with peroxidase as the reporter molecule and were visualized using diaminobenzidine as the peroxidase substrate to give a brown reaction product (see 54 for details of CYPI!B1 antibody and immunohistochemical techniques). Each tissue section has been ~ counterstained with hematoxylin, which stains nuclei blue and allows visualization of morphological detail. In each type of tumor, there is strong CYPIBI immunoreactivity in tumor cells. Panels: a, stomach

cancer;

rhabdomyosarcoma

tumor

CYPIBI1

(malignant

in nontumor cells.

b, esophageal

of muscle).

cancer; c, ovarian cancer, and d,

There

is no

immunoreactivity

for

—

— s a

:

«

+:

ve.

:

CYTOCHROME P450 CYP1B1

311

as determined by immunohistochemistry (71). The intra-acinar distribution of CYP1B1 parallels the localization of CYP1A2 within the liver acinus. In rat liver cells in culture CYP1B1 mRNA is expressed constitutively in hepatic stellate cells and myofibroblasts but not in hepatocytes or Kupffer cells (72). There is no constitutive expression of CYP1B1 protein in several tissues including kidney, lung, and uterus (8).

CONCLUSIONS This review outlines current knowledge regarding the regulation, metabolic specificity and tissue-specific expression of CYP1B1, which appears in many respects to be a unique form of P450. Despite relatively intensive investigation, there remain many important areas to be studied. For example, it is important particularly in the context of toxicology and drug metabolism to establish the precise metabolic specificity of CYP1B1 towards substrates and to identify its capability for activating and detoxifying carcinogens and other toxic chemicals. Also important is its regulation by individual Ah receptor agonists and factors regulating the cell-typespecific expression of this P450. However, this classic toxicological approach to P450s runs the risk of underestimating the potentially important biological significance of CYP1B1. Possibly one of the most important findings to date is our own observation of enhanced expression of CYP1B1 protein in a variety of human cancers (14) and the suggestion that CYP1B1 may be a marker of tumorigenesis. Given the frequency of its occurrence in tumors, there may also be an important endogenous function for this enzyme. Future research into the role of CYP1B1 will need to elucidate its regulation in tumors and its metabolism of anticancer drugs. If CYP1B1 can be shown to have a distinct metabolic specificity, this could lead to the development of cytotoxic anticancer prodrugs designed to be selectively activated by CYP1B1 in tumors. ACKNOWLEDGMENTS

Research in the authors’ laboratories has been supported by the Medical Research Council UK, the Cancer Research Campaign UK, Chief Scientist Office of the Scottish Executive, and Association for International Cancer Research.

Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Sutter TR, Tang YM, Hayes CL, Wo YYP, Jabs EW, et al. 1994. Complete cDNA sequence of a human dioxin-inducible

mRNA identifies a new gene subfamily of

cytochrome P450 that maps to chromosome 2. J. Biol. Chem. 269:13092-99 2. Sutter TR, Guzman K, Dold KM,

Green-

1991. Targets for dioxin:

genes

lee WF.

312

MURRAY ET AL

for plasminogen activator inhibitor-2 and interleukin-18. Science 254:415-18 . Nelson

DR,

Koymans

L,

Kamataki

T,

Stegeman JJ, Feyereisen R, et al. 1996. P450 superfamily: update on new sequences, gene mapping, accession numbers and nomenclature. Pharmacogenetics 6:1-42 . Tang YM, Wo Y-Y, Stewart J, Hawkins AL, Griffins CA, et al. 1996. Isolation and char-

acterization of the human CYP1B1 J. Biol. Chem. 271:28324-30

gene.

. Brown S, Wiebel FJ, Gelboin HV, Minna

JD. 1976. Assignment of a locus required for flavoprotein-linked monooxygenase expression to human chromosome 2. Proc. Natl. Acad. Sci. USA 73:4628-32

. Savas U, Bhattacharyya KK, Christou M, Alexander DL, Jefcoate CR. 1994. Mouse cytochrome P-450EF, representative of a new 1B subfamily of cytochrome P-450s. J. Biol. Chem. 269:14905-11 . Shen Z, Liu J, Wells RL, Elkind MM. 1994. cDNA cloning, sequence analysis, and induction by aryl hydrocarbons of a murine cytochrome P450 gene, Cyp1B1. DNA Cell Biol. 13:763-69 . Bhattacharyya KK, Brake PB, Eltom SE, Otto SA, Jefcoate CR. 1995. Identification of a rat adrenal cytochrome P450 active in polycyclic hydrocarbon metabolism as rat CYPIB1. J. Biol. Chem. 270: 11595-602 . Walker NJ, Gastel JA, Costa LT, Clark GC, Lucier GW, Sutter TR. 1995. Rat CYPIB1:

an adrenal cytochrome P450 that exhibits sex-dependent expression in livers and kidneys of TCDD-treated animals. Carcinogenesis 16:1319-27

. Savas

U, Carstens

C-P, Jefcoate

CR.

1997. Recombinant mouse CYPI1B1 expressed in Escherichia coli exhibits selective binding by polycyclic hydrocarbons and metabolism which parallels C3HI10T1/2 cell microsomes but differs from human recombinant CYP1B1. Arch. Biochem. Biophys. 347:181—92

jie Pottenger LH, Christou M, Jefcoate CR. 1991. Purification and immunological characterization of a novel cytochrome P450 from C3H/10T1/2 cells. Arch. Biochem. Biophys. 286:488-97 12: Otto S, Marcus C, Pidgeon C, Jefcoate C. 1991. A novel adrenocorticotropininducible cytochrome P450 from rat adrenal microsomes catalyzes polycyclic aromatic hydrocarbon metabolism. Endocrinology 129:970—-82 HEE Otto S, Bhattacharyya KK, Jefcoate CR. 1992. Polycyclic aromatic hydrocarbon metabolism in rat adrenal, ovary, and testis microsomes is catalyzed by the same novel cytochrome P450 (P450RAP). Endocrinology 131:3067—76 14. Murray GI, Taylor MC, McFadyen MC, McKay JA, Greenlee WF, et al. 1997. Tumor-specific expression of cytochrome P450 CYP1B1. Cancer Res. 57:3026-31 1S: Wo Y-YP, Stewart J, Greenlee WF. 1997. Functional analysis of the promoter for the human CYP/B/ gene. J. Biol. Chem. 272:26702-7 16. Shehin SE, Stephenson RO, Greenlee WF. 2000. Transcriptional regulation of the human CYPIB1 gene: evidence for involvement of an aryl hydrocarbon receptor response element in constitutive expression. J. Biol. Chem. 275:6770-76 iby Zhang L, Savas U, Alexander DL, Jefcoate CR. 1998. Characterization of the mouse CypIB1 gene. J. Biol. Chem. 273:5174-83 18. Kress S, Greenlee WF. 1997. Cell specific regulation of human CYP/A/ and CYP/B1/ genes. Cancer Res. 57:1264—69 Lo: Spink DC, Spink BC, Cao JQ, Gierthy JF, Hayes CL, Li Y, etal. 1997. Induction of cytochrome P450 1B1 and catechol estrogen metabolism in ACHN human renal adenocarcinoma cells. J. Steroid Biochem. Mol. Biol. 62:223-32 20. Christou M, Savas U, Spink DC, Gierthy JF, Jefccate CR. 1994. Co-expression of human CYP1A1 and a human analog of cytochrome P450-EF in response to

CYTOCHROME P450 CYP1B1

2,3,7,8-tetrachloro-dibenzo-p-dioxin

in

the human mammary carcinoma-derived MCE-7 cells. Carcinogenesis 15:725-32 2A, Spink DC, Spink BC, Cao JQ, DePasquale JA, Pentecost BT, et al. 1998 Differential

313

(C3H10T1/2) cells by protein kinase C (PKC). Biochem. Pharmacol. 57:619-30

30. Ryu DY, Levie PA, Fernandez-Salguero P, Gonzalez FJ, Hodgson E. 1996. Piperony1 butoxide and acenaphthylene induce cytochrome P450 1A2 and 1B1 in aromatic hydrocarbon-responsive receptor knock out mouse liver. Mol. Pharmacol. 50: 443-46

expression of CYP1A1 and CYP1B1 in human breast epithelial cells and breast tumor cells. Carcinogenesis 19:291—98 22: Spink BC, Fasco FJ, Gierthy JF, Spink DC. 1998. 12-O-tetradecanoylphorbol-3acetate upregulates the Ah receptor and differentially alters CYP1B1 and CYP1A1 in MCF-7 breast cancer cells. J. Cell Biochem. 70:289-96 23: McKay JA, Melvin WT, Ah-See AK, Ewen

ail, Savas U, Jefcoate CR. 1994. Dual regulation of cytochrome P450EF expression via the aryl hydrocarbon receptor and protein stabilization in C3H/10T1/2 cells. Mol. Pharmacol. 45:1152-59 32; Christou M, Savas U, Schroeder S, Shen

SWB, Greenlee WF, et al. 1995. Expres-

X, Thompson T, et al. 1995. Cytochromes

sion of cytochrome P450 CYP1B1 cancer. FEBS Letts. 374:270-72

in breast

24. Cheung Y-L, Kerr AC, McFadyen MCE, Melvin WT, Murray GI. 1999. Differential expression of CYP1A1,

CYP1A2,

CYP1B1 in human kidney tumours. Cancer Lett. 139:199-205 PIS. Chang C-Y, Puga A. 1998. Constitutive activation of aromatic hydrocarbon receptor. Mol. Cell. Biol. 18:525—35 26. Heath Pagliuso S, Rogers WJ, Tullis K, Seidel SD, Cenijn PH, et al. 1998. Acti-

vation of the Ah receptor by tryptophan and tryptophan metabolites. Biochemistry 37:11508-15

ue Brake PB, Jefcoate CR. 1995. Regulation of cytochrome P4501B1 in cultured rat adrenocortical ceils by cyclic adenosine 3’,5’-monophosphate and 2,3,7,8tetrachlorodibenzo-p-dioxin.

Endocrinol-

ogy 136:5034-41

28. Eltom SE, Zhang L, Jefcoate CR. 1999. Regulation of cytochrome P-450 (CYP) 1B1 in mouse hepa-1 variant cell lines: a possible role for aryl hydrocarbon receptor nuclear translocator (ARNT) as a suppressor of CYP1B1 gene expression. Mol. Pharmacol. 55:594-604 2M), Ikegwuonu FI, Christou M, Jefcoate CR. 1999. Regulation of cytochrome P4501B1 (CYP1B1) in mouse embryo fibroblast

P450 CYP1A1 and CYPI1B1 in the rat mammary gland: cell-specific expression and regulation by polycyclic aromatic hydrocarbons. Mol. Cell Endocrinol. 115: 41-50 33: Brake PB, Zhang L, Jefcoate CR. 1998. Aryl hydrocarbon regulation of cytochrome P4501B1 in rat mammary fibroblasts: evidence for transcriptional repression by glucocorticoids. Mol. Pharmacol. 54:825-33 34. Hayes CL, Spink DC, Spink BC, Cao JQ, Walker NJ, Sutter TR. 1996. 176Estradiol hydroxylation catalyzed by human cytochrome P450 1B1. Proc. Natl. Acad. Sci. USA 93:9776-81 35) Shimada T, Hayes CL, Yamazaki H, Amin S, Hecht

SS, et al. 1996. Activation

of

chemically diverse procarcinogens by human cytochrome P-450 1B1. Cancer Res. 56:2979-84

36. Crespi CL, Penman

BW,

Steimer

DT,

Smith T, Yang CS, Sutter TR. 1997. Development of a human lymphoblastoid

cell line constitutively expressing human CYP1B1 cDNA: substrate specificity with model substrates and promutagens. Carcinogenesis 12:83-89

ae Shimada T, Wunsch RM, Hanna IH, Sutter TR, Guengerich FP, Gillam EM. 1998.

Recombinant

human

cytochrome

P450

314

MURRAY ET AL

1B1 expression in Escherichia coli. Arch. Biochem. Biophys. 357:111—20

38. Luch A, Coffing SL, Tang YM, Schneider A, Soballa V, et al. 1998. Stable expression of human cytochrome P450 1B1 in V79 Chinese hamster cells and metabolically catalysed DNA adduct formation of dibenz[a,l]pyrene. Chem. Res. Toxicol. 11:686-95

Bo) Crofts FG, Trickland PT, Hayes CL, Sutter TR. 1997. Metabolism of 2-amino1-methyl-phenylimidazo[4,5-b] pyridine (PhIP) by human cytochrome P4501B1. Carcinogenesis 18:1793—98 40. Shimada T, Gillam EMJ, Sutter TR, Strickland PT, Guengerich FP, Yamazaki H. 1997. Oxidation of xenobiotics by recombinant human cytochrome P450 1B1. Drug Metab. Dispos. 29:617—22 41. Shimada T, Watanabe J, Kawajiri K, Sut-

47.

48.

49.

ter TR, Guengerich FP, et al. 1999. Cat-

alytic properties of polymorphic human cytochrome P450 1B1 variants. Carcinogenesis 20:1607-13 42. Spink DC, Hayes CL, Young NR, Christou M, Sutter TR, et al. 1994. The effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin on

estrogen metabolism in MCF-7 breast cancer cells: evidence for induction of a novel 17-estradiol 4-hydroxylase. J. Steroid Biochem. Molec. Biol. 51:251—58 43. Kim JH, Stansbury KH, Walker NJ, Trush MA, Strickland PT, Sutter TR. 1998. Metabolism of benzo[a]pyrene and benzl[a]pyrene-7,8-diol by human

cytochrome P450 1Bl. Carcinogenesis 19:1847-S3 44. Shimada T, Yamazaki H, Foroozesh M, Hopkins NE, Alworth WL, Guengerich FP. 1998. Selectivity of polycyclic inhibitors for human cytochrome P450s 1A1, 1A2., and 1B1. Chem. Res. Toxicol. 11:1048—56 45, Doostdar H, Burke MD, Mayer RT. 2000. Bioflavonoids: selective substrates and inhibitors for cytochrome P450 CYP1A and CYPIB1. Toxicology 144:31-38 46. Hakkola J, Pasanen M, Pelkonene O,

50.

Hukkanen J, Evisalmi S, et al. 1998. Expression of CYP1B1 in human adult and fetal tissues and differential inducibility of CYP1B1 and CYP1AI1 by Ah receptor ligands in human placenta and cultured cells. Carcinogenesis 18:391—97 Hellmold H, Rylanddr T, Magnusson M, Reihnér E, Warner M, Gustafsson J-A. 1998. Characterization of cytochrome P450 enzymes in human breast tissue from reduction mammoplasties. J. Clin. Endocrinol. Metab. 83:886—95 Vadlamuri SV, Glover DD, Turner T, Sarkar MA. 1998. Regiospecific expression of cytochrome P4501A1 and 1B1 in human uterine tissue. Cancer Lett. 122:143-50 Willey JC, Coy EL, Frampton MW, Torres A, Apostolakos MJ, et al. 1997. Quantitative RT-PCR measurement of cytochromes P450 1A1, 1B1 and 2B7 microsomal epoxide hydrolase, NADPH oxidoreductase expression in lung cells of smokers and nonsmokers. Am. J. Respir. Cell Biol. 17: 114-24 Stoilov I, Akarsu AN, Sarfarazi M. 1997. Identification of three different truncating mutations in cytochrome P4501B1 (CYPI1B1) as the principal cause of primary congenital glaucoma (bupthalmos) in families linked to the GLC3A locus on chromosome 2p21. Hum. Mol. Genet. 6:64147 . Rieder CRM, Ramsden DB, Williams AC.

1998. Cytochrome P450 1B1 mRNA in the human central nervous system. J. Clin. Pathol. Mol. Pathol. 51:138-42 Nnbo. Pupari R, Savela K, Nurminen T, Hukkanen J, Raunio H, et al. 2000. Expres-

sion of CYP1A1, CYPIB1 and CYP3A, and polycyclic aromatic hydrocarbonDNA adduct formation in bronchoalveolar macrophages of smokers and non-smokers. Int. J. Cancer 86:610-16 . Edwards RJ, Adams DA, Watts PS, Davies

DS, Boobis AR. 1998. Development of a comprehensive panel of antibodies against

CYTOCHROME P450 CYP1B1

the major xenobiotic metabolising forms of cytochrome P450 in humans. Biochem. Pharmacol. 56:377-87

54, McFadyen MCE,

Breeman

S, Payne

S,

Stirk C, Miller ID, et al. 1999. Immuno-

histochemical localization of cytochrome P450 CYP1B1 in breast cancer with monoclonal antibodies specific for CYP1B1. J. Histochem. Cytochem. 47:1457-64

215): Larsen

MC, Angus WGR, Brake PB, Eltom SE, Sukow KA, Jefcoate CR. 1998.

Characterization of CYP1B1 and CYPI1A1 expression human mammary epithelial cells: role of aryl hydrocarbon receptor in polycyclic aromatic hydrocarbon metabolism. Cancer Res. 58:2366—74

56. Eltom SE, Larsen MC, Jefcoate CR. 1998. Expression of CYPIB1 but not CYPIA1 by primary cultured human mammary stromal fibroblasts constitutively and in response to dioxin exposure: role of the Ah receptor. Carcinogenesis 19:1437-44

SVE Alexander DL, Eltom SE, Jefcoate CR. 1997. Ah receptor regulation of CYP1B1 expression in primary mouse-derived cells. Cancer Res. 57:4498—506 58. Bejjani BA, Lewis RA, Tomey KF, Anderson KL, Dueker DK, et al. 1998. Muta-

tions in CYP1B1, the gene for cytochrome P4501B1, are the predominant cause of primary congenital glaucoma in Saudia Arabia. Am. J. Hum. Genet. 62:325—33 a) Stoilov I, Akarsu AN, Alozie I, Child A, Barsoum-Homsy M, et al. 1998. Sequence analysis and homology modelling suggests that primary congenital glaucoma on 2p21 results from mutations disrupting either the hinge region or the conserved core structure of cytochrome P4501B1. Am. J. Hum. Genet. 62:573-84 60. Buters JTM, Sakai S, Richter T, Pineau T, Alexander DL, et al. 1999. Cytochrome P450 CYP1B1 determines susceptibility to 7,12-dimethylbenz[aJanthracene-induced

lymphomas. Proc. Natl. Acad. Sci. USA

96:1977-82 61. Bailey LR, Roodi N, Dupont WD, Parl

315

FF. 1998. Association of cytochrome P450 1B1(CYP1B1) polymorphism with steroid hormone receptor status in breast cancer. Cancer Res. 58:5038-41

62. Inoue K, Asao T, Shimada T. 2000. Ethnicrelated differences in the frequency distribution of genetic polymorphisms in the CYPI1A1 and CYPIB1 genes in Japanese and Caucasian populations. Xenobiotica 30:285—95 63. Li DN, Seidel A, Pritchard MP, Wolf CR, Friedberg T. 2000. Polymorphisms in P450 CYPIB1 affect the conversion of estradiol to the potentially carcinogenic metabolite 4-hydroxyestradiol. Pharmacogenetics 10:343-53 64. Liehr JG, Ricci MJ. 1996. 4-Hydroxylation of estrogens as marker of human mammary tumors. Proc. Natl. Acad. Sci. USA 93:3294-96 65. Liehr JG, Ricci MJ, Jefcoate CR, Hannigan EV, Hokanson JA, Zhu BT. 1995. 4-

Hydroxylation of estradiol by human uterine myometrium and uterine microsomes: implications for the mechanisms of uterine tumorigenesis. Proc. Natl. Acad. Sci. USA 92:9220—24 66. Shen Z, Wells RL, Elkind MM. 1994. Enhanced cytochrome P450 (Cyp/b1) expression, aryl hydrocarbon hydroxylase activity, cytotoxicity, and transformation of C3H 10T1/2 cells by dimethylbenz(a)anthracene in conditioned medium.

Cancer Res. 54:4052-56 67. Heidel SM, Czuprynski CJ, Jefcoate CR. 1998. Bone marrow stromal cells constitutively express high levels of cytochrome P4501B1 that metabolize 7,12dimethylbenz[a]anthracene. Mol. Pharmacol. 54:1000—6 68. Ryu D-Y, Hodgson E. 1999. Constitutive expression and induction of CYP1B1 mRNA in the mouse. J. Biochem. Mol. Toxicol. 13:249-51 69. Abel J, Li W, Dohr O, Vogel C, Donat S. 1996. Dose-response relationship of cytochrome P4501b1 mRNA induction

316

MURRAY ET AL

by 2,3,7,8-tetrachlorodibenzo-p-dioxin of

C57BL/6J and DBA/2J mice. Arch. Toxicol. 70:510-13

70. Tristcher

AM,

Seacat

AM,

Yager

JD,

Groopman JD, Miller BD, et al. 1996. In-

creased oxidative DNA damage in livers of 2,3,7,8-tetrachlordibenzo-p-dioxin treated

rats but not ovariectomized rats. Cancer Lett. 98:219-25 1p, Walker NJ, Crofts FG, Li Y, Lax SK Hayes CL, et al. 1998. Induction and

localization

of cytochrome

P450

1B1

(CYP1B1) protein in the livers of TCDDtreated rats: detection using polyclonal antibodies raised to histidine tagged fusion proteins produced and purified from bacteria. Carcinogenesis 19:395-402 TL: Piscaglia F, Knittel T, Kobold D, BarnikolWatanabe S, Di Rocco P, Ramadori G. 1999. Cellular localization of hepatic cytochrome 1B1 expression and its regulation by aromatic hydrocarbons and inflammatory cytokines. Biochem. Pharmacol. 58:157-65

Annu. Rey. Pharmacol. Toxicol. 2001. 41:317-45 Copyright © 2001 by Annual Reviews. All rights reserved

PHYSIOLOGICAL FUNCTIONS OF CYCLIC ADP-RIBOSE AND NAADP as Ca.Lcrum MESSENGERS Hon Cheung Lee Department of Pharmacology, University of Minnesota, Minneapolis, Minnesota 55455; e-mail: leehc @tc.umn.edu

Key Words

cADPR, inositol trisphosphate, ADP-ribosyl cyclase, CD38,

Ca** stores, ryanodine receptor @ Abstract

Cyclic ADP-ribose (cADPR) and nicotinic acid adenine dinucleotide

phosphate (NAADP) are two Ca”* messengers derived from NAD and NADP, respectively. Although NAADP is a linear molecule, structurally distinct from the cyclic cADPR, it is synthesized by similar enzymes, ADP-ribosyl cyclase and its homolog, CD38. The crystal structure of the cyclase has been solved and its active site identified. These two novel nucleotides have now been shown to be involved in a wide range of cellular functions including: cell cycle regulation in Euglena, a protist; gene expression in plants; and in animal systems, from fertilization to neurotransmitter release and long-term depression in brain. A battery of pharmacological reagents have been developed, providing valuable tools for elucidating the physiological functions of these two novel Ca*+ messengers. This article reviews these recent results and explores the implications of the existence of multiple Ca”* messengers and Ca’* stores in cells.

INTRODUCTION The Ca’*-mobilizing activities of cyclic ADP-ribose (CADPR) and nicotinic acid adenine dinucleotide phosphate (NAADP) were first described in sea urchin eggs (1-4). These nucleotides have since been shown to be highly effective in releasing Ca?* from internal stores in a wide variety of cells from protist and plant to human (reviewed in (5, 6)). The cyclic structure of cADPR based on X-ray crystallography is shown in Figure 1 (see color insert) (3). This novel cyclic nucleotide is derived

from NAD (2). The site of cyclization is at the N1-position of the adenine ring, which is linked to the anomeric carbon of the terminal ribose (3). The nicotinamide

group of the precursor NAD is released during the cyclization reaction. In contrast, NAADP is a linear metabolite of NADP (Figure 1). It is produced by exchanging the nicotinamide group of NADP with nicotinic acid (7). Despite the fact that the two molecules have clearly different structures and precursors, it is remarkable 0362-1642/01/0421-0317$14.00

317

318

LEE

that they are, nevertheless, synthesized by the same class of enzymes, which includes ADP-ribosyl cyclase and its homologue, CD38 (7). The mechanisms of the Ca**+-releasing action of these two molecules are also completely independent and are different from that of inositol trisphosphate (IP;) as well. Since the last

comprehensive review of the mechanisms of Ca** signaling by these two Ca** messengers in 1997 (5), major advances have been made. The crystal structure of ADP-ribosyl cyclase has been determined, and its active site, as well as that of CD38, has been identified (8,9). The cADPR-sensitive Ca** channels have been

successfully reconstituted into lipid bilayers, and their characteristics indicate they are similar to the ryanodine receptor (10, 11). Elevation of cADPR levels following

cellular activation has now been documented in protist, plant, and animal cells (12-14).

Both cADPR and NAADP

have been shown to mediate Ca*+ signaling

ina wide variety of cells, and the functions they are involved in include fertilization, cell cycle regulation, insulin secretion, muscle contraction, nitric oxide signaling, neurotransmitter release, and activation of gene expression, as well as long-term

depression. Cells thus possess multiple Ca** stores and multiple Ca** messengers for activating them. Ca?* mobilization as a signaling mechanism is emerging as far more versatile and complex than the unitary view having IP; as the sole messenger. In the following sections, these recent advances will be reviewed with particular emphasis on the physiological functions of cADPR and NAADP.

STRUCTURES AND MECHANISMS OF ADP-RIBOSYL CYCLASE AND CD38 The presence of an enzymatic activity producing cADPR was first described in sea urchin egg homogenates (1, 2). This activity has since been shown to be ubiquitous [(15, 16) and reviewed in (17, 18)]. A soluble protein of about 30 kDa purified from

Aplysia ovotestis was named ADP-ribosyl cyclase (19). Two homologs, CD38 and CD157, both of which are mammalian antigens, have been identified (20— 22). Both share about 30% sequence identity with the cyclase and are capable of synthesizing cADPR from NAD. Additionally, CD38 can effectively hydrolyze cADPR

to ADP-ribose

(23-26). This class of enzymes is thus multifunctional,

capable of catalyzing more than one reaction. Although CD38 was first described as a lymphocyte surface antigen, it has since been found in many other tissues, including eye and brain [(27, 28) and reviewed in (29-31)]. Itis present not only on the surface of cells but also in intracellular organelles (32), including the nucleus (33-35). In liver nuclei, CD38 is localized by immuno-electronmicroscopy to the inner nuclear envelope (35). In addition to these two CD molecules, a soluble form of the ADP-ribosyl cyclase has recently been purified from bovine brain (36). The brain enzyme is about 30 kDa, smaller than CD38, which is about 45 kDa.

However, it is catalytically

similar to CD38, possessing both cADPR-synthesizing and cADPR-hydrolyzing

activities (36). Cyclase-like enzymes are present not only in animal cells but also

in protist.

A membrane-bound enzyme of about 41 kDa has been purified from

FUNCTIONS OF cADPR AND NAADP

319

Euglena, which is catalytically similar to the Aplysia cyclase, possessing only cADPR synthesizing but not the hydrolyzing activity (37). In addition to metabolizing cADPR, both the Aplysia cyclase and CD38 can also catalyze a transglycosylation reaction exchanging the terminal nicotinamide group of the substrate NADP with nicotinic acid and producing NAADP (7). Which catalytic path these enzymes take is determined by pH. In acidic pH and in the presence of nicotinic acid, NAADP is the predominant product, whereas at neutral or alkaline pH, the enzymes mainly cyclize NAD to produce cADPR (7). The unusual acidic pH dependency has led to the proposal that the NAADP-sensitive Ca** signaling may play a role in the endocytic pathway, which is composed of acidic organelles such as endosomes (5, 31). Thus, internalization of surface CD38

can expose the enzyme to acidic conditions conducive to NAADP synthesis. How a single enzyme can use two different substrates and produce two structurally distinct Ca** messengers has been investigated using the Aplysia cyclase. The enzyme is a homodimer both in solution (38) and as crystals (39, 40). The dimer, formed by two bean-shaped monomers in a head-to-head fashion, resembles a donut with a central cavity. Figure 2 (see color insert) shows the secondary structures in one monomer and the van der Waals surfaces in the other. Four large B-structures constitute most of the carboxyl domain, which is separated by a central cleft from the amino domain containing a-helixes (39). The monomer is a very

compact molecule packaged by five intramolecular disulfide bonds, three in the amino domain and two in the carboxy! domain. Cocrystallization of the cyclase with nicotinamide, a substrate for the exchange reaction, shows that the active site is at a pocket near the central cleft (Figure 2) (8). In order to cyclize NAD, the substrate must be bound in a folded conformation such that the two ends of NAD are close enough for linkage. The pocket structure of the active site serves nicely to mold the substrate into such a configuration. Two

iryptophan residues, Trp!*° and Trp”’, are present lining the rim of the active site pocket and Trp'° is separated from the bound nicotinamide by only 2.8 A (8). These two residues can serve to position the nicotinamide group and the adenine ring of NAD. Indeed, changing them to glycine using site-directed mutagenesis reduces the cADPR synthesizing activity by several thousand-fold (8). The next catalytic step is likely to be the release of the nicotinamide group from NAD and the formation of an ADP-riboysl intermediate. The catalytic residue

is shown by site-directed mutagenesis to be a glutamic acid, Glu'””. Changing it to even a conservative residue, such as aspartic acid or glutamine, renders the enzyme essentially inactive (8). As shown in Figure 2, the catalytic glutamate residue is situated deep inside the active site pocket. The ADP-ribosyl intermediate could be an oxocarbenium ion as has been proposed for similar enzymes (41-44). The anionic nature of the glutamate residue could serve to stabilize the cationic oxocarbenium intermediate. This function can also be shared by another nearby

glutamate residue, Glu’, which when mutated results in significant reduction in enzymatic activity (8). Alternatively, the intermediate could involve covalent ADP-ribosylation as has been shown in the case of CD38 (45), which is described below.

320

LEE

Intramolecular attack of the activated anomeric carbon of the ADP-ribosyl intermediate by the nitrogen at the 1-position of the adenine ring would result in cyclization, producing cADPR. If NADP is used as a substrate and nicotinic acid is present, nucleophilic attack of the intermediate by nicotinic acid would produce NAADP instead. At low pH, acidic residues around the active pocket are neutralized, allowing easier access of nicotinic acid to the active site and thus

facilitating the exchange reaction. The fact that the cyclase can use either NAD or NADP as substrate suggests that the 2’-position of the adeninyl ribose of the bound substrate is pointing outward, away from the active pocket; thus, substrate binding would only be minimally affected, even if a bulky phosphate is present at the 2’-position, as in NADP. Homology modeling of CD38 using the crystal coordinates of the cyclase indicates that the residues corresponding to the four critical residues of the cyclase described above are all clustered in a pocket that is similar to the active site of the

cyclase (9). Glu’”° of CD38, which corresponds to Glu!” of the cyclase, is also a catalytic residue because replacing it with even conservative residues inactivates

the enzyme. Likewise, Trp'*> and Trp!*’, corresponding to the two positioning tryptophan residues in the cyclase, are also critical for the enzymatic activity (9). Therefore, CD38 and the cyclase are homologous not only in their primary sequences but also structurally. The catalytic scheme described above for the cyclase can account for the hydrolytic function of CD38 as well. It suffices to further assume that the active site of CD38 has high affinity for cADPR itself, such that cADPR can bind and be converted catalytically to the ADP-ribosy] intermediate. Accessibility of the active site pocket of CD38 to water may also be higher, allowing ready reaction of the ADP-ribosyl intermediate with water to produce ADP-ribose. Indeed, detailed kinetic analyses support the idea that a single intermediate is responsible for all the reactions catalyzed by CD38 (46). The nature of the intermediate has recently been investigated using a substrate-inhibitor of the enzyme, arabinosyl 2’-fluoro-2’-deoxynicotinamide mononucleotide (45). Incubation of CD38 with the compound results in the release of nicotinamide from it and the covalent attachment of the rest of the compound to the enzyme. The residue

where the covalent modification occurs is identified as Glu2”° (45), identical to the

catalytic residue identified by site-directed mutagenesis (9). This result suggests

that CD38 may be ADP-ribosylated covalently at Glu” during normal cataly-_ sis when NAD is used as a substrate. The catalytic model described above thus presents a unified mechanism that accounts for all the known enzymatic properties of both the cyclase and CD38 (30).

PHARMACOLOGY OF CALCIUM SIGNALING MEDIATED BY cADPR AND NAADP The class of pharmacological reagents that can selectively modify the cADPRinduced Ca?* release is the modulators of the ryanodine receptor (RyR). Caffeine,

a stimulator of the RyR, at high concentrations releases Ca2+ from the same stores

FUNCTIONS OF cADPR AND NAADP

321

TABLE 1 Agonistic and antagonistic analogs of cADPR aee ee ee en eee Compound Effect Active concentration and other properties cADPR

Agonistic

ECsy ~ 18-48 nM (1, 83)

3-deaza-cADPR

Agonistic

ECs) © 1 nM, metabolically stable (58)

2’A-deoxy-cADPR

Agonistic

ECs) © 58 nM (176)

Cyclic aristeromycin diphosphate ribose

Agonistic

ECs) ~ 80 nM, partial agonist, metabolically stable (177)

Caffeine

Agonistic

ECsy ~ 5 mM, cell permeant (47)

8-NH>-cADPR

Antagonistic

iCsy ~ 10 nM, competitive (54)

7-deaza-8-Br-cADPR

Antagonistic

ICsy © 0.7 4M, cell permeant and metabolically stable (55)

8-N3-cADPR

Antagonistic

0.45 uM, photoaffinity labeling reagent (57)

8-Br-cADPR

Antagonistic

ICsy © 1 uM (54)

3’a-O-methyl-cADPR

Antagonistic

ICsy © 5 uM (176)

Ruthenium red

Antagonistic

50 uM (47)

Procaine

Antagonistic

1 mM (47)

as CADPR, and.at low concentrations, it potentiates the effect of cADPR (47-49). Inhibitors of RyR, such as ruthenium red, procaine and Me?*, likewise block the

action of cADPR (47, 48, 50). The effect of ryanodine itself is more complicated.

In some systems, such as sea urchin eggs, ryanodine releases Ca** from the same stores as CADPR (47-49). In other systems, such as pancreatic acinar and neurons,

ryanodine blocks the action of cADPR (51,52). This dual effect of ryanodine is likely related to its known biphasic action on the RyR channel (53). As shown in Table 1, the effective concentrations of the RyR modulators needed are relatively high. The discovery of ADP-ribosyl cyclase provides a versatile method to synthesize specific reagents and some of these are also listed in Table 1. In addition to NAD, the cyclase can cyclize a variety of NAD analogs to produce corresponding analogs of cADPR. The first series of useful CADPR analogs synthesized in this manner is the 8-derivatives, 8-NH,-cADPR and 8-Br-cADPR; both are

antagonists of cADPR (54). 8-NH -cADPR

is much more effective than the 8-Br-

derivative, but the latter is found to be cell permeant (55, 56). A more novel antag-

onist is 7-deaza-8-Br-cADPR, which is not only cell permeant but is also metabolically stable (55). Another antagonist that has useful properties is 8-N3,-cADPR, which is photoactive and has been used as an affinity probe for labeling the cADPRreceptor (57). Equally novel is the metabolically stable agonist, 3-deaza-cADPR, which is more than an order of magnitude more potent than cADPR itself (58). The finding that the cyclase can also catalyze a base-exchange reaction to produce NAADP provides a similarly versatile method for synthesizing analogs of NAADP. Various analogs of NADP and nicotinic acid are used as substrates for the exchange reaction to produce a series of active NAADP analogs (59, 60)

322

LEE TABLE 2

Active analogs of NAADP

—

Compound

Active concentration and other properties

NAADP

ICsy © 0.5 nM, ECsy ~ 20 nM (61, 62)

Etheno-NAADP

ICs) © 60 nM, ECsy ~ 5 uM, fluorescent (60)

Etheno-aza-NAADP

ICs, ~ 60 nM, ECsy ~ 2.5 4M, fluorescent (60)

3PSA-ADP

ICsq © 0.5 uM, ECsy ~ 3 uM (59)

Deamino-NAADP

—ICsp © 0.5 4M, ECsy © 10 uM (59)

and some of these are listed in Table 2. A very unusual property of NAADP is that it can induce effective desensitization of the Ca** release mechanism even at subthreshold concentrations (61,62). In this manner, NAADP can function as its

own specific antagonist with an ICs, in the 1 nM range. All of the active analogs produced so far show a similar property as listed in Table 2 (59,60). Two of these analogs, etheno-NAADP and etheno-aza-NAADP, are fluorescent (60) and potentially can be used to visualize the NAADP-receptor. In addition to the listed analogs, Bay K 8644 (ICs) = 30 uM) and L-type Ca** channel blockers, such as diltazem (ICs) = 7 4M), nifedipine (IC;, = 11 uM), and verapamil (ICs) = 21 4M), can also block the NAADP-induced Ca** release (63). The required concentrations of these Ca** antagonists are relatively high, however. Inhibitors of the cyclase and CD38 are also available. Nicotinamide is cell permeant and inhibits the cyclase activity with an IC5, of about 1.5 mM (64, 65). It does not, however, truly inhibit but forces the reversal of the reaction. The cyclase is a reversible enzyme such that NAD is produced from cADPR in the presence of a high concentration of nicotinamide (23). Indeed, under the condition where the cyclase appears inhibited, there is a rapid exchange of the added nicotinamide with that of the NAD (7). Nevertheless, the membrane permeability of nicotinamide makes it a useful reagent. A much more potent inhibitor of CD38 is available. Arabinosyl 2’-fluoro-2’-deoxynicotinamide mononucleotide inhibits the enzyme with a K; of 169 nM (45). The battery of reagents described above thus provides valuable tools for pharmacologically dissecting the Ca** signaling pathways mediated by cADPR and NAADP, which are crucial in understanding the physiological functions of these

novel Ca** messengers.

PHYSIOLOGICAL FUNCTIONS OF cADPR AND NAADP

Modulation of the Ryanodine Receptor The first indication that cADPR may be an endogenous modulator of the RyR

comes from pharmacological evidence as described above (47-49). However, the

FUNCTIONS OF cADPR AND NAADP

323

action of cADPR is found to be complex and requires accessory proteins. Photoaffinity labeling shows that cADPR binds specifically to a 140-kDa protein (57), smaller in size than the RyR. Another cofactor is calmodulin, which greatly enhances the Ca**t-releasing effect of cADPR (49, 66-69). Similar stimulation of

muscle RyR by calmodulin has also been reported (70, 71). More direct evidence that the target of cADPR is the RyR comes from reconstituting the cADPR-sensitive channels isolated from two species of sea urchin eggs into lipid bilayers. The reconstituted channels are Ca?* selective and show concentration dependence on cADPR (10, 11). The conductance of the channel from

one species is very similar to the mammalian RyR and its cADPR-dependency is found to require the presence of a dialyzable factor (11). The channel from the other species of sea urchin has somewhat smaller conductance, but the pharmacology is similar to the mammalian RyR nonetheless, being blocked by inhibitors of RyR and enhanced by caffeine. In both cases, the Ca**-conducting activity of the cADPR-channel is blocked by W7 and trifluoperazine, indicating it is dependent on calrnodulin (10,11), which is consistent with the results described above (49,

66-69). Both cardiac and skeletal RyR channels have been reported to be activated by cADPR (72-74). Similarly, cADPR is found to increase the Ca* sensitivity of the

third type of RyR isolated from diaphragm muscle and reconstituted into bilayers (75). Despite this body of positive evidence, the subject remains controversial because of negative results reported earlier (70, 76, 77). Part of the controversy is

related to the fact that, as described above, the action of cADPR on the RyR requires accessory proteins. Conditions used to isolate and reconstitute the mammalian RyR that are not optimal for retaining these necessary factors can yield variable results. Instead of relying on the difficult technique of reconstitution, a more informative approach is therefore to assess the effect of cADPR on RyR in more intact systems. It is thus convincing that in permeabilized clonal PC12 cells, the responsiveness to cADPR is strictly correlated with expression of the cardiac RyR, and cell lines devoid of RyR are not responsive (78). Even more direct evidence comes from

analyzing the Ca** sparks in intact cardiac myocytes, which are due to elemental Ca’* release from a single or a few RyR channels (79). The frequency of these

Ca’+ sparks is found to be enhanced by cADPR (80, 81). These results are further supported by measurements of ryanodine binding in detergent permeabilized parotid acini, which possess all three types of RyR (82). The distribution of the RyR can be visualized using BODIPY-ryanodine, a fluorescent analog, and both cADPR and ryanodine competitively inhibit the fluorescent staining in a similar manner (82). Analogous results are observed in

microsomes isolated from T-lymphocytes. Ryanodine binds specifically to these membranes and the binding is modulated by cADPR in a concentration-dependent manner (14). Taken together, these results that come from the use of more native

systems provide very convincing evidence that the target of cADPR is indeed the RyR.

324

LEE

Elevation of CADPR Levels Following Cell Activation cellular If cADPR is indeed a second messenger for Ca**, it is expected that its deassay first The agonists. primary by n levels would increase following activatio urchin sea in activity easing Ca?*-rel its on scribed for cADPR is a bioassay based in egg homogenates (1, 83). Using this assay, the endogenous levels of cADPR g occurrin naturally is cADPR ating several rat tissues are determined, demonstr enhance to caffeine using improved later (84). The sensitivity of the bioassay is the Ca2+-releasing activity of cADPR (48, 85). Another assay for cADPR is the radio-receptor assay based on the high affinity and specific binding of cADPR to egg microsomes (17,86). Antibodies against cADPR, raised in rabbits (87) and chickens (88), are now available and a highly sensitive radio-immuno assay has been developed. Human leukemic HL-60 cells are progenitors that can be induced to differentiate into different hemopoietic linages such as monocytes or granulocytes. Progressive accumulation of cellular cADPR as measured using the radioimmuno-assay is found to accompany monocytic differentiation induced by retinoic acid (87). The elevation of cADPR is specific for retinoic acid; other inducers of differentiation, such as cAMP, produce no such effect. Similar to retinoic acid, another cellpermeant agonist, nitric oxide, can likewise produce elevation of cADPR levels.

This has been observed in both neuroscretory PC12 cells (78) and hippocampal slices (89). As will be discussed later, the effect of nitric oxide is mediated through

activation of the ADP-ribosyl cyclase activity by the cGMP-dependent protein kinase. In pancreatic islets, stimulation of insulin secretion by glucose is also accompanied by cADPR elevation (90). In this case, it has been proposed that the action of glucose is due to glycolytic production of ATP, which in turn modulates the ADP-ribosyl cyclase activity (91). The agonists described above all have their primary targets inside the cells. Agonists that are cell impermeant and act through stimulation of surface receptors can also elevate cellular cADPR levels. This is the case in T-lymphocytes. Stimulation of the T-cell receptor/CD3 complex by specific antibody OKT3 induces increases in cytoplasmic Ca** concentration and a sustained elevation of cADPR levels (14). In intestinal longitudinal muscles, activation of the surface cholecystokinin receptor likewise stimulates ADP-ribosyl cyclase activity, which is blocked by an antagonist of the receptor (92). Similar stimulation of the cyclase is seen in | adrenal chromaffin cells following activation of the acetylcholine receptor (69). In all of these cases, Ca?* influx accompanies surface receptor activation. It has been proposed that the influx may be responsible for stimulating the cyclase activity (69,92). Alternative evidence suggests that the activation of the cyclase by surface receptors can be linked through a G-protein. Thus, treatment of membranes isolated from NG108 cells with a muscarinic agonist stimulates cADPR synthesis by 2- to 3-fold, which is inhibited by prior incubation with cholera toxin (93). A similar stimulatory effect by an adrenergic agonist is seen in membranes isolated from cardiac myocytes, which is mimicked by GTP-y-S and blocked by the toxin (94).

FUNCTIONS OF cADPR AND NAADP

325

Agonist-induced elevation of cellular cADPR content is seen not only in animal cells but in plants and Euglena, a protist, as well. Treatment of the subepidermal cells of Aurea hypocotyls with abscisic acid, a hormone that regulates plants response to environmental cues, rapidly elevates cellular cADPR, which in turn causally activates abscisic acid-specific genes (13). Likewise, in Euglena, treatment with vitamin B12, an obligatory growth factor, induces rapid elevation of cADPR, preceding the increases in DNA synthesis and cell number (12). From the results described above, an intriguing generalization is that many of the primary agonists that signal through the cADPR-pathway are cell permeant. They include a gaseous messenger, nitric oxide, metabolic factors such as glucose and vitamin B12, and agonists that target intracellular receptors, such as retinoic

acid. This is in contrast to the IP3-signaling pathway, which is predominantly linked to surface receptor activation. Cells may thus employ two separate Ca**+ messengers, CADPR and IP3, to distinguish, respectively, permeant and impermeant signals. This distinction is clearly not absolute. As described above, surface receptor stimulation can, in some cases, activate the cADPR-pathway as well.

Mediation of Ca** Signaling by Nitric Oxide The finding that the ADP-ribosyl cyclase in sea urchin eggs is activated by cGMP (95), which is known to be raised by nitric oxide, suggests that the cADPRand nitric oxide-signaling pathways are linked (96). The cGMP-dependent stimulation of the egg cyclase requires ATP (97) and is inhibited by protein kinase inhibitors (95), indicating that it is likely to be mediated by cGMP-dependent kinase phosphorylation. Exposing live sea urchin eggs to nitric oxide activates Ca** mobilization, which can be blocked by 8-amino-cADPR, a specific antagonist of the cADPR-receptor (54), nicotinamide, an inhibitor of the egg cyclase (65), and cGMP-dependent kinase inhibitors (98), providing direct evidence that the two

pathways are linked. That nitric oxide can activate Ca** mobilization via the cADPR-pathway has also been observed in clonal PC12-16A cells, an RyR-expressing neurosecretory cell line (78). This suggests that the connection between nitric oxide and cADPR

may have relevance in neuronal function (99). Indeed neurons, such as

dorsal root ganglion neurons, are known to be responsive to cADPR (100, 101). Metabotropic glutamate receptor activation in these neurons evokes transient depolarization, Ca?+-activated inward currents and rises in intracellular Ca2*, which

can be mimicked by intracellular photorelease of caged cGMP or caged cADPR (101). 8-amino-cADPR and a cyclase inhibitor, nicotinamide, inhibit the effects

of both caged analogs. These results are consistent with the functional presence of the cGMP/cADPR-signaling pathway in neurons. The possibility that this pathway may be involved in higher integrative functions of the nervous system, such as long-term depression (LTD), has been examined using embryonic cerebellar cultures derived from transgenic mice. The gene for the neuronal nitric oxide synthase in these mice is ablated. Although the

326

PEE

ar cells, no cADPR-mechanism is present and operational in the cultured cerebell effect of CADPR or cGMP on LTD has been observed ( 102).

norThe issue has recently been reexamined using hippocampal slices from of t Treatmen 89). (56, LTD ting investiga for mal rats, a more natural paradigm by blocked be can which LTD, induces slices with a nitric oxide donor readily ryanodine or cell-permeant antagonists of cADPR, 8-Br-cADPR (54), or 7-deaza8-Br-cADPR (55), indicating that the effect of nitric oxide is mediated by cADPR (56). In addition to nitric oxide, cGMP levels in brain slices can be raised biochemically using an inhibitor of the cGMP-selective phosphodiesterase in conjunction

with an inhibitor of cAMP-dependent protein kinase (89). This treatment increases the cADPR levels in the slices and induces LTD, which can be similarly blocked by ryanodine or cADPR antagonists. Inhibitors of cGMP-specific protein kinase likewise prevent LTD induction either by nitric oxide or by biochemical elevation of cGMP

(56, 89). These results are consistent with nitric oxide raising cGMP

levels in hippocampal slices, which in turn stimulates phosphorylation of ADPribosyl cyclase by a cGMP-dependent kinase. The resulting elevation of cADPR and mobilization of the cADPR-dependent Ca** stores then induce LTD. As is described in the section after the next, this nitric oxide/eADPR-signaling pathway is operational not only in animal cells, but also in plants.

Cell Cycle Signaling in Euglena Euglena is an unicellular protist. So far, it is the most evolutionarily primitive cell that has been reported to possess the cADPR-pathway. Its ADP-ribosyl cyclase is a 40-kDa membrane-bound protein that is catalytically similar to the soluble Aplysia cyclase (37). Unlike the mammalian CD38, the Euglena cyclase does not have the cADPR hydrolyzing activity. Euglena synchronized by light-dark cycle divide simultaneously at the onset of darkness. Its cyclase activity increases markedly correlating with DNA synthesis and prior to cell division (12). Cell division can also be synchronized by addition of vitamin B 12, an essential growth factor, to quiescent cells deprived of the factor. The cellular levels of cADPR increase concomitant to an increase in the cyclase activity and precede cell division (12), which suggests a causal relationship. Isolated microsomes show that the cADPR-sensitive Ca?* stores are present and functional in Euglena. These microsomes are responsive to cADPR, and Ca’* release from them is inhibited by ruthenium red (12, 103). Likewise, caffeine can activate Ca** release from the same stores (103), indicating that the pharmacology is similar to the mammalian cADPR-mechanism. Evidence that cADPR may be involved in regulating the cell cycle in mammalian cells has also been reported (104). HeLa and 3T3 cells transfected with human CD38, a cADPR synthesizing enzyme (described above), show elevation

of intracellular cADPR, partial depletion of thapsigargin-sensitive calcium stores, and increase in basal free cytoplasmic calcium concentration. The cell doubling time of these cells is reduced to as much as 35% of the control cells that do not express CD38 (104). The consistent evidence between Euglena and mammalian cells strongly suggests that cADPR is an important regulator of cell cycle.

FUNCTIONS OF cADPR AND NAADP

327

Activation of Gene Expression in Plants The presence of cADPR-sensitive Ca’* channels in plants was first demonstrated in isolated vacuolar membranes, by using a Ca’* release assay and also by a patchclamping technique (105). The pharmacology of the plant channel has been shown to be very similar to the mammalian RyR (106, 107). The cADPR-pathway is now known to be important in regulating gene expression in plants. Treatment of plant cells with the hormone abscisic acid (ABA) elevates cADPR levels and activates specific genes. The gene expression can be blocked by 8-amino-cADPR (13), a specific antagonist of cADPR-receptor (54), and mimicked by microinjection of either cADPR itself or the Aplysia cyclase (13). The latter presumably would produce cADPR from endogenous NAD. Although IP; can activate the same type of gene expression, its antagonist, heparin, can only block the IP3-induced but not the ABA-induced gene expression, indicating that cADPR is selectively linked to the ABA-signaling pathway (13). Although pharmacological evidence using U73122, a phospholipase C inhibitor, suggests that IP; could also be involved (108), the inhibitor is now shown to be nonspecific and, in fact, is inhibitory to the cADPR-mechanism as well (109, 110).

That cADPR can indeed elevate intracellular Ca*+ by mobilizing the vacuolar Ca** store is demonstrated directly in guard cells (111). A consequence of the Ca** changes is the reduction in turgor pressure of the injected cell and closure of the stoma. ABA can induce similar Ca**t changes and stomatal closure (108, 111).

Both nicotinamide, an inhibitor of the ADP-riboysl cyclase, and 8-amino-cADPR specifically inhibit the ABA-induced stomatal closure (111). cADPR is thus an important messenger that mediates at least two effects of ABA: a short-term effect of closing the stoma to reduce water loss and a long-term effect of expressing specific genes to combat the environmental stress. Another important function in plants, in which cADPR plays a role, is mediating the activation of genes involved in the defense against pathogens. Virus infection of plants elevates nitric oxide synthase activity and induces expression of specific defense genes. The pattern of gene expression activated by viral infection is mimicked by treatment with a nitric oxide donor (112). Introducing either cGMP or cADPR into the cells also induces expression of these genes, which is blocked by ruthenium red (112), an inhibitor of the cADPR-dependent Cat channels. These results are consistent with nitric oxide raising cGMP

levels, which in turn leads

to production of cADPR and Ca** mobilization. The same nitric oxide/cADPRpathway thus operates in plants and rat brain (described above), indicating that it is a general signaling mechanism.

Cau Signaling in Oocytes Marine invertebrate eggs, such as sea urchin eggs, have long been favorite models for investigating mechanisms of Ca** signaling. Fertilization of the eggs is accompanied by a Ca** wave that starts at the sperm-egg fusion site and propagates across the entire egg. This wave serves as a triggering ionic signal that initiates an

328

LEE

intrinsic program, cumulating in cell division and the eventual development of a mature organism. Both the cADPR- and NAADP-mechanisms were first described to occur in sea urchin egg homogenates (1, 2,4). That they are functional in live eggs is demonstrated most convincingly using caged analogs of cADPR and NAADP specifically synthesized for the purpose (113, 114). The functional role of cADPR is assessed using specific inhibitors and is found to mediate, in conjunction with IP3, the Ca’* wave at fertilization, which can be blocked only by inhibiting both the cADPRand IP;-mechanisms (55, 115, 116).

Similar to sea urchin eggs, both Ascidian and starfish oocytes possess all three Ca’* signaling mechanisms. Fertilization of Ascidian eggs is accompanied by a rapid decrease in membrane Ca’* current, exocytosis and the initiation of a long-term oscillation of cytoplasmic Ca?* concentration (117, 118). Each of these functions is selectively served by one of the three Ca*~ signaling mechanisms (118). Thus, intracellular infusion of cADPR readily induces the membrane current decrease and exocytosis. NAADP, on the other hand, can activate the current changes but not the exocytosis (118). The cytoplasmic Ca?~ oscillation is most readily mimicked by infusion of IP;, which however, is ineffective in mediating either exocytosis or the current change (118). In starfish oocytes, the cADPR- and IP;-mechanisms are present both in the cytoplasm and the nucleus (119, 120). Microinjection of antibodies against calmodulin substantially inhibits the rapid nuclear Ca?* elevation induced by photolyzing caged cADPR co-injected into the nucleus (119), indicating that the mechanism is dependent on calmodulin and agrees with that seen in sea urchin eggs (49, 66). In the presence of specific inhibitors of the cADPR- and IP;-mechanisms, the pattern of Ca** changes associated with oocyte-maturation activated by the hormone, l-methyladenine, is selectively modulated, indicating that both Ca*~ signaling mechanisms are involved (119, 120). Additionally, after maturation, the NAADP-

sensitive Ca*+-release mechanism is found to be enhanced and appears to be associated more closely with the plasma membrane (121). Mammalian eggs, including those from mouse, pig and cow, are also responsive to cADPR (122-124). Similar to that observed in Ascidian oocytes (118), cADPR

is particularly effective, even at nanomolar concentrations, in triggering the cortical exocytosis (122), but is less so in inducing a cytoplasmic Ca** change (123), which suggests a preferential association of the cADPR-Ca*~ stores with the egg cortex.

Ca’** Signaling in Pancreatic Acinar Cells In addition to the three invertebrate eggs described above, pancreatic acinar cells and brain microsomes also possess both the cADPR- (52, 125, 126) and NAADP-

sensitive mechanisms (127, 128), in addition to the IP;-mechanism. Treatment of the acinar cells with cholecystokinin (CCK) activates repeated Ca2+ spiking in the apical secretory pole that contains a high density of zymogen granules. The Ca** spiking can be mimicked by infusion of cADPR, and its antagonist,

FUNCTIONS OF cADPR AND NAADP

329

8-amino-cADPR, can specifically block the activating effect of physiological concentrations of CCK (52, 129). Because it has not been possible to demonstrate production of IP; at this hormone concentration range, cADPR may be the primary Ca** messenger under this condition (130). Isolated zymogen granules can respond to cADPR with Ca’* release, which suggests that cADPR may in fact have a direct role in mediating granule exocytosis (131). The relative importance of the cADPR- and IP3;-mechanisms, however, can be metabolically regulated. Thus, high cellular glucose inhibits the cADPR-mechanism but enhances the IP3mechanism, making it dominant (132). Of the three Ca?* messengers, NAADP

is the most effective in the acinar

cells, being active in the nanomolar range as compared to the micromolar range for cADPR and IP; (128). Similar to that found in sea urchin eggs, the NAADPmechanism in acinar cells can be desensitized after exposure to a high concentration of NAADP itself (61,62). Prior desensitization of the NAADP-mechanism blocks

the hormone-induced Ca’* spiking, indicating that it plays a role in the hormonal signaling as well (128). A scheme has been proposed to describe how the three Ca?*-signaling mechanisms are coordinated in the acinar cells (128). Accordingly, CCK first activates a small and highly localized Ca’* release from the NAADP-sensitive stores, which by itself is not sufficient to trigger Ca’* spiking but requires the amplification by the cADPR-sensitive stores via Ca?+-induced Ca*+ release (CICR). Thus, 8-amino-

cADPR, by blocking the cADPR-receptor, inhibits not only cADPR-induced spiking but also the NAADP- and CKK-induced spiking (128, 129). That the cADPRantagonist can indeed inhibit CICR has been directly demonstrated in sea urchin egg microsomes (48,54). The Ca** signal amplified by the cADPR-mechanism further activates Ca2+ mobilization from the IP3-sensitive Ca’*+ stores, which ul-

timately are responsible for initiating Ca** spiking. Consistent with this notion is the observation that heparin, by blocking the CICR-activation of the IP3-stores, inhibits spiking induced by CCK, NAADP, cADPR and IP; (52, 128). This scheme thus proposes NAADP as a trigger, the cADPR-stores as an intermediate amplifier and the IP;-stores as the ultimate activator of hormone-induced Ca’? spiking.

Ca’* Signaling in Cardiac and Smooth Muscle Cells Myocytes contain an abundance of cardiac RyR. As described above, cardiac RyR reconstituted into bilayers is responsive to cADPR (72,73). Transfection and expression of cardiac RyR in cells also confer cADPR responsiveness (78). Therefore, it is puzzling that an early report suggests that cADPR does not regulate Ca’* release in myocytes (133). A series of subsequent studies shows otherwise (80, 81, 134-136). The crucial factor appears to be temperature. Myocytes need to be maintained at a physiological temperature of 36°C (136), which was not done in the earlier study (133). Perhaps the most direct and visual demonstration of the effects of cADPR in myocytes is the enhancement of the frequency of Car sparks (80, 81), which are due to elemental Ca*+ release from a group of RyR (79).

330

BEE

Intracellular release of cADPR from its caged analog

stimulates both the frequency

of Ca2+ sparks and the amplitude of the Ca** transient associated with the action

potential (80). Prior infusion of 8-amino-cADPR blocks the enhancing effects of cADPR on Ca’ sparks. In detergent permeabilized myocytes addition of cADPR produces similar enhancement of the spark frequency that is readily reversible (81). In both cases, the effect of cADPR

requires a period of time to develop

maximally, which suggests some intermediate processes are involved, such as binding to accessory proteins as described above. In addition to myocytes, three types of smooth muscles, including trachea (137), coronary artery (138) and intestinal longitudinal muscle (92), are also responsive to cADPR. In permeabilized tracheal smooth muscle, acetylcholine activates a long-lasting Ca** oscillation, which can be blocked by 8-amino-cADPR (137). A single Ca** transient is produced by cADPR alone, but it greatly increases the frequency and amplitude of the Ca?’* oscillation induced by subsequent exposure to acetylcholine. By sensitizing the Ca**-release mechanism, cADPR can thus effectively modulate the effects of agonists.

Ca’* Signaling in Lymphatic and Blood Cells Various types of lymphatic and blood cells, including macrophages (139), natural killer cells (140), lymphoma cells (141), T-lymphocytes (14, 142-144), and mononuclear cells (145), have been reported to be responsive to cADPR. Activation of the TCR/CD3 receptor complex of T-lymphocytes with a specific antibody elicits a fast increase in intracellular Ca** believed to be mediated mainly by IP;, which is followed by a sustained elevation owing to Ca*~ influx (14). The cellular level of cADPR increases and remains elevated, correlating temporally with the sustained Ca* influx. The influx is inhibited by 7-deaza-8-Br-cADPR, a

permeant antagonist of cADPR (55), in a concentration-dependent manner, indicating that cADPR is indeed responsible (14). It is proposed that cADPR mediates Ca’* influx by depleting the internal stores (144, 146), which in turn activates the capacitative Ca** entry mechanism (147). That cADPR can indeed mobilize the internal Ca** stores is directly shown in permeabilized lymphocytes (142, 143, 148). Consistent with RyR being the target of cADPR is the demonstration that RyR, particularly type 3, is present in lymphocytes and that cADPR can modulate ryan- . odine binding in microsomes (14). The biological consequence of activating the TCR/CD3 complex is cell proliferation, which is accompanied by expression of markers such as CD25 and HLA-DR. 7-deaza-8-Br-cADPR effectively inhibits both cell proliferation and the expression of markers in a concentration-dependent manner without much cytotoxic effect (14). These results provide strong evidence that cADPR is a second messenger important in regulating lymphocyte functions. Similar results on proliferation are found in mononuclear cells derived from cord blood (145). Treatment of mononuclear cells with a high concentration of

cADPR results in permeation of the messenger into the cells, sustained elevation of intracellular Ca**, and increases in the colony output as well as colony size.

FUNCTIONS OF cADPR AND NAADP

ool

These effects can be blocked by a cADPR-antagonist, 8-NH,-cADPR. A nonhydrolyzable analog, 3-deaza-cADPR (58), is even more effective than cADPR itself. These results, taken together with the cell cycle shortening in HeLa and 3T3 cells described above (104), further strengthen the evidence of the importance of cADPR in regulating cell growth and proliferation.

Regulation of Insulin Secretion in Pancreatic B-Cells Pancreatic B-cells respond to high concentrations of glucose by secreting insulin. Evidence suggests that cADPR plays an important role in this process. The cADPR level in the islets is raised by glucose (90, 149). Permeabilized islets respond to cADPR and secrete insulin (126). Cyclic ADP-ribose can release Ca”* from microsomes isolated from the islets as well (90, 149). Transgenic mice over-expressing CD38 (a cADPR-metabolizing enzyme described above) in their B-cells show elevated plasma levels of insulin, and their islets also secrete more insulin in response to glucose (150). Conversely, mice with their CD38 gene knocked out show lower insulin levels in the serum after challenging with glucose, and their islets likewise secrete less insulin in response to glucose (149). The cellular levels of cADPR in

the CD38-ablated islets also fail to increase following exposure to glucose. More intriguing is the detection of the presence of autoantibodies against CD38 in 10%—14% of diabetic patients. These antibodies can bind to and specifically modulate the enzymatic activity of CD38, which suggests that the cADPRmetabolizing enzyme is important in the development of diabetes (151, 152). Despite this impressive series of experiments, the subject has been somewhat controversial (153-155). Part of the reason is related to the type of cells used in the studies. Islets from diabetic mice (ob/ob) and cultured B-cells (RINmSF) do not respond to cADPR (90, 153-155). These cells also secrete much less insulin

in response to glucose and have a depressed level of CD38 and RyR (90, 156). Another source of variability could be the requirement of accessory factors, such as FK506-binding protein (157) and the calmodulin-dependent protein kinase II (67), which have been shown to be important in conferring the cADPR-response in B-cells. Conditions that do not preserve the correct functioning of these factors could produce negative results.

Potentiation of CICR and Transmitter Release in Neurons The mechanism of cADPR-induced Ca’* release as delineated in sea urchin eggs occurs Via sensitizing the Ca2+-induced Ca?+-release (CICR) mechanism ((47, 48)

and reviewed in (158)). In the presence of calmodulin, increasing the concentration

of cADPR reduces the concentration of Ca**+ (or Sr’*) required to activate CICR, such that at high enough concentrations of cADPR, even basal levels of Cae in the nanomolar range are sufficient (48, 49, 66). This potentiation mechanism

has been shown to also operate in intact bullfrog sympathetic neuron (51) and NG108-15 cells (159, 160). Thus, application of cADPR through a patch-clamp

pipette substantially augments the action potential- or depolarization-induced rises

332

LEE

in intracellular Ca**. The potentiation is inhibited by ryanodine, and Ca** imaging shows that CADPR enhances the spatial spread of the Ca’ signal triggered by the influx at the edge of the cell into the center (160). Another important neuronal function that involves cADPR is neurotransmitter release. As described above, cADPR is particularly effective in activating exocytosis in both sea urchin (83) and Ascidian eggs (118). This is also the case in adrenal chromaffin cells, which are neurosecretory. Permeabilized chromaffin cells respond to cADPR and release acetylcholine, which is blocked by pretreatment of the cells with imperatoxin, an inhibitory toxin of RyR (69). Analogous results are found in neurons. Liposomal delivery of cADPR to frog nerve-muscle preparations evokes a rapid increase in the number of quanta released, mainly owing to an increase in the number of functional sites without affecting other quantal parameters (161). In the buccal ganglion of Aplysia, microinjection of cADPR into a cholinergic presynaptic neuron rapidly increases both the intracellular Ca** and the postsynaptic response evoked by a presynaptic spike (162). Preloading the neuron with 8-amino-cADPR or ryanodine blocks the effects. Analyses of the postsynaptic responses show that the number of acetylcholine quanta released is increased following cADPR injection. Application of NAD, the precursor of cADPR, likewise increases the quantal release. Singlecell reverse transcriptase polymerase chain reaction confirms the presence of the cyclase in the presynaptic neuron (162). These results show that cADPR is effective in enhancing neurotransmitter release in both vertebrate and invertebrate neurons.

Ca’** Signaling in Salivary and Lacrimal Gland Cells Various preparations of the salivary gland have been used to investigate the effect of cADPR. Microsomes prepared from the parotid gland respond to cADPR as

assayed using a *Ca’* efflux technique (163). Permeabilized cells from both

the submandibular and the parotid glands respond to cADPR in a concentrationdependent manner, and these effects are blocked by many of the inhibitors of RyR, including imperatoxin, ruthenium red, benzocaine, and high concentrations of ryanodine. The cADPR-induced Ca?* rlease is potentiated by Sr°* and low concentrations of ryanodine. Calmodulin enhances the cADPR-effect, whereas W-7 depresses it (164, 165). The release is also blocked by 8-Br-cADPR (82), an antagonist of CADPR (54). None of these agents affect IP;-induced Ca?* release. The pharmacology is thus essentially identical to that seen in sea urchin eggs (reviewed in (5, 166)). Infusion of cADPR into intact gland cells likewise activates

Ca** release as measured by the Ca*+-activated Cl~ current (167). In these gland

cells, however, CAMP seems to have an intriguing modulating effect on the cADPRinduced Ca** release that is not observed in sea urchin eggs. The expression of RyR in these gland cells is detected by reverse transcriptasepolymerase chain reaction and can be visualized using a fluorescent analog of ryanodine (82, 164). The RyR is localized primarily in the basal pole whereas the

FUNCTIONS OF cADPR AND NAADP

333

IP3-receptor is predominantly in the apical pole (82). Specific binding of ryanodine to microsomes is competitively inhibited by cADPR. Both ryanodine and cADPR, likewise, competitively inhibit the fluorescent staining (82), providing visual and direct evidence that the target of cADPR is the RyR. Permeabilized cells from lacrimal acini, another fluid-secreting gland, have also been shown to release Ca** in response to cADPR, which is blocked by ryanodine (168, 169). Stimulation of the cells with phenylephrine activates Ca*+ release without IP; production. The Ca’* release is likewise inhibited totally by ryanodine, which suggests that cADPR is the mediator of the a-adrenergic receptor activation (168).

Intercellular Ca** Signaling Cells in tissues are commonly connected by gap junctions, forming communicating conduits through which ions and small signaling molecules can permeate. Whether cADPR can diffuse through the gap junctions and mediate intercellular Ca** signaling has been investigated in lens cells (110, 170). In permeabilized lens cells, cADPR specifically releases Ca?* from internal stores, and the release is

blocked by 8-amino-cADPR but not by heparin (170). Microinjection of cADPR into intact cells in monolayer culture not only elevates Ca” in the injected cell but also initiates a Ca** wave that spreads to several layers of adjacent cells (110). Co-injection of cADPR with a Ca** chelator suppresses the Ca’+ change in the injected cell but not the spreading of the Ca** wave to the adjacent cells, indicating it is CADPR and not Ca** that is diffusing to the adjacent cells. These results show that cADPR, like IP3, can permeate through gap junctions and serve as an intercellular Ca?* messenger. Intercellular signaling can occur extracellularly in a paracrine fashion as well. CD38 is present on the surface of many cells (reviewed in (29, 31)). It is homol-

ogous to the cyclase not only in sequence but also structurally (9). As shown in Figure 2, the cyclase is a dimer with a central cavity that is about the size of a molecule of cADPR. The cavity is lined with charged and hydrophilic residues (31). It has been proposed that CD38 could serve as a permeating channel for cADPR (5, 31). CD38 reconstituted into liposomes can indeed catalyze the transport of cADPR into the vesicles during its enzymatic synthesis from NAD (171). Similar transport of cADPR against its concentration gradient is seen in resealed human red cell ghosts that contain endogenous CD38. HeLa cells transfected with CD38 exhibit a transient elevation of intracellular Ca** following exposure to NAD; this suggests that cADPR is synthesized and transported across the cell membrane by surface CD38 to mobilize Ca?+ from internal stores (171). Several

cell types, including HeLa cells, are found to possess a transport mechanism at the cell membrane

that allows NAD

efflux (172). Thus, NAD

can be released

by these cells and serve a paracrine signaling function. Prolonged incubation of NAD with HeLa cells transfected with CD38 triggers endocytosis of surface CD38, which is accompanied by a sustained elevation of intracellular Ca

(172).

334

LEE

Evidence suggests that cytosolic NAD is transported into the endocytic vesicles and is converted by the vesicular CD38 to cADPR, which in turn, is transported back to the cytosol to effect mobilization of internal Ca** stores (172). This novel NAD/CD38-mechanism has been invoked to account for the topological paradox of ecto-CD38 affecting intracellular Ca’* signaling (173).

SEPARATE BUT INTERACTING CALCIUM STORES It is now clear that cells possess at least three independent mechanisms for mobilizing internal Ca’* stores. It is generally known that the IP,-mechanism is present mainly in the endoplasmic reticulum (ER). Fractionation studies show that the cADPR-mechanism co-purifies with the IP;-mechanism and glucose-6phosphatase,

a marker for the ER (1,4, 86). Because

the ER is believed to be

continuous with the nuclear envelope, it is also likely to contain both the cADPRand IP;-dependent Ca**t-release mechanisms, as depicted in Figure 3 (see color insert). This is shown to be the case in the isolated liver nuclei (33, 35, 174), which contain CD38 as well (33-35). This strategic localization of the whole cADPR-

signaling machinery in the nuclear envelope can allow the pathway to directly regulate gene expression. The colocalization of the cADPR- and IP;-mechanisms is not obligatory. For example, in parotid acinar cells, the two mechanisms are segregated to two opposite poles of the cell (82). Organelles other than the ER may also contain the cADPRmechanism. This is the case for pancreatic zymogen granules, an exocytotic organelle (131). Sea urchin egg cortical granules, another exocytotic organelle, and cortical ER likewise contain the cADPR-mechanism (Figure 3). This is suggested by the fact that cortical microsomes incorporated into lipid bilayers exhibit cADPR-dependent Ca** channels (11). The cortical localization of the mechanism can also account for the remarkable effectiveness of cADPR in inducing exocytosis and modulating the plasma membrane Ca** channels as observed in Ascidian eggs (Figure 3) (118). Thus, segregating the cADPR-mechanism to a specific region of the cell can allow a localized Ca** signal to regulate a selective function, such as membrane current or exocytosis, without necessarily activating the whole cell, providing a mean for fine tuning of signaling. The identity of the organelle containing the NAADP-mechanism is currently unknown. Fractionation studies indicate that the NAADP-stores are separable from the IP;- and cADPR-stores as well as from the mitochondria (4). The stores

appear to possess a novel Ca**-ATPase that is insensitive to thapsigargin (175). In Ascidian eggs, the NAADP-stores are segregated in the cortex, because their mobilization readily modulates membrane current but has very little effect on the cytoplasmic Ca** concentration. In contrast, activation of the IP3-stores triggers a cytoplasmic Ca** oscillation without eliciting membrane events (118). This is another example of segregated Ca** stores regulating distinct and selective functions. Another effect of the cortical stores is to insulate the Ca* changes in the region.

Lele, (Ci

Cyclic ADP-ribose (CADPR)

Figure 1 Structures of cADPR and NAADP. The N1-site of cyclization in cADPR is indicated by a dashed circle. The space-filling model is based on X-ray crystallography data (3). The three structural determinants of NAADP crucial to its Ca2t releasing activity are also indicated by dashed circles (59). The space-filling structure of NAADP is based on the crystal coordinates of NADP bound to isocitrate dehydrogenase and modeled using ChemBuilder3D (30, 31). Color code: oxygen, red; hydrogen, white; carbon, black; nitrogen, cyan; phosphorous, yellow.

G2

LEE

Figure 2

Crystal structure of ADP-ribosyl

cyclase. The upper panel shows a cyclase site. The structure is based on X-ray crystallography (8). The van der Waals surface is shown for the left monomer. The secondary structures are shown for the right monome r. Color code: a-helix, red: B-sheet, purple; coil, gray; disulfide bond, cyan. The lower panel shows the active site with three critical residues labeled. Color code: nicotinamide, yellow; carbon atom. green; ¢: oxygen, red; nitrogen, blue. All structures are rendered using the program MolMol ( 178). homo-dimer with nicotinamide (yellow) bound to the active

BE

@=3

Nucleus Figure 3 Separate but interacting Ca2* stores. Mobilization of locally segregated Ca2t 2 P =i): 3 5 stores regulates selective functions, such as membrane Ca“? channels (white cylinders) ; ; nD) ae ay eee) and exocytosis. Interaction between Ca*t stores via CICR can lead to propagative Ca*t waves or oscillations, resulting in global activation of the cell. The abundance of Ca2*ATPase in the cortical stores effectively insulates the cortical region (light blue) from the Ca2t changes in the cytosol (darker blue). A single enzyme, ADP-ribosyl cyclase, is responsible for synthesizing both cADPR and NAADP from NAD and NADP respectively. Representations: cADPR-mechanism, red tetramer; NAADP-mechanism, purple cylinder; IP3-mechanism, green tetramer; thapsigargin sensitive Ca2t-ATPase, yellow sphere; thapsigargin insensitive Ca2+-ATPase, brown sphere.

—

7

= _s

a Ja.)

:

a*

a

om

Qrenip

i> en

=e.

AP

rin Vir

adh e i=

7

-

2

FUNCTIONS OF cADPR AND NAADP

335

The abundance of Ca**-pumps in the cortical Ca?+ stores can effectively buffer the Ca’* in the cortex and minimize the influx to and from the cytosol (Figure 3). Although the three types of Ca’* stores are separate, they can nevertheless functionally interact. Thus, a localized increase in NAADP by photolyzing its caged analog using a focused laser beam can elicit a Ca*+ wave that propagates across the entire sea urchin egg (6). On the other hand, a global increase by whole-field photolysis evokes Ca?” oscillation, which has been proposed to be due to interaction between the NAADP- and cADPR-stores (5, 61, 114). That mobilization of

the NAADP-stores can affect modulation of the IP;-induced Ca** oscillation has been observed in Ascidian eggs (118). In pancreatic acinar cells, results described above suggest that the Ca**+ released from the NAADP-stores serves as a triggering signal that is sequentially amplified by the cADPR- and IP3-stores through the CICR mechanism (Figure 3) (128). The presence of separate but interacting Ca’* stores that respond specifically to distinct messengers provides cells with a versatile mechanism for signaling. Specific and selective functions can be regulated by mobilizing locally segregated stores. Globally, cell activation can occur through interaction between stores, generating propagative Ca** waves. Ca?" signaling can be extended even beyond cell boundaries through gap junctions or through a paracrine mechanism. The versatility of Ca** mobilization as a signaling mechanism makes it perfectly suited for responding to a wide range of environmental stimuli. ACKNOWLEDGMENTS My research is supported by grants from the National (HD17484, GM60333, and GM61568).

Institutes of Health

I thank Richard Graeff for proofreading

the manuscript. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED ie Clapper DL, Walseth TF, Dargie PJ,LeeHC. 1987. Pyridine nucleotide metabolites stimulate calcium release from sea urchin egg microsomes desensitized to inositol trisphosphate. J. Biol. Chem. 262:9561—-68 . Lee HC, Walseth TF, Bratt GT, Hayes RN,

Clapper DL. 1989. Structural determination

of a-cyclic metabolite of NAD* with intracellular Ca*+-mobilizing activity. J. Biol. Chem. 264:1608-15

_ Lee HC, Aarhus R, Levitt D. 1994. The crystal structure of cyclic ADP-ribose. Nature Struct. Biol. 1:143-44

4. Lee HC, Aarhus R. 1995. A derivative of NADP mobilizes calcium stores insensitive

to inositol trisphosphate and cyclic ADPribose. J. Biol. Chem. 270:2152-57

. Lee HC. 1997. Mechanisms of calcium signaling by cyclic ADP-ribose and NAADP. Physiol. Rev. 77:1133-64 . Lee HC. 2000. NAADP: An emerging calcium signaling molecule. J. Memb. Biol. 173:1-8 Aarhus R, Graeff RM, Dickey DM, Walseth

TF, Lee HC.

1995.

ADP-ribosyl

cyclase and CD38 catalyze the synthesis ofa

LEE

calcium-mobilizing metabolite from NADP. J. Biol. Chem. 270:30327-33 . Munshi C, Thiel DJ, Mathews

II, Aarhus

R, Walseth TF, et al. 1999. Characterization

Cyclic ADP-ribose: metabolism and calcium mobilizing function. Vit. Horm. 48: 199-257 18. Lee HC. 1994. Cyclic ADP-ribose: a cal-

cium

of the active site of ADP-ribosyl cyclase. J. Biol. Chem. 274:30770-77 . Munshi C, Aarhus R, Graeff R, Walseth TE, Levitt D, et al. 2000. Identification

of the enzymatic active site of CD38 by site-directed mutagenesis. J. Biol. Chem. 275:21566-71

10. Perez CF, Marengo JJ, Bull R, Hidalgo C. 1998. Cyclic ADP-ribose activates caffeine-sensitive calclum channels from sea urchin egg microsomes. Am. J. Physiol. 274:C430-39 Nil Lokuta AJ, Darszon A, Beltran C, Valdivia HH. 1998. Detection and functional characterization of ryanodine receptors from sea urchin eggs. J. Physiol. 510.1: 155-64 12: Masuda W, Takenaka S, Inageda K, Nishina H, Takahashi K, et al. 1997. Oscil-

lation of ADP-ribosyl cyclase activity during the cell cycle and function of cyclic ADP-ribose in a unicellular organism, Euglena Gracilis. FEBS Lett. 405:104—6 13), Wu Y, Kuzma J, Marechal E, Graeff R, Lee HC, et al. 1997. Abscisic acid signaling through cyclic ADP-ribose in plants. Science 278:2126-30 . Guse AH, da Silva CP, Berg I, Skapenko AL, Weber K, et al. 1999. Regulation of

calcium signalling in T lymphocytes by the second messenger cyclic ADP-ribose. Nature 398:70-73 . Rusinko N, Lee HC. 1989. Widespread occurrence in animal tissues of an enzyme catalyzing the conversion of NAD* into a cyclic metabolite with intracellular

Ca?*+-mobilizing activity. J. Biol. Chem. 264:11725-31 . Lee HC, Aarhus R. 1993. Wide distribution

of an enzyme that catalyzes the hydrolysis of cyclic ADP-ribose. Biochim. Biophys. Acta 1164:68—74 . Lee HC, Galione

A, Walseth

TR

1994,

mobilizing

metabolite

of NAD~.

Mol. Cell. Biochem. 138:229-35

jee Lee HC, Aarhus R. 1991. ADP-ribosyl cyclase: an enzyme that cyclizes NAD~ into a calcium-mobilizing metabolite. Cell Regul. 2:203-9 20. Jackson DG, Bell JI. 1990. Isolation of a cDNA

encoding the human

CD38

(T10)

molecule, a cell surface glycoprotein with an unusual discontinuous pattern of expression during lymphocyte differentiation. J. Immunol. 144:281i-15 Oe States DJ, Walseth TF, Lee HC. 1992. Similarities in amino acid sequences of Aplysia ADP-ribosyl cyclase and human lymphocyte antigen CD38. Trends Biochem. Sci. 17:495 22: Itoh M, Ishihara K, Tomizawa H, Tanaka H, Kobune Y, et al. 1994. Molecular cloning of murine BST-1 having homology with CD38 and Aplysia ADP-ribosyl cyclase. Biochem. Biophys. Res. Commun. 203:1309-17 23.

Kim H, Jacobson EL, Jacobson MK. 1993.

Synthesis and degradation of cyclic ADPribose by NAD glycohydrolases. Science 261:1330-33 24.

Lee HC,

Zocchi

E, Guida L, Franco

L,

Benatti U, et al. 1993. Production and hydrolysis of cyclic ADP-ribose at the outer surface of human erythrocytes. Biochem. Biophys. Res. Commun. 191:639-45 . Howard M, Grimaldi JC, Bazan JF, Lund -

FE, Santos-Argumedo L, et al. 1993. Formation and hydrolysis of cyclic ADPribose catalyzed by lymphocyte antigen CD38. Science 262:1056—59 26. Takasawa S, Tohgo A, Noguchi N, Koguma T, Nata K, et al. 1993. Synthesis and hydrolysis of cyclic ADP-ribose by human leukocyte antigen CD38 and inhibition of the hydrolysis by ATP. J. Biol. Chem. 268:26052-54

FUNCTIONS OF cADPR AND NAADP

337

2a Khoo KM, Chang CF. 1999. Characterization and localization of CD38 in the vertebrate eye. Brain Res. 821:17—25 28. Fernandez JE, Deaglio S, Donati D, Sao-

38. Munshi C Baumann C, Levitt D, Bloom-

boda Beusan I, Corno F et al. 1998. Anal-

3), Prasad GS, McRee DE, Stura EA, Levitt

ysis of the distribution of human CD38 and of its ligand CD31 in normal tissues. J. Biol. Regul. Homeostatic Agents 12: 81-91 2S). Lee HC, Munshi C, Graeff R. 1999. Structures and activities of cyclic ADP-ribose, NAADP and their metabolic enzymes. Mol. Cell. Biochem. 193:89-98 30. Lee HC. 1999. A unified mechanism for enzymatic synthesis of two calcium messengers, cyclic ADP-ribose and NAADP. Biol. Chem. 380:785—93 She Lee HC. 2000. Enzymatic functions and structures of CD38 and homologs. Chem Immunol. 75:39-5S9 32 Mizuguchi M, Otsuka N, Sato M, Ishii Y, Kon S, et al. 1995. Neuronal localization of CD38 antigen in the human brain. Brain Res. 697:235-40 335 Adebanjo OA, Anandatheerthavarada HK, Koval

AP,

Moonga

BS,

Biswas

G,

et al. 1999. A new function for CD38/

ADP-ribosyl

cyclase

in nuclear

Ca**+

homeostasis. Nature Cell Biol. 1:409-14 hoo KM, Chang CF. 2000. Localization 34. of plasma membrane CD38 is domain specific in rat hepatocyte. Arch. Biochem. Biophys. 373:35-43 39. Khoo KM, Han M-K, Park JB, Chae SW, Kim U-H, et al. 2000. Localization of the cyclic ADP-ribose-dependent calcium signaling pathway in hepatocyte nucleus. J. Biol. Chem. 275:24807-17 36. Matsumura N, Tanuma S. 1998. Involvement of cytosolic NAD* glycohydrolase in

cyclic ADP-ribose metabolism. Biochem. Biophys. Res. Commun. 253:246-52 Bile Masuda W, Takenaka S, Tsuyama S, Inui _H, Miyatake K, et al. 1999. Purification and characterization of ADP-ribosyl cyclase from Euglena gracilis. J. Biochem. 125:449-53

field VA, Lee HC. 1998. The homo-dimeric

form of ADP-ribosyl cyclase in solution. Biochim. Biophys. Acta 1388:428-36 DG, Lee HC, et al. 1996. Crystal structure

of Aplysia ADP ribosy] cyclase, a homologue of the bifunctional ectozyme CD38. Nature Struct. Biol. 3:957-64

40. Prasad GS, Levitt DG, Lee HC, Stout CD. 1996. Crystallization of ADP-ribosyl cyclase from Aplysia californica. Proteins 24:138-40 41. Muller-Steffner H, Muzard M, Oppenheimer N, Schuber F. 1994. Mechanistic

implications of cyclic ADP-ribose hydrolysis and methanolysis catalyzed by calf

spleen NAD*

glycohydrolase. Biochem.

Biophys. Res. Commun. 204:1279-85 42. Bell CE, Eisenberg D. 1996. Crystal structure of diphtheria toxin bound to nicotinamide adenine dinucleotide. Biochemistry 35:1137-49. 43. Tao W, Grubmeyer C, Blanchard JS. 1996. Transition state structure of Salmonella Typhimurium orotate phosphoribosyltransferase. Biochemistry 35:14—21 44. Barth H, Preiss JC, Hofmann F, Aktories K. 1998. Characterization of the catalytic site of the ADP-ribsoyltransferase Clostridium botulinum C2 toxin by site-directed mutagenesis. J. Biol. Chem. 273:29506-11 45. Sauve AA, Deng HT, Angeletti RH, Schramm VL. 2000. A covalent intermediate in CD38 is responsible for ADPribosylation and cyclization reactions. J. Am Chem. Soc. 122:7855-59 46. Sauve AA, Munshi C, Lee HC, Schramm VL. 1998. The reaction mechanism for CD38. A single intermediate is responsible for cyclization, hydrolysis, and base-exchange chemistries. Biochemistry 37:13239-49 47. Galione A, Lee HC, Busa WB. 1991. Ca?*induced Ca** release in sea urchin egg homogenates: modulation by cyclic ADPribose. Science 253:1143-46

338

EEE

48. Lee HC. 1993. Potentiation of calcium- and caffeine-induced calcium release by cyclic ADP-ribose. J. Biol. Chem. 268:293—99 49. Lee HC, Aarhus R, Graeff RM. 1995S. Sensitization of calcium-induced calcium release by cyclic ADP-ribose and calmodulin. J. Biol. Chem. 270:9060—66 50. Graeff RM, Podein RJ, Aarhus R, Lee HC. 1995. Magnesium ions but not ATP inhibit cyclic ADP-ribose-induced calcium release. Biochem. Biophys. Res. Commun. 206:786-91 Sl. Hua SY, Tokimasa T, Takasawa S, Furuya Y, Nohmi M, et al. 1994. Cyclic ADPribose modulates Ca2+ release channels for activation by physiological Ca?* entry in bullfrog sympathetic neurons. Neuron 12:1073-79 5), Thorn P, Gerasimenko O, Petersen OH. 1994. Cyclic ADP-ribose regulation of ryanodine receptors involved in agonist evoked cytosolic Ca** oscillations in pancreatic acinar cells. EMBO J. 13:2038-43 53: Meissner G. 1994. Ryanodine receptor/ Ca?*+ release channels and their regulation by endogenous effectors. Annu. Rev. Physiol. 56:485—508 54. Walseth TF, Lee HC. 1993. Synthesis and characterization of antagonists of cyclic-ADP-ribose-induced Ca** release. Biochim. Biophys. Acta 1178:235-42 55: Sethi JK, Empson RM, Bailey VC, Potter BVL, Galione A. 1997. 7-Deaza8-bromo-cyclic ADP-ribose, the first membrane-permeant, hydrolysis-resistant cyclic ADP-ribose antagonist. J. Biol. Chem. 272:16358-63 . Reyes-Harde M, Potter BVL, Galione A, Stanton PK. 1999. Induction of hippocampal LTD requires nitric-oxide-stimulated

58. Wong L, Aarhus R, Lee HC, Walseth TF. 1999. Cyclic 3-deaza-adenosine diphosphoribose: a potent and stable analog of cyclic ADP-ribose. Biochim. Biophys. Acta 1472:555—64 ay) Lee HC, Aarhus R. 1997. Structural determinants of nicotinic acid adenine dinucleotide phosphate important for its calcium-mobilizing activity. J. Biol. Chem. 272:20378-83 60. Lee HC, Aarhus R. 1998. Fluorescent analogs of NAADP with calcium mobilizing activity. Biochim. Biophys. Acta 1425:263-71 61. Aarhus R, Dickey DM, Graeff RM, Gee KR, Walseth TF, et al. 1996. Activation and inactivation of Ca** release by NAADP*. J. Biol. Chem. 271:8513-16 62. Genazzani AA, Empson RM, Galione A. 1996. Unique inactivation properties of NAADP-sensitive Ca-* release. J. Biol. Chem. 271:11599-602 . Genazzani AA, Mezna M, Dickey DM, Michelangeli F, Walseth TF, et al. 1997.

Pharmacological properties of the Ca?*release

.

.

66.

PKG activity and Ca** release from cyclic ADP-ribose-sensitive stores. J. Neurophysiol. 82:1569-76

Mh Walseth TF, Aarhus R, Kerr JA, Lee HC. 1993. Identification of cyclic ADP-ribosebinding proteins by photoaffinity labeling. J. Biol. Chem. 268:26686-91

67.

mechanism

sensitive

to NAADP

in the sea urchin egg. Brit. J. Pharm. 121:1489-95 Inageda K, Takahashi K, Tokita K, Nishina H, Kanaho Y, et al. 1995. Enzyme properties of Aplysia ADP-ribosy] cyclase - Comparison with NAD glycohydrolase of CD38 antigen. J. Biochem. 117:125-31 Sethi JK, Empson RM, Galione A. 1996. Nicotinamide inhibits cyclic ADP-ribosemediated calcium signalling in sea urchin eggs. Biochem. J. 319:613-17 Lee HC, Aarhus R, Graeff R, Gurnack ME, Walseth TF. 1994. Cyclic ADP ribose activation of the ryanodine receptor is mediated by calmodulin. Nature 370:307-9 Takasawa S, Ishida A, Nata K, Nakagawa K, Noguchi N, et al. 1995. Requirement of calmodulin-dependent protein kinase II in cyclic ADP-ribose-mediated intracel-

lular Ca?+ mobilization. J. Biol. Chem. 270:30257-S9

FUNCTIONS OF cADPR AND NAADP

68. Tanaka Y, Tashjian AH, Jr. 1995. Calmodulin is a selective mediator of Ca*tinduced Ca?* release via the ryanodine

receptor-like Ca**+ channel triggered by cyclic ADP-ribose. Proc. Natl. Acad. Sci. USA 92:3244-48 69. Morita K, Kitayama S, Dohi T. 1997. Stimulation of cyclic ADP-ribose synthesis by acetylcholine and its role in catecholamine release in bovine adrenal chromaffin cells. J. Biol. Chem. 272:21002-9 70. Tripathy A, Xu L, Mann G, Meissner G. 1995. Galmodulin activation and inhibition of skeletal muscle Ca**+ release channel (Ryanodine Receptor). Biophys. J. 69:106— 19 cae Ikemoto T, linc M, Endo M. 1995. Enhancing effect of calmodulin on Ca**-induced Ca**+ release in the sarcoplasmic reticulum of rabbit skeletal muscle fibres. J. Physiol. 487.3:573-82 q2: Meszaros LG, Bak J, Chu A. 1993. Cyclic ADP-ribose as an endogenous regulator of the non-skeletal type ryanodine receptor

Ca**+ channel. Nature 364:76-79 U3: Sitsapesan R, McGarry SJ, Williams AJ. 1994. Cyclic ADP-ribose competes with ATP for the adenine nucleotide binding site on the cardiac ryanodine receptor Cart: release channel. Circ. Res. 75:596—-600 74. Sitsapesan R, Williams AJ. 1995. Cyclic ADP-ribose and related compounds activate sheep skeletal sarcoplasmic reticulum Ca? release channel. Am. J. Physiol. 268:C1235—40 The Sonnleitner A, Conti A, Bertocchini F, Schindler

H, Sorrentino

V. 1998.

Func-

tional properties of the ryanodine receptor type 3 (Ryr3) Ca?* release channel. EMBO J. 17:2790-98 76. Fruen BR, Mickelson JR, Shomer NH, Velez

P, Louis

CF.

1994.

Cyclic

ADP-

ribose does not affect cardiac or skeletal muscle ryanodine receptors. FEBS Lett. 352:123-26 ;

We Chen SRW, Li XL, Ebisawa K, Zhang L. 1997. Functional characterization of the

339

recombinant type 3 Ca?* release channel (ryanodine receptor) expressed in HEK293 cells. J. Biol. Chem. 272:24234—46 78. Clementi E, Riccio M, Sciorati C, Nistico G, Meldolesi J. 1996. The type 2 ryanodine receptor of neurosecretory PC12 cells is activated by cyclic ADP-ribose. Role of the nitric oxide/cGMP pathway. J. Biol. Chem. 271:17739-45 1S: Cheng H, Lederer WJ, Cannell MB. 1993. Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science 262:740-44 80. Cui Y, Galione A, Terrar DA. 1999. Effects of photoreleased cADP-ribose on calcium transients and calcium sparks in myocytes isolated from guinea-pig and rat ventricle. Biochem. J. 42:269-73

81. Lukyanenko sparks and

V, Gyorke S. 1999. Ca?+

Ca*+ waves in saponinpermeabilized rat ventricular myocytes. J. Physiol. 521.3:575—85

82. Zhang X, Wen J, Bidasee KR, Besch HR, Jr., Wojcikiewicz

RJ, et al. 1999. Ryan-

odine and inositol trisphosphate receptors are differentially distributed and expressed in rat parotid gland. Biochem. J. 340:519— 27 83. Dargie PJ, Agre MC, Lee HC. 1990. Comparison of Ca*+ mobilizing activities of cyclic ADP-ribose and inositol trisphosphate. Cell Regul. 1:279-90

84. Walseth TF, Aarhus R, Zeleznikar RJ, Jr., Lee HC. 1991. Determination of endogenous levels of cyclic ADP-ribose in rat tissues. Biochim. Biophys. Acta 1094:113-—20

85. Walseth TF, Wong L, Graeff RM, Lee HC. 1997. A bioassay for determining endogenous levels of cyclic ADP-ribose. Meth. Enzymol. 280:287-94. 86. Lee HC. 1991. Specific binding of cyclic ADP-ribose to calcium-storing microsomes from sea urchin eggs. J. Biol. Chem. 266:2276-81 87. Takahashi K, Kukimoto I, Tokita K, Inageda K, Inoue S, et al. 1995. Accumulation of cyclic ADP-ribose measured by

340

LEE

a specific radioimmunoassay in differentiated human leukemic HL-60 cells with all-trans-retinoic acid. FEBS Lett. 371: 204-8

we Graeff RM, Franco L, Deflora A, Lee HC. 1998. Cyclic GMP-dependent and -independent effects on the synthesis of the calcium messengers cyclic ADP-

88. Graeff RM, Walseth TF, Lee HC. 1997. A radio-immunoassay for measuring endogenous levels of cyclic ADP-ribose in tissues.

ribose and nicotinic acid adenine dinucleotide phosphate. J. Biol. Chem. 273:118-25 98. Willmott N, Sethi JK, Walseth TF, Lee

Meth. Enzymol. 280:230-41

89. Reyes-Harde M, Empson R, Potter BVL, Galione A, Stanton PK. 1999. Evidence of

a role for cyclic ADP-ribose in long-term synaptic depression in hippocampus. Proc. Natl. Acad. Sci. USA 96:4061—66 90. Takasawa S, Akiyama T, Nata K, Kuroki M, Tohgo

A, et al. 1998.

Cyclic

ADP-

ribose and inositol 1,4,5-trisphosphate as alternate second messengers for intracellu-

lar Ca*+ mobilization in normal and diabetic beta-cells. J. Biol. Chem. 273:2497— 500 il Okamoto H. 1999. The CD38-cyclic ADPribose signaling system in insulin secretion. Mol. Cell. Biochem. 193:115-18 92. Kuemmerle JF, Makhlouf GM. 1995. Agonist-stimulated cyclic ADP ribose. Endogenous modulator of Ca?*-induced Ca** release in intestinal longitudinal muscle. J. Biol. Chem. 270:25488—-94 08: Higashida H, Yokoyama S, Hashii M, Taketo

M,

Higashida

M,

et al.

1997.

Muscarinic receptor-mediated dual regulation of ADP-ribosyl cyclase in NG108-15 neuronal cell membranes. J. Biol. Chem. 272:31272-77 94, Higashida H, Egorova A, Higashida C, Zhong Z-G, Yokoyama S, et al. 1999. Sympathetic potentiation of cyclic ADP-ribose formation in rat cardiac myocytes. J. Biol. Chem, 274:33348-54 O5; Galione A, White A, Willmott N, Turner M, Potter BV, et al. 1993. cGMP

mobi-

lizes intracellular Ca** in sea urchin eggs by stimulating cyclic ADP-ribose synthesis. Nature 365:456—-59 96. Lee HC. 1994. A signaling pathway involving cyclic ADP-ribose, cGMP, and nitric oxide. News Physiol. Sci. 9:134-37

HC, White AM, et al. 1996. Nitric oxide-

induced mobilization of intracellular calcium via the cyclic ADP-ribose signaling pathway. J. Biol. Chem. 271:3699-705 99. Clementi E, Meldolesi J. 1997. The cross-

talk between nitric oxide and Ca** - a story with a complex past and a promising future. Trends Pharm. Sct. 18:266—69 100. Currie KP, Swann K, Galione A, Scott RH. 1992. Activation of Ca2*-dependent currents in cultured rat dorsal root ganglion neurones by a sperm factor and cyclic ADP-ribose. Mol. Biol. Cell 3:1415-25 101. Pollock J, Crawford JH, Wootton JF, Seabrook GR, Scott RH. 1999. Metabotropic glutamate receptor activation and intracellular cyclic ADP-ribose release Ca** from the same store in cultured DRG neurones. Cell Calcium 26:139-48 . Linden

DJ, Dawson

TM,

Dawson

VL.

1995. An evaluation of the nitric oxide/ cGMP/cGMP-dependent protein kinase cascade in the induction of cerebellar long-term depression in culture. J. Neurosct. 15:5098-105

103. Masuda W, Takenaka S, Tsuyama S, Tokunaga M, Yamaji R, et al. 1997. Inositol 1,4,5-trisphosphate and cyclic ADP-

ribose mobilize Ca**+ in a protist, Euglena Gracilis. Comp. Biochem. Physiol. 118:279-83 104. Zocchi E, Daga A, Usai C, Franco L, Guida L, et al. 1998. Expression of CD38 increases intracellular calcium concentration and reduces doubling time in HeLa and 3T3 cells. J. Biol. Chem. 273:801724 105. Allen GJ, Muir SR, Sanders D. 1995.

FUNCTIONS OF cADPR AND NAADP

Release of Ca** from individual plant vacuoles by both InsP3 and cyclic ADPribose. Science 268:735—37 106. Allen GJ, Schroeder JI. 1998. Cyclic ADP-ribose and ABA signal transduction. Trends Plant Sci. 3:123-25 107. Muir SR, Sanders D. 1996. Pharmacology of Ca** release from red beet microsomes suggests the presence of ryanodine receptor homologs in higher plants. FEBS Lett. 395:39-42 108. Staxen I, Pical C, Montgomery LT, Gray JE, Hetherington

AM,

et al. 1999. Ab-

scisic acid induces oscillations in guardcell cytosolic free calcium that involve phosphoinositide-specific phospholipase C. Proc. Natl. Acad. Sci. USA 96:1779-84 109. Lee S-J, Madden PJ, Shen SS. 1998. U73122 blocked the cGMP-induced calcium release in sea urchin eggs. Exp. Cell Res. 242:328-40

110. Churchill GC, Louis CF. 1998. Roles of

SE

Ca**, inositol trisphosphate and cyclic ADP-ribose in mediating intercellular Ca** signaling in sheep lens cells. J. Cell Sci. 111:1217-25 Leckie CP, McAinsh MR, Allen GJ, Sanders D, Hetherington AM. 1998. Abscisic acid-induced stomatal closure mediated by cyclic ADP-ribose. Proc. Natl. Acad. Sci. USA 95:15837-42

2:

Durner

J, Wendehenne

D, Klessig DF.

1998. Defense gene induction in tobacco by nitric oxide, cyclic GMP, and cyclic ADP-ribose. Proc. Natl. Acad. Sci. USA 95:10328-33 13% Aarhus R, Gee K, Lee HC. 1995. Caged cyclic ADP-ribose:- synthesis and use. J. Biol. Chem. 270:7745-49 114. Lee HC, Aarhus R, Gee KR, Kestner T. 1997. Caged nicotinic acid adenine dinucleotide phosphate:- Synthesis and use. J. Biol. Chem. 272:4172-78 1S Lee HC, Aarhus R, Walseth TF. 1993. Calcium mobilization by dual receptors

during fertilization of sea urchin eggs. Science 261:352-55

116. Galione

A, McDougall

A, Busa

341 WB,

Willmott N, Gillot I, et al. 1993. Redun-

dant mechanisms of calcium-induced calcium release underlying calcium waves during fertilization of sea urchin eggs. Science 261:348-52

NIL7f. Albrieux M, Sardet C, Villaz M. 1997. The two intracellular Ca?* release channels, ryanodine receptor and _ inositol 1,4,5-trisphosphate receptor, play different roles during fertilization in Ascidians. Dev. Biol. 189:174-85 118. Albrieux M, Lee HC, Villaz M. 1998. Calcium signaling by cyclic ADP-ribose, NAADP, and inositol trisphosphate are involved in distinct functions in Ascidian oocytes. J. Biol. Chem. 273:14566—74 IED} Santella L, Deriso L, Gragnaniello G, Kyozuka E. 1998. Separate activation of the cytoplasmic and nuclear calcium pools in maturing starfish oocytes. Biochem. Biophys. Res. Commun. 252:1-4 120. Santella L, Kyozuka K. 1997. Effects of 1-methyladenine on nuclear Ca*+ transients and meiosis resumption in starfish oocytes are mimicked by the nuclear injection of inositol 1,4,5-trisphosphate and cADP-ribose. Cell Calcium 22: 11-20 Wil, Santella L, Kyozuka K, Genazzani AA, De Riso L, Carafoli E. 2000. Nicotinic acid adenine dinucleotide phosphateinduced Ca?* release. Interactions among distinct Ca*+ mobilizing mechanisms in starfish oocytes. J. Biol. Chem. 275:8301— 6 1222 Ayabe T, Kopf GS, Schultz RM. 1995. Regulation of mouse egg activation: presence of ryanodine receptors and effects of microinjected ryanodine and cyclic ADP ribose on uninseminated and inseminated eggs. Development 121:2233-44 123. Yue C, White KL, Reed WA, Bunch TD. 1995.

The

existence

of inositol

1,4,5-

trisphosphate and ryanodine receptors in mature bovine oocytes. Development 121:2645-54

342

LBE:

124. Machaty Z, Funahashi H, Day BN, Prather RS. 1997. Developmental changes in the intracellular Ca** release mechanisms in porcine oocytes. Biol. Reprod. 56:921-30 12>: White AM, Watson SP, Galione A. 1993.

Cyclic ADP-ribose-induced Ca** release from rat brain microsomes. FEBS Lett. 318:259-63 126. Takasawa S, Nata K, Yonekura H, Okamoto H. 1993. Cyclic ADP-ribose in insulin secretion from pancreatic beta cells. Science 259:370-73 12 Bak J, White P, Timar G, Missiaen L, Genazzani AA, et al. 1999. Nicotinic acid adenine dinucleotide phosphate triggers Ca** release from brain microsomes. Curr. Biol. 9:751-54 . Cancela JM, Churchill

1999. Coordination

GC, Galione

A.

of agonist-induced

Ca’*-signalling patterns by NAADP in pancreatic 74-16

acinar

cells.

Nature

398:

129. Cancela JM, Petersen OH. 1998. The cyclic ADP ribose antagonist 8-NH)cADP-ribose blocks cholecystokinin-

evoked cytosolic Ca** spiking in pancreatic acinar cells. Pflug. Arch. 435:74648 130. Matozaki T, Goke B, Tsunoda Y, Rodriguez M, Martinez J, et al. 1990. Two functionally distinct cholecystokinin receptors show different modes of actions on Ca*+ mobilization and phospholipid hydrolysis insolated rat pancreatic acini. J. Biol. Chem. 265:6247—54 iil Gerasimenko OV, Gerasimenko JV, Belan PV, Petersen OH. 1996. Inositol trispho-

Sw.

sphate and cyclic ADP-ribose-mediated release of Ca**+ from single isolated pancreatic zymogen granules. Cell 84:47380 Cancela JM, Mogami H, Tepikin AV, Petersen OH. 1998. Intracellular glucose switches between cyclic ADP-ribose and inositol trisphosphate triggering of cy-

tosolic Ca**+ spiking. Curr. Biol. 8:865— 68

138; Guo XQ, Laflamme MA, Becker PL. 1996. Cyclic ADP-ribose does not regulate sarcoplasmic reticulum Ca?* release in intact cardiac myocytes. Circ. Res. 79:147-S1 134. Rakovic S, Galione A, Ashamu GA, Potter BVL, Terrar DA. 1996. A specific cyclic ADP-ribose antagonist inhibits cardiac excitation-contraction coupling. Curr. Biol. 6:989-96 133: Rakovic S, Shigeo lino YC, Galione A, Ashamu GA, Potter BVL, et al. 1999. An antagonist of cADP-nbose inhibits arrhythmogenic oscillations of intracellu-

lar Ca** in heart cells. J. Biol. Chem. 274:17820-27 136. lino S, Cui Y, Galione A, Terrar DA. 1997. Actions of cADP-ribose and its antagonists on contraction in guinea pig isolated ventricular myocytes:- Influence of temperature. Circ. Res. 81:879-84 . Prakash

YS,

Kannan

MS,

Walseth

TF,

Sieck GC. 1998. Role of cyclic ADPribose in the regulation of [Ca**], in porcine tracheal smooth muscle. Am. J. Physiol. 43:C1653—60 . Kannan

MS, Fenton

AM,

Prakash

YS,

Sieck GC. 1996. Cyclic ADP-ribose stimulates sarcoplasmic reticulum calcium release in porcine coronary artery smooth muscle. Am. J. Physiol. 39:H 8016

139; Ebihara S, Sasaki T, Hida W, Kikuchi Y, Oshiro T, et al. 1997. Role of cyclic ADP-ribose in ATP-activated potassium currents in alveolar macrophages. J. Biol. Chem. 272:16023-29 140. Inngjerdingen M, Al-Aoukaty A, Damaj B, Maghazachi AA. 1999. Differential utilization of cyclic ADP-ribose pathway by chemokines to induce the mobilization of intracellular calcium in NK cells. Biochem. Biophys. Res. Commun. 262:467-72 141. Bourguignon LY, Chu A, Jin H, Brandt NR. 1995. Ryanodine receptor-ankyrin interaction regulates internal Ca2+

343

FUNCTIONS OF cADPR AND NAADP

release in mouse T-lymphoma J. Biol. Chem. 270:17917-22

cells.

Enhanced

insulin secretion in CD38expressing transgenic mice. J. Biol. Chem. 270:30045—50

142. Guse

AH, Dasilva CP, Emmrich F, Ashamu GA, Potter BVL, et al. 1995.

. Ikehata

Characterization of cyclic adenosine diphosphate-ribose-induced Ca** release in T lymphocyte cell lines. J. Immunol. 155:3353-59

A,

GA, Potter BVL, et al. 1996. Regulation

144. Guse

AH,

Berg

I, Dasilva

CP,

ies to CD38 ADP-ribose tients with sulin release 48:2309--15

try induced by cyclic ADP-ribose in intact T-lymphocytes. J. Biol. Chem. 272: 8546-50

ADP-ribose increases intracellular free calcium concentration and stimulates proliferation of human hemopoietic progenitors. FASEB J. 14:680—90 146. Guse AH. 1999. Cyclic ADP-ribose: A novel Ca*+ mobilising second messenger. Cell. Signal 11:309-16 147. Putney JW, Jr. 1986. A model for receptorregulated calcium entry. Cell Calcium 7:1-12 148. Takemura H, Imoto K, Sakano S, Kaneko M, Ohshika H. 1996. Lysophosphatidic acid-sensitive intracellular Ca*+ store does not regulate Ca?* entry at plasma membrane in Jurkat human T-cells. Biochem. J. 319:393—97

149. Kato

I, Yamamoto

Noguchi

N, Takasawa

Y,

Fujimura

M,

S, et al. 1998.

CD38 disruption impairs glucoseinduced increases in cyclic ADP-ribose,

[Ca*+], and insulin secretion. J. Biol. Chem. 274:1869-72

150. Kato I, Takasawa S, Akabane A, Tanaka O, Abe H, et al. 1995. Regulatory role of CD38 (ADP-ribosyl cyclase/cyclic ADP-ribose hydrolase) in insulin secretion by glucose in pancreatic beta cells.

al.

K,

Tohgo

1998.

Auto-

R, Antonelli A, et al. 1999. Autoantibod-

ter BVL, Mayr GW. 1997. Ca*+ en-

Franco L, et al. 2000. Extracellular cyclic

J, Nata

T, et

152: Pupilli C, Giannini S$, Marchetti P, Lupi

Pot-

145. Podesta M, Zocchi E, Pitto A, Usai C,

Satoh

antibodies against CD38 (ADP-ribosyl cyclase/cyclic ADP-ribose hydrolase) that impair glucose-induced insulin secretion in noninsulin-dependent diabetes patients. J. Clin. Invest. 102:395— 401

143. Guse AH, Dasilva CP, Weber K, Ashamu of cADP-ribose-induced Ca** release by Mg? and inorganic phosphate. J. Biol. Chem. 271:23946-53

F,

Nakazawa

(ADP-ribosy] cyclase/cyclic hydrolase) in Caucasian padiabetes:- Effects on infrom human islets. Diabetes

1153: Islam MS, Larsson O, Berggren PO. 1993. Cyclic ADP-ribose in beta cells. Science 262:584-86 154. Rutter GA, Theler JM, Li G, Wollheim CB. 1994. Ca?* stores in insulinsecreting cells - Lack of effect of cADP ribose. Cell Calcium 16:71—80 155: Webb DL, Islam MS, Efanov AM, Brown G, Kohler

M, et al. 1996.

Insulin

ex-

ocytosis and glucose-mediated increase in cytoplasmic free Ca*+ concentration in the pancreatic beta-cell are independent of cyclic ADP-ribose. J. Biol. Chem. 271:19074-79

156. Takasawa S, Nata K, Yonekura H, Okamoto H. 1993. Cyclic ADP-ribose in B cells. Science 262:585

aie

Noguchi N, Takasawa S, Nata K, Tohgo

A, Kato I, et al. 1997. Cyclic ADP-ribose

binds to FK506-binding protein 12.6 to release Ca2+ from islet microsomes. J. Biol. Chem. 272:3133-36 158. Lee HC. 1996. Modulator and messenger functions of cyclic ADP-ribose in calcium signaling. Re. Prog. Horm. Res. 51:355— 88 USI, Higashida H, Robbins J, Egorova A, Noda M, Taketo M, et al. 1995. Nicotinamide adenine dinucleotide regulates

344

LEE

muscarinic

receptor-coupled

Kt

(M)

channels in rodent NG108-15 Cells. J. Physiol. 482:317—23 160. Empson RM, Galione A. 1997. Cyclic ADP-ribose enhances coupling between voltage-gated Ca2+ entry and_ intracellular Ca?+ release. J. Biol. Chem. 272:20967-70 161. Brailoiu E, Miyamoto D. 2000. Inositol trisphosphate and cyclic adenosine diphosphate-ribose increase quantal transmitter release at frog motor nerve terminals: Possible involvement of smooth endoplasmic reticulum. Neuroscience 95:927-31 162. Mothet JP, Fossier P, Meunier FM, Stinnakre J, Tauc L, et al. 1998. Cyclic ADPribose and calcium-induced calcium release regulate neurotransmitter release at acholinergic synapse of Aplysia. J. Physiol. 507.2:405-14

163. Ozawa T, Nishiyama A. 1997. Characterization of ryanodine-sensitive Ca2+ release from microsomal vesicles of rat parotid acinar cells:- regulation by cyclic ADP-ribose. J. Memb. Biol. 156: 231-39 164. Zhang XJ, Wen JY, Bidasee KR, Besch HR, Rubin RP. 1997, Ryanodine receptor

expression is associated with intracellular Ca?* release in rat parotid acinar cells. Am. J. Physiol. 42:C1306-14 165. Yamaki H, Morita K, Kitayama S, Imai Y, Itadani K, et al. 1998. Cyclic ADP-ribose induces Ca**+ release from caffeine-

insensitive Ca?* pools in canine salivary gland cells. J. Dent. Res. 77:1807-16 166. Lee HC. 1995. Cyclic ADP-ribose: A mediator of a calcium signaling pathway. In Ryanodine Receptors, ed. V. Sorrentino, pp. 31-50. Boca Raton, Fl.: CRC 167. Lee MG, Xu X, Zeng WZ, Diaz J, Kuo TH, et al. 1997. Polarized expression of Ca*+ pumps in pancreatic and salivary gland cells-Role in initiation and propa-

gation of [Ca**+]; waves. J. Biol. Chem. 272:15771-76

J, Jorgensen TD, Dissing S.

168. Gromada

1995. The release of intracellular Ca**

in lacrimal acinar cells by alpha-, betaadrenergic and muscarinic cholinergic stimulation: the roles of inositol triphosphate and cyclic ADP-ribose. Pflug. Arch. 429:751-61 169. Gromada J, Jorgensen TD, Dissing S. 1995. Cyclic ADP-ribose and inositol 1,4,5-triphosphate mobilizes Ca** from distinct intracellular pools in permeabilized lacrimal acinar cells. FEBS Lett. 360:303-6 170. Churchill GC, Louis CF. 1999. Imaging of intracellular calcium stores in single permeabilized lens cells. Am. J. Physiol. 276:C426—34 gale Franco L, Guida L, Bruzzone S, Zocchi E, Usai C, et al. 1998. The transmembrane glycoprotein CD38 is a catalytically active transporter responsible for generation and influx of the second messenger cyclic ADP-ribose across membranes. FASEB J. 2:1507-20 . Zocchi

E Usai C, Guida

L, Franco

L,

Bruzzone §, et al. 1999. Ligand-induced internalization of CD38 results in intracellular Ca2* mobilization: role of

NAD* transport across cell membranes. FASEB J. 13:273-83 . Deflora A, Guida

L, Franco

L, Zocchi

E. 1997. The CD38/Cyclic ADP-ribose system:- A topological paradox. Int. J. Biochem. Cell Biol. 29:1149-66 174. Gerasimenko OV, Gerasimenko JV, Tepikin AV, Petersen OH. 1995. ATPdependent accumulation and inositol trisphosphate- or cyclic ADP-ribose-

mediated

release

of Ca**

from

the

nuclear envelope. Cell 80:439-44 17S: Genazzani AA, Galione A. 1996. Nicotinic acid-adenine dinucleotide phosphate

mobilizes

Ca*+ from a_thapsigargin-

insensitive pool. Biochem. J. 315:721—25 176. Ashamu GA, Sethi JK, Galione A, Potter BVL. 1997. Roles for adenosine ribose hydroxyl groups in cyclic adenosine

FUNCTIONS OF cADPR AND NAADP

5'-diphosphate ribose-mediated Ca?* release. Biochemistry 36:9509-17 WIT, Bailey VC, Fortt SM, Summerhill RJ, Galione A, Potter BVL. 1996. Cyclic aristeromycin diphosphate ribose: A potent and poorly hydrolysable Ca** minic

345

of cyclic adenosine diphosphate ribose. FEBS Lett. 379:227-28

178. Koradi R, Billeter M, Wuthrich K. 1996. MOLMOL: a program for display and analysis of macromolecular structures. J. Mol. Graphics 14:51-—55

iM

Oh ie \c

ie?

vl)

;

haar

Cs.

i Saute

' _

a

Annu. Rey. Pharmacol. Toxicol. 2001. 41:347-66

UsE OF BIOMARKERS AND SURROGATE ENDPOINTS IN DRUG DEVELOPMENT AND REGULATORY DECISION MAKING: Criteria, Validation, Strategies! LJ Lesko and AJ Atkinson, Jr Office of Clinical Pharmacology and Biopharmaceutics, Center for Drug Evaluation and Research, Food and Drug Administration, Rockville, Maryland 20852, and Clinical Center, National Institutes of Health, Bethesda, Maryland 20892-1504; e-mail: LeskoL @ cder.fda.gov, [email protected]

Key Words

regulatory review, clinical endpoints

@ Abstract In the future, biomarkers will play an increasingly important role in all phases of drug development, including regulatory review. However, only a few of these biomarkers will become established well enough to serve in regulatory decision making as surrogate endpoints, thereby substituting for traditional clinical endpoints. Even generally accepted surrogate endpoints are unlikely to capture all the therapeutic benefits and potential adverse effects a drug will have in a diverse patient population. Accordingly, combinations of biomarkers probably will be needed to provide a more complete characterization of the spectrum of pharmacologic response. In the future, pharmacogenomic approaches, including those based on differential expression of gene arrays, will provide panels of relevant biomarkers that can be expected to transform the drug development process.

INTRODUCTION Biological markers (biomarkers) can serve many unique firmation of diagnoses, monitoring treatment effects or prediction of clinical outcomes. In this review, we focus tial uses of biomarkers as indicators of drug exposure.

purposes, including condisease progression, and on the current and potenWe take a broad view of

exposure to include drug doses, dosing rates and duration of treatment, and sys-

temic plasma concentrations. We evaluate how the relationship between exposure and the magnitude of biomarker response may be applicable for predicting the efficacy or safety of a drug or drug product. !The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

347

348

LESKO

#® ATKINSON

There is a high level of interest in biomarkers in the pharmaceutical industry, which is faced with the ever increasing cost of research and development, and with growing pressure to accelerate the rate of bringing new drugs to the marketplace. In this context, biomarkers show considerable promise for improving the efficiency and informativeness of drug development and regulatory decision making. For years, a limited number of biomarkers have been used by regulatory agencies as a basis for approval and market access of several drugs. There is a legal basis for this, as well as a common set of biomarker characteristics that provide regulatory authorities with a level of certainty sufficient to allow some biomarkers to be used as surrogates for definitive clinical endpoints. However, there continues to be extensive debate about widespread reliance on biomarkers as substitutes for more traditional evidence of clinical efficacy (1,2). Accordingly, there is a vital need to establish consensus on the basic principles needed to properly develop, evaluate, and validate biomarkers.

DEFINITIONS AND BACKGROUND At least some of the controversy surrounding the use of biomarkers as surrogates for clinical endpoints reflects ambiguity in the terminology used by members of the different disciplines that are concerned with the design, execution, analysis, and evaluation of clinical trials. A number of recent attempts have been made to clarify this terminology (3,4). A synthesis of some proposed working definitions is as follows: (a) biological marker (biomarker)—a physical sign or laboratory measurement that occurs in association with a pathological process and that has putative diagnostic and/or prognostic utility; (b) surrogate endpoint—a biomarker that is intended to serve as a substitute for a clinically meaningful endpoint and is expected to predict the effect of a therapeutic intervention; and (c) clinical endpoint—a clinically meaningful measure of how a patient feels, functions, or survives. The hierarchical distinction between biomarkers and surrogate endpoints is intended to indicate that relatively few biomarkers will meet the stringent criteria

that are needed for them to serve as reliable substitutes for clinical endpoints. In fact, not all clinical endpoints are equally definitive and they can be further categorized as follows: (a) intermediate endpoint—a clinical endpoint that is not the ultimate outcome but is nonetheless of real clinical benefit; and (b) ultimate outcome—a Clinical endpoint such as survival, onset of serious morbidity, or

symptomatic response that captures the benefits and risks of an intervention. In some cases, the clinical benefit of an intermediate endpoint such as exercise tolerance may be important to patients even though this benefit is not associated with improvement in the clinical outcome of increased survival. However, in other cases, when the ultimate outcome is considered, the clinical benefit of an interme-

diate endpoint is more than offset by the adverse effects of drug therapy. For example, quinidine was used for many years to maintain normal sinus rhythm in patients who previously had atrial fibrillation. Maintenance of normal sinus rhythm was beneficial to some patients because it was associated with increased cardiac output and a decreased risk of systemic embolization from the

BIOMARKERS AND SURROGATE ENDPOINTS

349

PROARRHYTHMIA

ATRIAL FIBRILLATION

N

—>

SURVIVAL

eS

CARDIOVERSION

Figure 1 Path diagram illustrating the potential of the adverse proarrhythmic effects of quinidine therapy (broken line) to outweigh its potentially beneficial effects (solid line) in maintaining normal sinus rhythm (NSR) in patients with previous atrial fibrillation.

right atrium. Although meta-analysis confirmed that patients treated with quinidine remained in normal sinus rhythm longer than those who were untreated, it was found that quinidine therapy was associated with increased mortality (5). The path diagram shown in Figure | can be used to illustrate this apparent therapeutic paradox. This example deals with an intermediate clinical endpoint, but unanticipated adverse consequences of drug therapy are a frequent confounding factor when biomarkers are relied on as surrogates for definitive clinical endpoints. This limitation underlies much of the controversy surrounding the use of surrogate endpoints as the basis for regulatory evaluation of new therapeutic entities (1-3). Biomarkers and surrogate endpoints in current use usually consist of either physiological or laboratory measurements. Several biomarkers and surrogate endpoints commonly used for a number of therapeutic drug classes are listed in Table 1, together with their corresponding clinical endpoints. Even these commonly used biomarkers vary with respect to their acceptance as surrogate endpoints. Thus, blood pressure and cholesterol are the only two cardiovascular biomarkers currently accepted as surrogate endpoints (3). The results of the CAST (cardiac arrhythmia suppression trial) study have shown that suppression of ventricular arrhythmias can no longer substitute for survival in evaluating the efficacy of antiarrhythmic drugs (6). Similarly, increases in bone mineral density do not necessarily reflect decreases in fracture rate in patients treated with fluoride (7), and

decreases in serum levels of prostate-specific antigen may not correlate with a decrease in tumor growth (8). Although the biomarkers listed in Table 1 differ with respect to their ability to substitute for definitive clinical endpoints, they all have some degree of clin-

ical utility. Biomarkers such as these have traditionally been identified through

350

LESKO

TABLE 1

#8 ATKINSON

Examples of biomarkers and surrogate endpoints*

OT

Therapeutic class

Biomarker/surrogate

Clinical endpoint

Physiologic markers Antihypertensive drugs Drugs for glaucoma Drugs for osteoporosis Antiarrhythmic drugs

{Blood pressure {Intraocular pressure *Bone density { Arrhythmias

| Stroke Preservation of vision

Negative culture

Clinical cure *Survival {Morbidity {Coronary artery disease Tumor response

Laboratory markers Antibiotics Antiretroviral drugs Antidiabetic drugs Lipid-lowering drugs Drugs for prostate cancer

*CD4 count, | viral RNA

{Blood glucose {Cholesterol | Prostate-specific antigen

| Fracture rate 4 Survival

“Reproduced with permission from Reference 61.

studies of pathophysiology or epidemiology that have established their biological plausibility. Thus, clinical and epidemiological evidence indicated that high blood pressure was associated with an increased incidence of atherosclerotic cardiovascular disease, heart failure, stroke, and kidney failure (9). The mechanistic linkage between hypertension and cerebral hemorrhage and infarction was further established by pathophysiologic studies in man and in animal models (10). This linkage provided an initial basis of construct validity for believing that reductions in high blood pressure might be reflected in improved clinical outcomes (11). However, experience acquired through well-controlled clinical trials is needed to provide the criterion validity that is the best support for a particular biomarker (11). Thus, a large clinical trial, in which over 4000 patients with elevated serum cholesterol levels and coronary artery disease were studied, was required to establish that cholesterol-lowering drugs could have a favorable impact on overall mortality as well as on the occurrence of cardiovascular events (12). However,

there is a hierarchy in the level of criterion validity provided by clinical trials. Temple (3) has emphasized that one level of support for a biomarker is obtained when a number of drugs of the same pharmacologic class have consistent effects on the marker and on a relevant clinical endpoint. Even greater support is provided for a biomarker when this consistency can be demonstrated by drugs from different pharmacologic classes. As a result, both Rolan (11) and Temple (3) have concluded that biomarkers, paradoxically, are the least innovative, and thus in

many situations the least useful, when their validity is best established.

USES OF BIOMARKERS IN DRUG DEVELOPMENT Certainly a high level of stringency is required when a biomarker response is substituted for a clinical outcome and is proposed as the basis for regulatory approval of an application to market a new drug. However, biomarkers need not be validated

BIOMARKERS AND SURROGATE ENDPOINTS

351

as rigorously in order to play other important roles, such as facilitating our understanding of disease mechanisms and natural history, expediting the development of new drugs, addressing regulatory concerns related to dose-exposure-response relationships, and even assisting with some aspects of clinical practice. For example, a few tumor markers, such as prostate-specific antigen and a-fetoprotein, are used to help diagnose and monitor the treatment response of patients with prostate and hepatocellular carcinoma (13). Despite their clinical utility, changes in these biomarkers would not constitute an appropriate regulatory basis for new drug approval unless accompanied by appropriate clinical evidence of disease response. Nonetheless, these and other biomarkers can play an important role in a number of phases of new drug development, and in the regulatory review of investigational new drugs (INDs) and new drug applications (NDAs).

Drug Discovery and Preclinical Development Epidemiologic studies that link changes in a biomarker to pathophysiology can play an important role in identifying a suitable therapeutic target. For example, the association of elevated serum cholesterol levels with an increased incidence of coronary heart disease provides an underlying rationale for developing drugs that lower cholesterol by inhibiting 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase (14).

Biomarkers also play an important role in the preclinical assessment of potentially beneficial and harmful effects of a new drug candidate. Screening tests in animals using biomarkers, such as blood pressure lowering, provide important demonstration that a compound is likely to have the intended therapeutic activity in patients. Biomarkers for potential toxicity play an equally important role. For example, a drug found to prolong the QT interval in animals may warn of potential cardiovascular risk in subsequent clinical studies. Pharmacokinetic-pharmacodynamic (PK-PD) studies with biomarkers may be particularly useful (15, 16). In one instance, PK-PD studies showed good correla-

tion between the hypotensive effects of an antiarrhythmic drug in dogs and humans (17). Blood levels measured when adverse events such as seizures occur in animal

toxicology studies may help guide the design of dose escalation studies in humans and serve as a surrogate for preventing similar adverse events in humans (18). Breimer & Danhof (19) have provided additional examples in which whole-animal, mechanism-based PK-PD studies have been used to forecast the results of human PK-PD studies and to guide dose selection and dose escalation strategies.

Early Phase Clinical Development Biomarkers are perhaps most useful in the early phases of drug development, when measurement of clinical endpoints may be too time-consuming or cumbersome to provide timely proof of concept or dose-ranging information. An example would be a study in which different doses of zafirlukast, a leukotriene antagonist, were

352

LESKO

#® ATKINSON

administered to asthmatic subjects to assess the efficacy of this agent in preventing leukotriene D,-induced bronchoconstriction (20). Plasma concentration measurements made in conjunction with this study also demonstrated a plasma concentration threshold of 5 ng/ml that was required for therapeutic effects, supporting evidence of a receptor-mediated mechanism of action for this drug. Mildvan et al (21) have proposed a biomarker classification scheme for use in clinical trials of antiretroviral drugs and have emphasized the important role that CD4~ T-lymphocyte counts and measurement of HIV-1 plasma RNA concentrations have played in the early clinical development of these drugs. In some cases, a new biomarker is needed to facilitate the development of a

novel compound. For example, the proportion of hemoglobin molecules modified to have a high oxygen-binding affinity (%7MOD) was used in the early phase evaluation of tucaresol, a drug designed to prevent hemoglobin S (HbS) polymerization and subsequent hemolysis and painful crises in patients with sickle cell disease (11). The extent to which HbS is polymerized depends on the erythrocyte concentration of deoxygenated HbS, and the scientific rationale for this biomarker was based on the observation that HbS polymerization is inhibited when 20%-30% of hemoglobin is maintained in the oxy-conformation. Measurements of %MOD were included in the initial Phase I studies of tucaresol to demonstrate the oral dose range needed to obtain %MOD values of 19%—26% (22). This marker was also used to guide tucaresol dosage in the initial studies in patients with sickle cell anemia (23). Additional endpoints used in this study were lactic

dehydrogenase and bilirubin concentrations as markers of hemolysis, and percentage of irreversibly sickled cells. Although the validity of these markers has not been confirmed in extensive clinical trials of antisickling therapy, this example illustrates the potential utility of even novel markers during early phase drug development.

Late-Phase Clinical Development Several studies with simvastatin, an HMG-CoA reductase inhibitor, can be cited to illustrate the continued use of a biomarker throughout a clinical development program. Serum cholesterol measurements were used as a biomarker in a Phase II dose-ranging study (24). The efficiency of this study is indicated by the facts that — only four study centers were needed to enroll the 43 patients who participated in the study, and the study duration was only 6 weeks. Although daily simvastatin doses of 80 mg were well tolerated, the study indicated that near-maximal effects were obtained with a daily dose of only 20 mg. The 20-mg/day dose was then selected as the starting dose for the subsequent placebo-controlled Phase III trial, in which 444 patients with coronary heart disease were followed in 94 centers for a median of 5.4 years (12). Serum cholesterol measurements were used in this pivotal trial as the basis for further simvastatin dose adjustments, the goal of treatment being to reduce serum total cholesterol to 3.0-5.2 mmol/liter (117-200 mg/dl). Two aspects of this Phase III trial merit particular emphasis. First, the primary endpoint of the study

BIOMARKERS AND SURROGATE ENDPOINTS

353

was total mortality. By showing that the relative risk of death for patients receiving simvastatin compared with placebo-treated patients was 0.70 (95% confidence interval: 0.58—0.85,

p =

0.003), the study provided the first strong evidence for

advancing serum cholesterol measurements from biomarker to surrogate endpoint status. Second, by selecting for this study an initial simvastatin dose of 20 mg/day as the minimal dose for satisfactory effect (MDSE) rather than the maximally tolerated dose, the sponsor avoided the all-too-common pitfall of registering a starting dose that was subsequently found to be excessive (11). A final benefit from this Phase III study is that serum cholesterol measurements are now used routinely in clinical practice as a biomarker to guide simvastatin dose adjustments. Despite the current acceptance of cholesterol lowering as a surrogate endpoint, it should be pointed out that other drugs that lower cholesterol may have adverse effects that outweigh their benefit. For example, probucol, a drug structurally unrelated to statin HMG-CoA reductase inhibitors, has pronounced lipid-lowering effects but also prolongs the electrocardiographic QT interval and has caused torsades de pointes ventricular tachycardia in some patients (25). Accordingly, it cannot be assumed a priori that any cholesterol-lowering drug will have beneficial effects on survival. Biomarkers that reflect disease prognosis may also be useful in developing eligibility criteria and stratification groups in late-phase clinical trials. Baseline plasma HIV-1 RNA concentrations and CD4~ T-lymphocyte counts have been shown to be independent prognostic markers of clinical progression in patients receiving antiretroviral therapy for HIV-related disease (26). This supports the established practice of using CD4* T-lymphocyte counts as entry and stratification criteria for clinical trials of antiretroviral therapy (21). Although these biomarkers have also provided a basis for the accelerated approval of a number of antiretroviral drugs (27), federal regulations stipulate that accelerated approvals based on a surrogate endpoint are subject to the requirement of further studies to demonstrate clinical benefit (28). Perhaps the most widespread application of surrogate endpoints in late-phase clinical development is in the substitution of drug concentration measurements for clinical endpoints in the registration of new drug formulations and generic drug products. Federal regulations state that measurement of either blood concentrations or urine excretion rates of a drug may be used to demonstrate that a new formulation has bioavailability comparable to that of the reference material (29).

EVALUATION AND VALIDATION OF BIOMARKERS The scientific program for evaluating biomarkers must be planned as early as possible in the drug discovery and preclinical period of drug development with a blueprint to bring that biomarker into clinical trials and to establish the link between the biomarker and the clinical outcome. There is a critical need for rigorous assessment of the procedures and criteria used to evaluate or validate biomarkers in order for them to gain widespread acceptance. As emphasized above, the extent

354

LESKO

#& ATKINSON

of rigor depends on the intended use of the biomarker (11). Most of the emphasis and work in the biomarker field has been geared toward efficacy uses. This is generally because the primary goal of early phase drug development is to establish proof of concept by gathering short-term evidence in screening clinical trials. On the other hand, many adverse drug reactions have a relatively low incidence, and long-term drug exposure is needed to resolve safety issues and reveal infrequent but important adverse events. Nonetheless, increased effort needs to be expended on the development and evaluation of improved biomarkers for drug toxicity. The ultimate value of a biomarker will depend on whether it is assessed in an exploratory or observational type of study, or in a definitive or confirmatory study. The most desirable paradigm for evaluation of biomarkers is provided by adequate and well-controlled clinical studies that (a) define standardized relationships be-

tween drug exposure and response, (b) test hypotheses regarding mechanism of drug action, and (c) provide estimates of the magnitude of benefit. The size and duration of the treatment effect are essential aspects of biomarker evaluation, but sample size and study design are also important. Adequate and well-controlled studies to evaluate biomarkers are often not attempted or are not feasible during drug development. However, one should not focus too strongly on developing a biomarker just to serve as a surrogate endpoint. For example, the evaluation or validation of a biomarker intended to be examined

in a Phase II proof-of-therapeutic concept study may be based on a well-controlled study with a relatively small number of subjects and a short duration of treatment. Such a study may be called observational because it lacks the study power to test a hypothesis, but it can provide valid data to assess the strengths and limitations of a biomarker. This may be acceptable for addressing such exploratory questions as proof-of-therapeutic concept or even certain regulatory questions about dose or dosage regimen changes, but it would be inadequate for establishing a biomarker as a surrogate for a clinical outcome.

Evaluation of Biomarkers Evaluation of a biomarker can be how many of the characteristics context of its use. Characteristics been described by several authors should include the following.

based on an exploratory process of determining of an ideal biomarker are met relative to the of biomarkers that underpin their utility have (3, 30, 31). Ideally, the attributes of a biomarker

1. Clinical relevance, in that the marker provides evidence to support a theoretically rational basis for use, such as the ability to reflect some

measurement of, or change in, a physiologic or pathologic process or activity over a relatively short period of time. The marker is influenced by exposure to a drug and is believed or assumed to be related to the drug’s presumed pharmacologic action or intended clinical effect. There should be a strong, mechanistic molecular or biochemical basis for the biomarker

BIOMARKERS AND SURROGATE ENDPOINTS

355

in which it is positioned early or late in the causal chain of pathological events leading to the clinical endpoint. This obviously requires an understanding of the pathophysiology of a disease and of a drug’s mechanism of action. However, one must recognize that diseases frequently have multiple causal pathways. 2. Sensitivity and specificity to treatment effects, defined as the ability to detect the intended measurement or change in the target patient population via a given mechanism, without interference from other pharmacologic or

clinical effects of the drug unrelated to the drug’s mechanism of therapeutic action. The caveat here is that, as shown in Figure 1, drugs have both intended and unintended actions. 3. Reliability, defined as the ability to measure analytically the biomarker or change in biomarker with acceptable accuracy, precision, robustness, and

reproducibility. This refers to the quality and variability of the assay for quantitating the biomarker. 4. Practicality, defined as noninvasiveness or only modest invasiveness in order to obviate inconvenience and discomfort to healthy volunteers or patients

5. Simplicity, for routine utilization without the need for sophisticated equipment or operator skill, extensive time commitment, or low measurement cost. This is needed to facilitate widespread acceptance of the biomarker for use in drug development and in subsequent clinical practice. Validation of a biomarker is a complex part of the evaluation process. The criteria for validation are defined by the nature of the question that the biomarker is intended to address, the degree of certainty that is required for the answer, and the assumptions about the relationship between changes in the biomarker and clinical endpoints. Validation has been described as not being an all-or-none (binomial) variable, such as the outcome of an efficacy trial, but a continuous variable that varies during the drug development process as new information and data are obtained (11). There are multiple dimensions to biomarker validation that encompass important elements of study design and data analysis, including statistical assessment. There are also multiple pathways to validation of a biomarker for an intended use, and validation data itself is likely to arise from the totality of evidence provided progressively by preclinical animal studies, early Phase I and Phase II clinical studies in healthy volunteers or patients, and late-phase efficacy and safety trials in patients with the targeted disease. _ Typically, validation takes into account the following properties of a biomarker and criteria for validation.

1. Sensitivity, referred to as the ability of an appropriate biomarker or a change in biomarker to be measured with adequate precision, and with sufficient magnitude of change, to make it sensitive enough to reflect a

356

LESKO

#® ATKINSON

meaningful change in important clinical endpoints. Sensitivity also describes the quality of the relationship between the magnitude of change in the biomarker and the magnitude of changed in the clinical endpoint because a high level of correlation, unfortunately, does not necessarily prove a cause-effect relationship. 2. Specificity, referred to as the ability of a biomarker or a change in biomarker to distinguish patients who are responders to an intervention from those who are nonresponders in terms of changes in clinical endpoints. Specificity defines the extent to which a biomarker explains all or most of the changes in a clinical endpoint. 3. Bioanalytical assessment of the laboratory or test measurement of the biomarker in terms of accuracy, precision, reproducibility, range of use, and variability. 4. Probability of false positives, defined by situations in which a desired change in a biomarker is not reflected by a positive change in a clinical endpoint or, even worse, is associated with a negative change in a clinical endpoint. 5. Probability of false negatives, defined by situations in which no change or a small observed change in a biomarker fails to signal a positive, meaningful change in a clinical endpoint. 6. A PK-PD model that has been shown to predict future clinical outcomes or suitable dose adjustments based on biomarker measurement. This establishes the correlation between changes in the biomarker and changes in drug exposure, measured as plasma concentration or dose. One of the challenges here is to prospectively plan and properly implement the model and to determine which metrics of drug exposure and biomarker time course are best able to predict clinical outcomes. There are some patient factors (e.g. age, gender, race, and genetics), disease factors (e.g. stage and progression), and drug factors (e.g. metabolism and protein binding) that may modify treatment effects on biomarkers but are not themselves directly affected by a drug. Many of these factors may necessitate adjusting treatment effects on biomarkers and, thus, may affect the validity with which a biomarker can be applied to all patients with a disease. For example, the cognitive status of elderly patients with Alzheimer’s disease or the etiology of hypertension in African-American patients may influence the predictive value of biomarkers that are influenced by drug exposure.

Validation of Biomarkers as Surrogate Endpoints A surrogate endpoint can be thought of as a biomarker that can be definitively substituted for a clinically meaningful endpoint in an efficacy trial. There is value in biomarkers as surrogate endpoints only to the extent to which they can predict

BIOMARKERS AND SURROGATE ENDPOINTS

357

long-term clinical outcome and serve as confirmatory evidence of efficacy. Many authors have published their validation criteria for surrogate endpoints. The most rigorous standards are those of Fleming & DeMets (2), who stipulated that both of the following conditions must be satisfied: (a) The surrogate endpoint must be correlated with the true clinical outcome; and (b) as initially proposed by Prentice (32), the surrogate endpoint must fully capture the net effect of treatment on clinical outcome. More recently, Temple (3) has laid out examples of evidence that support and evidence that does not support the use of biomarkers as surrogate endpoints. Extensive clinical evidence is needed, and the process of rigorous scientific and

Statistical assessment can be time-consuming and expensive. In many cases, the time and effort needed to validate a surrogate endpoint, to the extent that it is accepted by regulatory authorities, may exceed the time and effort that would be expended in measuring the clinical outcome directly. Our current state of knowledge and lack of public consensus on validation of biomarkers as surrogate endpoints makes it impossible to provide specific steps or guidelines that can be followed. However, many authors have suggested approaches to validating surrogate endpoints and several criteria can be summarized as follows (30-33).

1. Biological plausibility should provide a mechanistic basis for using the surrogate endpoint. 2. Epidemiological or survey studies of the natural history of the disease should support surrogate status by establishing the statistical relationship between the biomarker and the clinical endpoint under basal conditions (30).

3. Adequate and well-controlled clinical trials should provide an estimate of the expected benefit in terms of clinical endpoints that can be derived mathematically or mechanistically from an estimate of the change in the potential surrogate endpoint. Ideally, an appropriate dose- or exposure-response relationship would be established as supplemental support for surrogate status. 4. The analysis should include a consideration of potential adverse reactions unrelated to the clinical endpoints predicted by the surrogate endpoint. 5. An exposure-response model should be developed that mathematically describes and predicts relationships between drug doses or plasma concentrations, and surrogate endpoints and clinical outcomes. Verification of these predictions is important. 6. The development and validation of biomarkers and surrogate endpoints should be built into the drug development process, beginning with the preclinical phase. 7. It may be helpful to conduct a meta-analysis of multiple clinical trials to look across and within studies to determine the consistency of effects

358

LESKO

#® ATKINSON

following interventions with various drug classes and within different stages of the disease (33).

Although there are many useful biomarkers, only a few of them are validated as surrogate endpoints and used to document the efficacy of a drug. Hence, there is a risk that the potentially useful applications of biomarkers will be overlooked in ill-advised attempts to elevate them to surrogate status.

USE OF BIOMARKERS IN REGULATORY REVIEW Observational and definitive clinical trials are conducted during the course of drug development to address a number of questions posed by the sponsor. The use of biomarkers for developing dose response or PK-PD relationships has been shown to increase the efficiency and informativeness of clinical studies (34). In particular, biomarker-based data can provide answers to questions related to dose and dosage regimens needed for the product’s label. These are of considerable interest to sponsors, as they may help to assure success in the marketplace through product differentiation. Similarly, regulatory authorities can pose questions to the database submitted in an application or dossier and frequently rely on dose-response or PKPD relationships to better understand the effects of a drug and to address their own questions related to drug dose and dose adjustments that might be needed in light of patient factors that introduce variability in average exposure-response relationships. There have been several excellent publications that provide a regulatory perspective on the use, benefits, and risks of biomarkers and surrogate endpoints in regulatory decisions leading to market access of new drugs (1,3). For ordinary approvals, there are relatively few, well-established surrogate endpoints. They include blood pressure and serum cholesterol for cardiovascular drugs, blood sugar and glycohemoglobin for antidiabetic drugs, plasma testosterone levels for prostate anticancer drugs, and tumor size for antineoplastic agents. Viral RNA load and CD4* T-lymphocyte counts are the most well-known surrogate endpoints used for accelerated approval of antiretroviral drugs. However, accelerated approval also has been based on reductions in tumor size and decreased rate of gastrointestinal polyp formation as surrogate endpoints (R Temple, personal communication).

Legal Basis The Food and Drug Administration (FDA) has a legal basis for using surrogate endpoints in ordinary and accelerated drug approvals leading to market access of new drugs or drug products (28, 35). The standards for linking a biomarker to a clinical outcome are higher for ordinary approvals than for accelerated approvals. This difference is based on consideration of many factors, including the degree of scientific evidence needed to support biomarker surrogacy, public health needs, relative risk/benefit ratio, and availability of alternative treatments.

BIOMARKERS

AND SURROGATE ENDPOINTS

359

Surrogate Endpoints as Confirmatory Evidence The FDA Modernization Act of 1997 (36) states that confirmatory evidence, when

combined with evidence from one adequate and well-controlled study, can support effectiveness as required for ordinary drug approvals. This provision has generated significant interest in the potential role of biomarkers to function as surrogate endpoints in providing that confirmatory evidence. The quantity of evidence needed to support effectiveness, other than two adequate and well-controlled clinical trials, is discussed in Section II of the FDA Guidance for Industry, entitled “Providing Clinical Evidence of Effectiveness for

Human Drug and Biological Products” (37). This guidance states that one adequate and well-controlled clinical efficacy study can sometimes be supported by evidence from a well-controlled study or studies using a pharmacologic effect, as a biomarker, that is not an established surrogate endpoint. Acceptance of this evidence of efficacy is based on (a) the quantity of evidence showing that there is a strong theoretical or mechanistic link between the pharmacologic effect and clinical outcome; and (b) the quantity of data showing that there is a strong link between the pharmacologic effect and clinical outcome based on prior experience with the pharmacological class, and a clear understanding of the pathophysiology and mechanism of drug action. Further discussion is necessary with regard to study design and data analysis in order to clarify the nature and use of biomarker data, linked to dose and/or plasma drug concentrations, to serve as potential confirmatory evidence. To facilitate that discussion, the FDA is in the process of writing a guidance for industry that deals with exposure-response relationships.

Other Regulatory Uses of Biomarkers Biomarkers that are imperfect surrogate endpoints for any of a variety of reasons often are useful in addressing regulatory questions, and sponsors are encouraged not to abandon attempts to bridge biomarkers to clinical outcomes once Phase III efficacy trials are underway. Obviously, these biomarkers should have a rational and reasonable

link to clinical outcome,

and there should be a hy-

pothesis that supports their use and that makes them relevant to decision making by a regulatory agency. Aside from the use of surrogate endpoints in adequate and well-controlled clinical trials to support the effectiveness necessary for market access, biomarkers that do not meet standards for becoming surrogate endpoints have other value in regulatory decision making and may be used in analyses that complement the results of adequate and well-controlled efficacy trials. For new chemical entities, these biomarkers are frequently incorporated in observational studies that are conducted routinely in Phase I or Phase II drug development. The FDA Guidance for Industry, entitled “Providing Clinical Evidence of Effectiveness for Human Drug and Biological Products” (37), provides an important

360

LESKO

#® ATKINSON

perspective on the potential usefulness of biomarkers, surrogate endpoints, and PK-PD relationships. Some examples of these uses are provided below (37).

le The most obvious example is the use of plasma drug and/or metabolite levels as surrogate endpoints for efficacy and safety in the approval of generic drug products. . For new chemical entities, biomarkers can support, but not replace, clinical outcome results from efficacy studies when they are proximal to the clinical outcome, or they can measure real clinical benefit to the patient, such as increasing exercise tolerance or improving pulmonary function. . Biomarkers that are more distal in the causal chain leading to the clinical outcome and were investigated in early clinical trials may be suitable for assessing the clinical significance of changes in systemic drug exposure due to intrinsic and extrinsic patient factors, such as age, gender, smoking habit, degree of renal impairment, and drug-drug interactions. Several clinical pharmacology regulatory guidances recommend that sponsors define therapeutic equivalence limits using PK-PD relationships for studies of drug-drug interactions, and the effects of renal or hepatic impairment to determine label claims and the need to adjust doses (38, 39). For example, HMG-CoA reductase inhibition or bleeding times may be used as biomarkers to assess drug-drug interactions with cholesterol-lowering statins and anticoagulants, respectively. . Biomarkers may be useful for subgroup analyses of efficacy or safety data from adequate and well-controlled clinical trials or from meta-analysis of several clinical studies to identify covariates that were expected to account for differences in response. The level of certainty provided by a subgroup analysis using a biomarker is dependent on many factors, including attributes of the biomarker, as described above, and whether the hypothesis of a subgroup difference was developed prestudy or post hoc.

. Biomarkers are useful in providing adequate evidence to bridge from a preexisting database of efficacy to support an efficacy decision in new situations or settings. These include approval of drugs or drug products for different populations (e.g. pediatric and ethnic groups) when certain conditions are met, different dosage forms (e.g. a controlled-release product for an established immediate-release product), different routes (e.g. parenteral vs oral), and different dosage regimens (e.g. three times a day vs twice a day). The ICH E5 Guidance, entitled “Ethnic Factors in the Acceptability of Foreign Clinical Data” (40), addresses the use of biomarkers specifically for bridging efficacy and safety data across ethnic groups in various regions of the world (40).

Risks Over the years, important lessons have been learned about the risks involved in assuming causal relationships between presumed surrogate endpoints and clinical

BIOMARKERS AND SURROGATE ENDPOINTS

outcomes

361

(6,41). These risks arise when (a) treatment intervention affects the

surrogate endpoint coincidentally, but not the desired clinical outcome because the causal pathways differ mechanistically; (b) many treatment interventions (e.g. different antihypertensive drugs) affect the same surrogate endpoint (e.g. blood pressure) but account for different changes in desired clinical outcome, such as the incidence of stroke, myocardial infarction, or congestive heart failure (42); (c) the prediction of clinical outcome following treatment intervention (e.g. propranolol) by a surrogate endpoint such as blood pressure is dependent on the demographic, environmental, disease, age, or genetic factors in the patient population (e.g. elderly African-American vs Caucasian or young African-American patients) (43); and/or (d) the proposed surrogate endpoint (e.g. ventricular premature beats) does not encompass other actions of the drug, in particular, those related to adverse reactions and safety. This was the case with the cardiac arrhythmia suppression trial (CAST) (6). Not only are drugs likely to have adverse effects that are not reflected by changes in a single biomarker, but a single biomarker may not indicate the full therapeutic benefit of a drug. For example, there appears to be an inflammatory component of coronary heart disease that accounts for the fact that the combination of C-reactive protein, an inflammatory biomarker, and lipid measurements predicts the relative risk of myocardial infarction better than when either marker is used alone (44). Pravastatin has been shown not only to lower serum cholesterol levels but to reduce plasma concentrations of C-reactive protein (45). It is likely that this apparent antiinflammatory effect of pravastatin accounts for the fact that, in the West of Scotland Coronary Prevention Study, the incidence of coronary heart disease events in patients treated with this drug was lower than that predicted from their cholesterol levels and a combination of other risk factors that did not reflect inflammatory response (46). It is clear that most biomarkers are unlikely to capture all the effects of a drug, and thereby fulfill the most stringent criterion for a surrogate endpoint, although it is desirable for the totality of evidence to lean in that direction. Consequently, there will probably be a trend in the future for clinical trials in this and in many other therapeutic areas to incorporate panels of biomarkers that can reflect more adequately the full spectrum of relevant potential therapeutic and toxic drug effects. For example, it seems likely that future clinical trials with statin HMG-CoA reductase inhibitors will incorporate both C-reactive protein and serum cholesterol as biomarkers.

FUTURE DIRECTIONS The rapid expansion of genomic information has focused considerable attention on the potential influence of genetic polymorphisms on response to drug therapy and has led to the development of pharmacogenomics as an important new field of scientific endeavor. Advances in pharmacogenomics are likely to result in the development of biomarkers that will play important roles as entry, stratification,

362

LESKO

#® ATKINSON

or exclusion criteria for clinical trials and, subsequently, will guide the optimization of drug prescribing for individual patients. In the area of cholesterol-lowering therapy, for example, it was found in the Regression Growth Evaluation Statin Study (REGRESS) that the Taq1B polymorphism in the CETP gene that codes for cholesterol ester transfer protein (CETP) affects not only the rate of progression of coronary atherosclerosis but also the extent to which patients benefit from pravastatin therapy (47). Coronary atherosclerosis appears to progress more rapidly in patients who are homozygous for the Taq1B allele. However, these also are the patients in whom pravastatin seems to be most effective. Patients homozygous for the B2 allele showed the least progression of atherosclerosis over the two-year study period, but there was no difference in disease progression between patients who were treated with pravastatin and those who received a placebo. Disease progression and pravastatin response were intermediate in patients who were B1B2 heterozygotes. This experience indicates that the efficiency of clinical trials could be enhanced considerably by using pharmacogenomic biomarkers to guide patient enrollment and stratification. To date, the potential role of pharmacogenomic biomarkers perhaps is best illustrated by the clinical development program and labeled indications for trastuzumab, a humanized monoclonal antibody against human epidermal growth factor receptor-2 (HER2) (48). HER2 is overexpressed in 25%—30% of patients with breast cancer and is associated with a more aggressive clinical course and shortened survival in these patients. One of the entry criteria for the more than 1000 women with breast cancer who participated in the Phase I, Phase II, and pivotal Phase III clinical trials of trastuzumab was that they have metastatic cancer that overexpressed HER2. Now that the new drug application for trastuzumab has been approved, the labeling states that trastuzumab therapy is indicated for patients with metastatic breast cancer whose tumors overexpress HER2 (49).

Despite the fact that measurement of HER2 overexpression has been central to the development and current clinical use of trastuzumab, only 20% of patients identified by this biomarker have responded to trastuzumab treatment (50). It is likely that additional biomarkers will be needed to increase this low predictive ability. In addition, even though FDA-approved immunohistochemical and flurorescence in situ hybridization (FISH) methods are now available for assessing HER2 overexpression, interpretation is operator dependent and quality-control programs are not in place to assess whether individual laboratories can perform the test accurately and reproducibly (50). Clearly, problems of this sort also will need to be surmounted as other pharmacogenomic biomarkers are incorporated in late-phase clinical development programs and in clinical practice. On the other hand, age, sex, diet, and other environmental exposures are contex-

tual factors that may affect the relationship between genetic substrate and disease

susceptibility. For example, a common nucleotide substitution (C677T) in the N(5,10)-methylenetetrahydrofolate reductase (MTHFR) gene reduces enzyme ac-

tivity and causes moderate elevations in plasma concentrations of homocyst eine (51). Because there is evidence that homocysteinemia may lead to atheroscl erosis

in individuals whose dietary intake of folic acid is inadequate, this provides a

BIOMARKERS AND SURROGATE ENDPOINTS

363

rationale for conducting studies in which the plasma homocysteine concentrationlowering effects of vitamin therapy are evaluated (52, 53). In these trials, plasma concentrations of plasma homocysteine wiil probably serve as a biomarker, and special attention may be focused on individuals who are homozygous for the C677T mutation in the MTHFR gene. Because the pathophysiology of most common diseases is multifactorial, singlegene mutations have generally only a limited correlation with the occurrence of a disease or with its progression or response to therapy. Accordingly, future developments in pharmacogenomics are likely to focus on the use of microarrays to study the differential expression of as many as 10,000 genes in a single experiment (54,55). The efficient identification and monitoring of relevant induced gene products has the potential to provide biomarkers that will be particularly useful in developing drugs for conditions that are difficult to track by currently available methods. Microarray techniques seem particularly well suited to monitoring disease progression or therapeutic response based on serial analysis of gene expression (SAGE) (56).

The large-scale study of gene expression marks the transition from structural to functional genomics and will focus increasing attention on the bioinformatic intrastructure needed to support microarray and other high-throughput methods of generating pharmacogenomic data (57,58). The massive amount of data that will be collected will need to be stored, processed, and analyzed by standardized relational database management systems, and a number of these have been described recently (56, 59, 60). The use of advanced bioinformatic techniques will facilitate

the serial transition of data to information and then to knowledge. Linked with pharmacogenomics, it can be anticipated to enhance the entire drug discovery and development process, from mapping disease genes, to stratifying patients and providing improved early-response monitoring in clinical trials, to ultimately making allele-specific therapeutics a clinical reality (59). Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED il Tempie RJ. 1995. A regulatory authority’s opinion about surrogate endpoints. In Clinical

Measurement

ed. WS

Nimmo,

in GT

Drug Tucker,

Evaluation, pp.

3-22.

New York: Wiley . Fleming TR, DeMets DL. 1996. Surrogate end points in clinical trials: Are we being misled? Ann. Intern. Med. 125:605-13 . Temple R. 1999. Are surrogate markers adequate to assess cardiovascular disease drugs? JAMA 282:790-95

4. NIH Definitions Working Group. 2000. Biomarkers and surrogate endpoints in clinical research: defintions and conceptual model. In Biomarkers and Surrogate Endpoints: Clinical Research and Applications, ed. GJ Downing, pp. 1-9. Amsterdam: Elsevier . Coplen SE, Antman EM, Berlin JA, Hewitt

P, Chalmers TC. 1990. Efficacy and safety of quinidine therapy for maintenance of normal sinus rhythm after cardioversion:

364

LESKO

#® ATKINSON

a meta-analysis of randomized trials. Circulation 82:1106—-16

control

. Echt DS, Liebson PR, Mitchell B, Peters RW, Obias-Manno D, et al. 1991. Mor-

tality and morbidity of patients receiving encainide, flecainide or placebo: the cardiac arrhythmia suppression trial. NV. Engl. J. Med. 324:781-88 . Riggs BL, Hodgson

SF, O’Fallon WM, Chao EYS, Wahner HW, et al. 1990. Effect

of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N. Engl. J. Med. 322:802-9 . Eisenberger MA, Nelson WG.

.

10. iI.

V2,

1996. How

much can we rely on the level of prostatespecific antigen as an end point for evaluation of clinical trials? A word of caution! J. Natl. Cancer. Inst. 88:779-81 Chobanian AV. 1983. The influence of hypertension and other hemodynamic factors in atherogenesis. Prog. Cardiovasc. Dis. 26:177-96 Russell RWR. 1975. How does bloodpressure cause stroke? Lancet 2:1283—85 Rolan P. 1997. The contribution of clinical pharmacology surrogates and models to drug development—a critical appraisal. Br. J. Clin. Pharmacol. 44:219-25 Scand. Simvastatin Surv. Study Group. 1994. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 344:1383-—89

leh, Bidart J-M, Thuillier F, Augereau C, Chalas J, Daver A, et al. 1999. Kinetics of serum tumor marker concentrations and usefulness in clinical monitoring. Clin. Chem. 45:1695—707 14. Gotto AM Jr, LaRosa JC, Hunninghake D, Grundy SM, Wilson PW, et al. 1990. The cholesterol facts: a summary of the evidence relating dietary fats, serum cholesterol, and coronary heart disease: a joint statement by the American Heart Association

and

the National

Heart,

Lung,

and Blood Institute. Circulation 81:1721— 38

i}. Levy G. 1993. The case for preclinical pharmacodynamics. In J/ntegration of Pharmacokinetics, Pharmacodynamics, and Toxicokinetics in Rational Drug Development, ed. A Yacobi, JP Skelly, VP Shah, LZ Benet, pp. 7-13. New York: Plenum 16. Sheiner LB, Steimer J-L. 2000. Pharmacokinetic/pharmacodynamic modeling in drug development. Annu. Rev. Pharmacol. Toxicol. 40:67—95 1h Eudeikis JR, Henthorn TK, Lertora JJL, Atkinson AJ Jr, Chao GC, Kushner W. 1982. Kinetic analysis of the vasodilator and ganglionic blocking actions of Nacetylprocainamide. J. Cardiovasc. Pharmacol. 4:303-9 18. Blumer J, Strong JM, Atkinson AJ Jr. 1973. The convulsant potency of lidocaine and its N-dealkylated metabolites. J. Pharmacol. Exp. Ther. 186:31—36 ek Breimer DD, Danhof M. 1997. Relevance of the application of pharmacokineticpharmacodynamic modelling concepts in drug development. The “Wooden Shoe’ paradigm. Clin. Pharmacokinet. 32:25967 . Smith LJ, Glass M, Minkwitz MC. 1993. Inhibition of leukotriene D,-induced bronchoconstriction in subjects with asthma: a concentration-effect study of ICI 204, 219. Clin. Pharmacol. Ther. 54:430-36 . Mildvan

D, Landay

A, De

Gruttola

V,

Machado SG, Kagan J. 1997. An approach to the validation

of markers for use in AIDS clinical trials. Clin. Inf. Dis. 24:76474

. Rolan PE, Parker JE, Gray SJ, Weatherley BC, Ingram J, et al. 1993. The pharmacokinetics, tolerability and pharmacodynamics of tucaresol (S89C80; 4[2-formyl-

3-hydroxyphenoxymethyl] benzoic acid), a potential anti-sickling agent, following oral administration to healthy subjects. Br. J. Clin. Pharmacol. 35:419-25 23% Arya R, Rolan PE, Wootton R, Posner J, Bellingham A. 1996. Tucaresol increases oxygen affinity and reduces haemolysis in

BIOMARKERS AND SURROGATE ENDPOINTS subjects with sickle cell anaemia. Br. J. Haematol. 93:817-21

24. Mol MJTM, Erkelens DW, Gevers Leuven JA, Schouten JA, Stalenhoef AFH.

1986. Effects of synvinolin (MK-733) on plasma

lipids in familial hypercholesterolaemia. Lancet 2:936-39

HS Reinoehl J, Frankovich D, Machado

C,

Kawasaki R, Baga JJ, et al. 1996. Probucol-

associated tachyarrhythmic events and QT prolongation: importance of gender. Am. HeartJ. 131:1184-91

26. Marschner IC, Collier AC, Coombs RW, D’ Aquila RT, DeGruttola V, et al. 1998.

Use of changes in plasma levels of human immunodeficiency virus type 1 RNA to assess the clinical benefit of antiretroviral therapy. J. Infect. Dis. 177:40-47 ile Deyton L. 1996. Importance of surrogate markers in evaluation of antiviral therapy for HIV infection. JAMA 276:159-60 28. US Gov. Print. Off. 1997. Code of Federal Regulations, Title 21, Vol. 5, Part 314, Subpart H. Washington, DC: US Gov. Print. Off. HW). US Gov. Print. Off. 1997. Code of Federal Regulations, Title 21, Vol. 5, Part 320, Sub-

part B. Washington, DC: US Gov. Print. Off.

30. Boissel J-P, Collet J-P, Moleur P, Haugh M. 1992. Surrogate endpoints: a basis for a rational approach. Eur. J. Clin. Pharmacol. 43:235-44 Bil Ronald Nancy Reagan Res. Inst. Alzheimer’s Assoc. Natl. Inst. Aging Work. Group. 1998. Consensus report of the working group on molecular and biochemical markers of Alzheimer’s disease. Neurobiol. Aging 19:109-16 32. Prentice RL. 1989. Surrogate endpoints in clinical trials: definition and operational criteria. Stat. Med. 8:431-40 BS: Hughes MD, DeGruttola V, Welles SL. 1995. Evaluating surrogate markers. J. Acquir. Immune Defic. Syndr. 10(Suppl. 2): S1-8 : 34. Reigner BG, Williams PE, Patel IH,

Steimer

JL, Peck C, van

365

Brummelen

P.

1997. An evaluation of the integration of pharmacokinetic and pharmacodynamic principles in clinical drug development. Experience within Hoffmann La Roche. Clin. Pharmacokinet. 33:142—52 30. US Gov. Print. Off. 1997. Federal Food, Drug and Cosmetic Act. Sect. 505(d)(5).

Washington, DC: US Gov. Print. Off. 36. US Gov. Print. Off. 1998. Food and Drug Administration Modernization Act. Sect. 115.

Washington,

DC:

US

Gov.

Print.

Off. SM, Food Drug Admin. Cent. Drug Eval. Res., Cent. Biol. Eval. Res. 1998. Guidance for industry: providing clinical evidence of effectiveness for human drug and biological products. http://www.fda.gov/cder/ guidances

38. Food Drug Admin. Cent. Drug Eval. Res., Cent. Biol. Eval. Res. 1999. Guidance for industry: metabolic drug-drug interaction studies in vivo—study design, data analysis and impact on dosing and labeling. http://www.fda.gov/cder/guidances 3D), Food Drug Admin. Cent. Drug Eval. Res.,

Cent.

Biol.

Eval.

Res.

1998.

Guidance for industry: pharmacokinetics in patients with impaired renal function—study design, data analysis and impact on dosing and labeling. http://www.fda.gov/cder/guidances 40. Int. Conf. Harmonization. 1997. ICH E5 guidance: ethnic factors in the acceptability of foreign clincal data. http://www. Fda.gov/cder/guidance/2293Eml.pdf 41. Packer M, Carver JP, Rodeheffer RT, Ivanhoe

RJ,

DiBianco

R, et ai.

1991.

Ef-

fect of oral milrinone on mortality in severe chronic heart failure. The PROMISE Study Research Group. N. Engl. J. Med. 325:1468-75

42. Psaty

BM,

Siscovick

DS,

Weiss

NS,

Koepsell TD, Rosendaal RF, et al. 1996.

Hypertension and outcomes research. From clinical trails to clinical epidemiology. Am. J. Hypertens. 9:178-83

366

LESKO

#® ATKINSON

43. Prichard BN, Graham BR, Cruickshank JM. 2000. New approaches to the uses of beta blocking drugs in hypertension. J. Hum. Hypertens. 14 (Suppl. 1):S63—68 44, Ridker PM. 1999. Evaluating novel cardiovascular risk factors: Can we better predict heart attacks? Ann. Intern. Med. 130:933SY 4S. Ridker PM, Rifai N, Pfeffer MA, Sacks F, Braunwald

E, Cholesterol

Newcomer LM, Upson B, et al. 1992. A prospective study of plasma homocyst (e)ine and risk of myocardial infarction in

DoE

Recurrent

Events Invest. 1999. Long-term effects of pravastatin on plasma concentration of C-reactive protein. Circulation 100: 230-35 46. West Scot. Coron. Prevent. Study Group. 1998. Influence of pravastatin and plasma lipids on clinical events in the West of Scotland Coronary Prevention Study (WOSCOPS). Circulation 97:1440-45 47. Kuivenhoven JA, Jukema JW, Zwinderman AH, de Knijff P, McPherson R, et al. 1998. The role of a common variant of the cholesterol ester transfer protein gene in the progression of coronary atherosclerosis. N. Engl. J. Med. 338:86—93 48. Steven Shak, Herceptin Multinatl. Invest. Study Group. 1999. Overview of the trastuzumab (Herceptin) anti- HER2 monoclonal antibody clinical program in HER2overexpressing metastatic breast cancer. Semin. Oncol. 26 (Suppl. 12):71-77 49. Physician’s Desk Reference. 2000. Montvale, NJ: Med. Econ. 54th ed.

50. Ravdin PM. 1999. Should HER2 status be routinely measured for all breast cancer patients? Semin. Oncol. 26 (Suppl. 12):117— 73) Sle Ellsworth DL, Manolio TA. 1999. The emerging importance of genetics in epidemiologic research. HI. Bioinformatics and statistical genetic methods. Ann. Epidemiol. 9:207-24 oye: Stampfer MJ, Malinow R, Willett WC,

54.

SEE

56.

Si

58.

.

US physicians. JAMA 268:877-81 Selhub J, Jacques PF, Bostom AG, D’ Agostino RB, Wilson PWF, et al. 1995. Association between plasma homocysteine concentrations and extracranial carotidartery stenosis. N Engl. J. Med. 332: 286-91 Debouck C, Goodfellow PN. 1999. DNA microarrays in drug discovery and development. Nat. Genet. 21:48-50 Hiltunen MO, Niemi M, Yla-Herttuala S. 1999. Functional genomics and DNA array techniques in atherosclerosis research. Curr. Opin. Lipidol. 10:515-19 Carlisle AJ, Prabhu VV, Elkahloun A, Hudson J, Trent JM, et al. 2000. Development of a prostate cDNA microarray and statistical gene expression analysis package. Mol. Carcinog. 28:12—22 Ermolaeva O, Rastogi M, Pruitt KD, Schuler GD, Bittner ML, et al. 1998. Data management and analysis for gene expression arrays. Nat. Genet. 20:19-23 Bassett DE Jr, Eisen MB, Boguski MS. 1999. Gene expression informatics—it’s all in your mine. Nat. Genet. 21:51—55 Braxton S, Bedilion T. 1998. The integration of microarray information in the drug development process. Curr. Opin. Biotechnol. 9:643-49

60. Scherf U, Ross DT, Waltham M, Smith LH, Lee JK, et al. 2000. A gene expression database for the molecular pharmacology of cancer. Nat. Genet. 24:236—44 6l. Atkinson AJ Jr. 2001. Physiological and laboratory markers of drug effect. In Principles of Clinical Pharmacology, ed. AJ Atkinson

Jr, CE

Daniels,

RL

Dedrick,

C Grudzinskas, SP Markey. New York: Academic. In Press

Annu. Rey. Pharmacol. Toxicol. 2001. 41:367-401 Copyright © 2001 by Annual Reviews. All rights reserved

CELLULAR RESPONSES TO DNA DAMAGE Chris J Norbury and Ian D Hickson Imperial Cancer Research Fund Laboratories, Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, OX3 9DS, United Kingdom; e-mail: norbury @icrf.icnet.uk, hickson@ icrf.icnet.uk

Key Words DNA damaging agents, DNA repair, genome instability, cell cycle checkpoints, apoptosis @ Abstract Cells are constantly under threat from the cytotoxic and mutagenic effects of DNA damaging agents. These agents can either be exogenous or formed within cells. Environmental DNA-damaging agents include UV light and ionizing radiation, as well as a variety of chemicals encountered in foodstuffs, or as air- and water-borne agents. Endogenous damaging agents include methylating species and the reactive oxygen species that arise during respiration. Although diverse responses are elicited in cells following DNA damage, this review focuses on three aspects: DNA repair mechanisms, cell cycle checkpoints, and apoptosis. Because the areas of nucleotide excision repair and mismatch repair have been covered extensively in recent reviews (1-6), we restrict our coverage of the DNA repair field to base excision repair and DNA double-strand break repair.

BASE EXCISION REPAIR The major forms of DNA damage arising from the actions of endogenous agents are (a) hydrolytic depurination, (b) hydrolytic deamination of cytosine and 5methylcytosine bases, (c) formation of covalent adducts with DNA, and (d) oxidative damage to bases and to the phosphodiester backbone of DNA. The vast majority of these ‘“‘small” lesions are repaired by the base excision repair (BER) pathway. The BER pathway has been reviewed in detail elsewhere (7-16) and is depicted in Figure 1. DNA glycosylases, of which there are several classes (see below), recognize abnormal DNA bases and catalyze hydrolytic cleavage of the N-glycosyl bond linking the base to the sugar. Following the generation of an apurinic/apyrimidinic (AP) site, in most cases the BER pathway can proceed utilizing a common set of proteins. The objective of the AP endonucleases and phosphodiesterases is to generate a single nucleotide gap containing 3’ hydroxyl and 5’ phosphate termini that permits a DNA polymerase to fill the gap. Finally, a DNA ligase can seal the remaining nick. Although this represents the primary pathway for BER, there are variations on the theme. First, instead of an AP endonuclease

cleaving the phosphodiester backbone 5’ to the AP site, some glycosylases remove 0362-1642/01/0421-0367$14.00

367

P

Excision of damaged base[A] by a DNA glycosylase

T

| Cleavage of the phosphodiester

backbone 5 to the AP site by an AP endonuclease

By

1 nucleotide gap repair

‘Long-patch’ repair Nick translation

Excision of 5 deoxyribose phosphate and filling of 1 nucleotide gap

by DNA PolB or DNA Pol 8/PCNA

by DNA Pol B

OH Pe P

7

A T

A T

al Ligation by DNA ligase III / XRCC1

a) | Cleavage of ’flap’ by FEN 1

and PCNA; ligation by ligaseI 2 A af

al

DNA DAMAGE RESPONSES

369

the damaged base and cleave the phosphodiester backbone 3’ to the resulting AP site via B-elimination (so-called AP lyases). In this case, the unsaturated sugar fragment left on the 3’ site of the nick is removed by a phosphodiesterase to generate the necessary single nucleotide gap. One source of this phosphodiesterase action is a second activity exhibited by the AP endonuclease (17-21). A second variation in BER is for the polymerization step of BER to involve a gap larger than one nucleotide (Figure 1). In this so-called “long-patch” pathway, approximately 2-10 nucieotides are excised and replaced by the combined actions of a DNA polymerase (usually Polé or ¢ but possibly Pol) proliferating cell nuclear antigen (PCNA), replication factor C (RF-C), and an endonuclease that cleaves “flap”

structures (FEN-1) (22-25). The long-patch pathway seems to predominate when repair is initiated at oxidized or reduced AP sites generated by X-rays or chemical agents, whereas the single nucleotide gap repair pathway occurs when “regular” AP sites are generated. In the single nucleotide gap pathway the XRCC1 protein directs specific interactions with both Polf and DNA ligase III (12, 26-29), and hence recruits this ligase to sites of ongoing repair. In long-patch BER the nick is probably sealed by DNA ligase I. Both branches of BER have been reconstituted in vitro (30-33).

The protein components of the BER pathway (see below) have been conserved both structurally and functionally during evolution, underscoring the vital role that BER plays in defending genome integrity. This role is further emphasized by the finding that disruption of BER function in mammalian cells is generally not compatible with viability. Thus, for example, mice deficient in HAP1

XRCC1

(34) and

(35) die during embryonic development.

DNA Glycosylases for Repair of “Inappropriate” DNA Bases Inappropriate bases include those not normally found in DNA, such as uracil, as well as normal bases found in the wrong context, such as thymine arising from deamination of 5-methyl cytosine. Uracil arises in DNA through deamination of cytosine residues, or through incorporation of dUTP during DNA replication. Uracil DNA glycosylase (UDG) excises uracil residues from DNA but not from RNA (7, 36). Probably because of a requirement to efficiently distinguish between uracil and thymine in DNA, which are very similar in structure, UDG has little flexibility in its range of substrates, being limited to uracil and its derivatives such as 5’ fluorouracil (7, 36). Recent studies have provided a structural basis for this

substrate specificity (37, 38). Owing to the shape of the active site pocket, the target uracil nucleotide must adopt an extrahelical conformation. This “nucleotide

< Figure 1 Schematic representation of the base excision repair pathway. A damaged base (A) is excised by a-DNA glycosylase to generate an apurinic/apyrimidinic site. After apurinic/apyrimidinic endonuclease cleavage the pathway bifurcates into single-nucleotide and long-patch routes. See text for details.

370

NORBURY

#® HICKSON

flipping” mechanism is emerging as a conserved feature of the action of several BER enzymes. Human cells contain another UDG activity, designated hSMUGI, which prefers ssDNA containing uracil residues as a substrate (39). The DNA of many organisms, including man, includes cytosine residues methylated at the C° position. Upon deamination of 5-methylcytosine, thymine is generated in the context of a G:T mispair. Any repair system for correction of this mispair has two obvious requirements: The G:T mispair must be repaired to G:C, and thymine residues in A:T pairs must not serve as substrates. The glycosylase responsible for repair of G:T mispairs is thymine DNA glycosylase (40). However, thymine DNA glycosylase is not specific for thymine residues, because it also recognises uracil when mispaired with guanine (41). This “uracil glycosylase” activity has been conserved in organisms that do not methylate cytosine residues.

Glycosylases for Repair of Oxidized DNA Bases Base oxidation in cellular DNA can arise following exposure to ionizing radiation, radiomimetic chemicals, and intracellular reactive oxygen species (ROS) (7-11). In Escherichia coli oxidized purines are excised by the FPG protein (also known as MutM and FAPY glycosylase) (42). FPG removes imidazole ring—opened derivatives, such as FAPY-guanine and FAPY-adenine and 7,8-dihydro-8 oxoguanine (8-OG) (Figure 2) (7). 8-OG can mispair with A and therefore generates GC > TA transversion mutations. fpg mutants show an increased frequency of GC > TA mutations, consistent with a defect in 8-OG repair (7, 43, 44).

Yeast and mammals contain proteins that perform roles analogous to those of FPG. The Saccharomyces cerevisiae Ogg1 protein has a substrate specificity similar to that of FPG, although their primary sequences are unrelated (45, 46). Instead, Ogg1 shows sequence similarity to the E. coli endonuclease (endo) III protein (see below). The human homolog of Ogg! excises 8-OG paired with cytosine by flipping the 8-OG residue into an extrahelical configuration (47). The ability of Ogg1 to discriminate between 8-OG and guanine appears to be conferred by a single hydrogen bond between an active site glycine residue and the 8-OG moiety. 8-OG residues that have escaped repair prior to DNA replication can mispair with adenine. 8-OG:A mispairs are relatively poor substrates for Ogg1, presumably

to exclude the use of the adenine residue as a template during DNA repair synthesis. All cells, therefore, express an enzyme that converts 8-OG:A mispairs to 8-OG:C | through acting as an adenine glycosylase (48-50). The gene encoding this activity in E. coli is mutY (51-54). Because cells counteract 8-OG residues in DNA through the combined actions of FPG/Ogg1 and MutY, fog mutY double mutants show an extremely high rate of GC + TA mutations (48). The MutY protein shows Significant sequence similarity with E. coli endo III, including a conserved helixhairpin-helix (HhH) motif and cysteine residues that create a binding site for an iron-sulfur cluster of the (4Fe-4S)?* type (55). MutY is also conserved in evolu-

tion, and the human homolog, MYH, has a very similar substrate specificity to that of its bacterial counterpart (56, 57).

N

Z

Ox0-g-O1pAUIp-g‘7

HN

(Ot

“HN

YNG

!

VNGN

NH

le

YNG

NSN

Bs .O

‘ee. 1

N

HO

VNG

ate

ZN

"HN

OXx0-g-OlpAYIp-g‘7

CTHN

ne NON

O.

9ulsoUl

o

cH N

Hal

Ox0-g-O1pAyIp-g‘/

NON

a

suluape

OH

jloeun

sojdur Jo exg oy} samyon Jo ns

paziprxosutind

Sayesq JOJ ns 24}

[330/SSPID NdAJO

~p-OUlWIP-9'Z “p-OUIWRID-9'¢ p -G-OUILUEID-9' -hxouphu-G-AxoupAy -N-S-AxoupAy -N-S-AxoupAy aulpiAD aulpluAdopiue -OPILWEWO}|AYau Woy SuUlpIULACOpie (auiuenb-Ade}) SUIPIWULWAd (auiuepe-Adey)

suluenb

VNG

andy =z

Woy

‘

N

HN

0 TL Y

N

-AxoupAyu-s "SOWIAZUD

DNA DAMAGE RESPONSES

371

372

NORBURY

#® HICKSON

Endo III excises numerous oxidation products of pyrimidines (7, 58-60). Many substrates result from reaction of ROS at the 5,6-double bond, and include cytotoxic lesions such as thymine and cytosine glycols. However, endo III also recognizes ring-opened, ring-contraction and ring-fragmentation products of cytosine and thymine residues (Figure 3). Budding yeast express two endo II homologs, Ntgl and Ntg2, the former of which lacks the characteristic iron-sulfur cluster

(61,62). The substrate specificities of these two enzymes are similar but nevertheless different from that of endo III. The only known human endo III homolog, hNTH1, displays the major structural motifs (HhH and Fe-S cluster) found in endo III and has a substrate specificity similar to that of its bacterial counterpart (63, 64).

Glycosylases for Repair of Alkylated Bases A selection of the wide range of products of base alkylation is shown in Figure 4. These adducts are repaired by a glycosylase designated 3-methyladenine DNA glycosylase (3-MAG) (reviewed in 7, 13). Two such enzymes, AlkA and Tag, exist in E. coli, the former being inducible as part of the adaptive response to DNA alkylation. A single major 3-MAG activity with a broad substrate specificity similar to that of AIkA exists in eukaryotes (7, 13). 3-MAG also efficiently excises 1,N°-ethenoadenine (65). 3-MAG is important for repair of 3-methyladenine and 1,N°-ethenoadenine residues in vivo, and for the protection of mammalian

cells

against the cytotoxic effects of certain alkylating agents (66-68). The ability of AlkA to recognize both positively charged and extended aromatic substrates derives from the presence of numerous aromatic side chains lining the catalytic cleft (69, 70). The crystal structure of AIkA indicates a nucleotide flipping mechanism of action. Alkylation at the O° position of guanine is potentially highly mutagenic because of the efficiency with which O°-alkylguanine mispairs with thymine, leading to GC — AT transition mutations. This lesion, which is also cytotoxic, is repaired by a conserved and specialized protein, O°-alkylguanine DNA alkyltransferase (O°-AT), reviewed elsewhere (71,72). The expression of O°-AT allows cells to resist the cytotoxic and mutagenic effects of a variety of alkylating agents, many of which are used in chemotherapy (71, 72).

AP Endonucleases in BER AP endonucleases counteract the cytotoxic and mutagenic potential of AP sites. These lesions arise via spontaneous hydrolysis of the N-glycosyl bond linking a base to the phosphodiester backbone, following damage to DNA by alkylating agents or ROS, and through the action of DNA glycosylases (7, 73). There are two major families of AP endonucleases, termed the exonuclease (exo) II and endo IV families, which derive their name from the two AP endonucleases expressed in E. coli (7,9, 11,73). In addition to AP endonuclease activity, these enzymes possess phosphodiesterase activity for removal of fragmented sugar residues, such as phosphoglycolate, from the 3’ terminus of strand breaks induced by oxidants.

O ~S

| AY

UBXO||\Y

ON

HN

jloBIA JOoA}B

H

O

OH

eoun

uu jOoAjH

sajduexy Jo oy)

OH

ONOH

HN

SID) 8B (SUB1]

SHN

|AYay

KK

samnjons Jo

BUIpIADOIPAYIpP

-9‘G-AxoipAu-9

OyN

HN

\Auoye}

O@NEZO

ES x

poziprxo

auIpIJOZEpI

ouIprULAd SaBSqns JO} OY}

wil

aansiy ¢=

Hi

yee

O

ean

H

a

-|[Aoweqied-;-Suesl -AxOupAU-S

OH

ulojuepAy

Qo OH

NH

ulojuepAyjAUyow

ron

O

HN

Sis HN

-G-AXOIpAY-G

OPUS[]]

-AxOupAyIp-G‘p-OXO-Z

IES

SLIDJO

QuIpNAD

OH

O

~AxO.pAU-S

ON

OH

8H

HON

eae!

aulwAyy -O1pAYIP-9°G -9-AxOpAy-9-AxoipAy-s

HN >844 OH S

"SOWAZUD

joein -AxoJpAu-S

QO.

OH

“HN

—-SulLUAYOupAY

H N

oo OQ.

O

O

Rae

HN

O

woeunoupAy

H ON

Fi

OL

O

HN LN :

DNA DAMAGE RESPONSES

9SID) (SUBI]

aD

#® HICKSON NORBURY

374

zi ousuie-

-OURUJO-E', N auluenb N'L auluenb

G iS

-OUBYIO-E', N auluenb

oa

auluepe -OUdy}e-,N‘I

N

-(\AyjeAxoupAy-Z)-2 N°H

(@)

O

=p ans

vie

—,

NOH

eae

?HN

Ae)

~p N

Cory

"HO yna N- &e

—SUlUape|AYjJa-E

O

ep

auiuenb

G HN

-(\AyyoAxouja-Z)-Z

O°HO°HO

a

“ah?

bs Ne

auluenfiAujaw-g

eb.

de “HN ne Ie

oe

auluenbiAyjew-2

ily,

3H

wo,

O

auluenb

“SOUIAZUDJO SSPID OYIA-E OY) 1OJ Saye.NSQns Jo somnjonys oy) Jo sojdwiexg

NG NG

euiuenb

ue

Ne

Io) N

ID

ars

all ue tae Ie? CF O

ol)

ue

-(\AuJaosojyd-z)-2

HN

cd HN

| Queue -(\AuenB-/)-siq-2'

DNA DAMAGE RESPONSES

SiS

This activity can also remove the 3’ terminal groups left by AP lyases following £-elimination at AP sites (see above). AP lyase activity is a feature of the endo III and FPG families of glycosylases (7, 9, 11). The catalytic mechanism of action of both families of AP endonucleases has been revealed by X-ray crystallography. The human homolog of exo II, HAP1 (APE/Ref-1), shows structural similarity to DNase I (74). The abasic deoxyribose lies in an extrahelical conformation and is stabilized in the HAP1 active site by interactions with specific residues (75). Several active site residues required for the hydrolysis reaction have been identified by site-directed mutagenesis (17, 73, 76-80).

Structural studies of E. coli endo IV have revealed that both the abasic

deoxyribose and its partner nucleotide are flipped out of the duplex (81). Structural and site-directed mutagenesis studies have suggested that HAP1 might coordinate different steps of the BER process through engaging in protein:protein interactions (12). HAP1 interacts with and displaces glycosylases that are bound to the AP sites generated by their action. Moreover, following cleavage of the phosphodiester backbone, HAP 1 remains bound to the nicked DNA until direct interactions with Polf initiate the subsequent phosphodiesterase/ polymerization steps, helping to explain how BER can be such an efficient and rapid process in vivo.

Late Stages in BER Pol can catalyze both the 5’ phosphodiesterase and polymerization steps in BER (Figure 1). Perhaps surprisingly, recent data using Polf-deficient mouse cells indicate that only the phosphodiesterase action of Polf is essential for protection against methyl methane sulfonate (82). As well as binding to a HAP1:DNA complex, Polf makes interactions with the XRCC1 protein, which appears to play a scaffolding role in BER rather than any specific enzymatic function. XRCC1 then recruits DNA ligase III to sites of ongoing repair (83).

DNA DOUBLE-STRAND BREAK REPAIR DNA double-strand breaks (DSBs) arise in cellular DNA via a number of differ-

ent routes, including exposure to ionizing radiation and radiomimetic chemicals, through the interaction of a replication fork with a single-stranded break in the template, and in a programmed manner during meiosis and immunoglobulin gene rearrangement (reviewed in 84,85). The three primary pathways for repair of DNA DSBs (Figure 5) are discussed below.

Homologous Recombination Knowledge of the mechanisms of homologous recombination (HR) and the protein components of the HR pathway has come primarily from genetic analyses in

yeasts and subsequent biochemical studies. In S. cerevisiae HR 1s dependent upon

376

NORBURY

#® HICKSON

DNA double-strand break repair

Y~ rays/

tLradiomimetic

ae | Formation of dsb

=

Strand

Endo-

invasion

nucleolytic trimming

pee

Bee . Ligation

DNA DAMAGE RESPONSES

377

the RADS52 epistasis group of genes, which includes RAD50, RADS1, RADS2, RAD54, RAD55, RAD5S7, RADS9, MREI1, and XRS2 (reviewed in 84, 86). Rad51 is the eukaryotic homolog of E. coli RecA, which initiates HR through catalyzing homologous pairing and DNA strand exchange. Rad51 has a substantially reduced catalytic rate compared with RecA. To compensate for this apparent catalytic deficiency, eukaryotic Rad51 acts in concert with other proteins. Replication Protein A (RPA) enhances the efficiency of DNA strand-exchange reactions mediated by Rad51, but only if added to reactions after Rad51 has created the characteristic nucleoprotein filaments on ssDNA in which strand exchange occurs (87-91). Rad52 and the Rad55/57 heterodimer both counteract the inhibitory effect of RPA, presumably by mediating exchange of Rad51 for RPA on ssDNA (87-91). Moreover, Rad54 can stimulate Rad51-catalyzed pairing of homologous DNA molecules in vitro (92, 93). Rad54 is a DNA-dependent ATPase of the Swi/Snf family whose precise role is unclear (94). Several mammalian Rad55/57-related proteins have been

identified, but whether any of these act as functional homologs remains unclear. Two of these, XRCC2 and XRCC3, were identified by functional complementation

of X-ray-sensitive hamster cell mutants defective in DSB-induced HR (95). HR functions are vital for repair of DSBs (reviewed in 84-86). In a recent

model proposed by Van Dyck et al (96) Rad52 acts to initiate DSB repair through binding DNA ends. This binding protects DNA ends from exonucleolytic attack and facilitates end-to-end joining. Rad52 then loads Rad51 onto the DNA through direct protein:protein interaction. This end-binding function probably explains why Rad52 also acts during the single-strand annealing (SSA) pathway (see below). It

is possible that Rad52 acts in partial competition with the Ku DNA end-joining complex to direct DSB repair down the HR pathway instead of the NHEJ pathway (Figure 5). In S. cerevisiae RAD52 is essential for DSB repair and HR (86). However, RAD52~‘~ mice are viable and fertile, and their cells do not display deficiency in DSB repair (97). RAD54~/~ mice are also viable and fertile, despite having a cellular defect in the repair of DNA cross-links. These mice are hypersensitive to y-rays at the embryonic, but not adult, stages (98,99). In contrast, RAD5/ a

mice die during early embryonic development. The roles played by other DSB repair proteins are not fully elucidated. In yeast the Rad50/Mrel 1/Xrs2 complex provides nuclease activity and is implicated in

10 oligonucleotides. This information does not jibe with the one in eight (12.5%) activity shown in a number of careful studies. As suggested previously, it is possible that the published literature represents a selection bias and that actually up to eightfold more unique oligomers than the 1655 (i.e. 95%, whereas levels of control PKC-A/i mRNA were not markedly changed. In contrast to the original report, identical co-downregulation of PKC-a@ and PKC-¢ was observed also when

Lipofectin™ was used as the carrier. This co-downregulation may be caused by irrelevant cleavage. There is a contiguous 11-base match between Isis 3521 and the PKC-¢ mRNA, which, in theory and certainly in cell-free systems, is more than sufficient for RNase H competency. Isis 3522, a 20-mer phosphorothioate oligonucleotide which is targeted to the 5’ region of the PKC-~@ mRNA, inhibits PKC-a@ protein and mRNA expression, but does not inhibit PKC-¢ expression. Here, only a 4-base region of complementarity exists between the oligonucleotide and the PKC-¢ target. Such a small region of homology is probably insufficient for RNase H competency, and so irrelevant cleavage does not occur. However, the relative ease with which potential irrelevant cleavage was found in this system suggests that it is a far more common problem than usually contemplated. If so, then it may be very difficult to relate an observed phenotype to a specific knockout, precisely because the knockout is not as specific as assumed.

Potential Elimination of Irrelevant Cleavage Conceptually, this is straightforward; if RNase H activity is eliminated, then the

potential for irrelevant cleavage is also eliminated.

However, the elimination of

410

LEBEDEVA

® STEIN

RNase H competency unfortunately creates additional problems. A partial decrease in the RNase H competency of an oligonucleotide will occur if the backbone charge is reduced. However, if charge reduction is excessive, insolubility and formulation

difficulties will be the result. Furthermore, the ability of cationic carriers to deliver the oligonucleotide will be drastically diminished, and antisense efficacy will be compromised. Uncharged molecules are usually not very active antisense effectors, because they must depend on steric blockade of translation for efficacy. This is usually ineffective because the 80S elongating ribosomes have intrinsic unwinding ability (60,61), and they can probably read through the steric

block. Charge may be retained by substituting nuclease-resistant but non-RNAse H-competent 2’-O-alkyloligoribonucleotides throughout the length of the backbone. This retains the property of aqueous solubility, but efficacy will most likely be greatly diminished, probably also because of ribosome-promoted unwinding. Perhaps the most effective strategy yet devised to reduce yet not eliminate RNase H competency is the use of so-called “gap-mers,” which have six to eight 2’-Oalkylloligoribonucleotides (alkyl = methyl or methoxyethoxy) at the 3’ and 5’ termini and, to retain RNase H activity, a central core of six to eight oligodeoxyribonucleotides. For purposes of nuclease resistance, the entire backbone contains phosphorothioate linkages. These oligonucleotides tend to be somewhat expensive to synthesize, and, although extremely promising, their use is not yet as widespread as perhaps it should be. Another strategy for eliminating irrelevant cleavage calls for the replacement of RNase H competency by RNase P competency [see Altman (62) and Forster & Altman (63)]. RNase P, like RNase H, is a ubiquitous cellular enzyme, functioning to cleave the S’ terminus of precursor transfer RNA (tRNA) molecules to generate a mature tRNA. If a synthetic oligonucleotide [called an external guide sequence (EGS)] targeted to an mRNA is designed to mimic certain structural features of precursor tRNA (i.e. to incorporate a stem and seven-residue loop hairpin, in

addition to the two hybridizing arms linked to the 3’ and 5’ ends of the stem loop), then in cell-free systems RNase P will cleave the target RNA at the junction between the single-stranded leader sequence and the duplex formed with the EGS (64). Ma et al (64) have recently developed a series of nuclease-resistant,

serum-

stable EGSs that efficiently induce RNase P cleavage in vitro of a 29-mer derived from the hepatitis B virus genome. More recently we have used an EGS whose sequence was extrapolated from Isis 3521 to downregulate PKC-a expression in T24 bladder carcinoma cells. Two carriers were used, Lipofectin™ and

LipofectACE™ (GIBCO BRL; Rockville, MD), with identical results—almost 95% downregulation of protein and mRNA expression. In contrast to what we observed with phosphorothioate oligonucleotides, this occurred in the absence of any downregulation of PKC-¢ mRNA or protein expression. Excellent downregulation of bel-xL protein and mRNA expression in T24 cells was also demonstrated by the use of the EGS technology. However, EGSs are presently difficult and expensive to synthesize.

-

ANTISENSE OLIGONUCLEOTIDES

411

Antisense Oligonucleotides Targeted to bcl-2: Tissue Culture Successes, Problems of Data Interpretation, and Initiation of Clinical Trials Bcl-2 is an important antiapoptotic protein found in a wide variety of human cancer cells. Its inhibition would theoretically sensitize cells to cytotoxic chemotherapy, and a number of at least partially successful studies in tissue culture and experimental animals have been performed that have led to the initiation of clinical trials. Initially, antisense phosphodiester and phosphorothioate oligonucleotides targeted to the bcl-2 mRNA were used to inhibit the growth in culture of 697 human leukemia cells (65). The oligonucleotide targeted the translation initiation site of the human bcl-2 mRNA. Both classes of oligomer decreased cell proliferation: The phosphorothioates were more potent inhibitors, but, given the benefit of a decade of research, the experiments with the phosphodiesters should now probably be discounted. It was proposed the phosphorothioate bcl-2 antisense oligonucleotides induced cell death through sequence-specific mechanisms, but in retrospect this is difficult to accept because they were not delivered to the cells by a carrier and the concentrations were high (25 jxM). These problems (i.e. lack of carrier delivery, high concentration, and use of only a single control) recur in almost all of the reports that target bcl-2 except for a few that are specifically noted below. Furthermore, this oligonucleotide was empirically chosen to target the translation initiation site of the bcl-2 mRNA, a method no longer considered to be an appropriate one for identifying active oligonucleotides. In other work with the phosphorothioate oligonucleotide used by Reed et al (65), Bcl-2 protein expression was inhibited in acute myeloid leukemia cells. This was associated with decreased duration of cell survival and a diminution in the number of clonogenic cells in culture (66). Subsequent to bcl-2 downregulation, treatment with daunorubicin and 1-6-D-arabino-furanosyl-cytosine (AraC) increased cell kill (66, 67). Durrieu et al (67) used the same oligonucleotide as Reed et al (65), but the concentration was lower (1 44M) because delivery was accomplished with cationic lipids. However, only a single control oligomer was used. In another set

of experiments, in 7 of 17 samples of myeloblasts from acute myeloid leukemia patients, treatment with the antisense oligonucleotide resulted in a significant decrease in the expression of Bcl-2 protein (68). This, in turn, was accompanied by increased apoptosis in response to AraC. These experiments were more rigorously controlled than those of Reed et al (65), and the oligonucleotide concentration was

lower (5 4M). However, no carrier was used for delivery, introducing questions of specificity. . Phosphodiester and phosphorothioate antisense oligonucleotides directed against the first six codons of the Bcl-2 mRNA

(denoted G3139;

Genta, Inc.,

Lexington, Mass.) were also successfully used in non-Hodgkin lymphoma cell lines. A putatively specific reduction in Bcl-2 mRNA levels within | day of treatment was demonstrated by Kitada et al (69), in SU-DHL-4 cells.

A commensurate

412

LEBEDEVA

® STEIN

reduction in Bcl-2 protein level occurred only after 3 days, presumably because of the long half-life of the Bcl-2 protein, and this reduction was associated with a drastic decrease in cellular viability. Treatment of RS11846 lymphoma cells with G3139 similarly reduced Bcl-2 expression and sensitized these cells to AraC and methotrexate (70). However, the oligonucleotide concentration was extreme

(200 M), no carrier was used for delivery, and the authors of that report used only a single control sequence. Overexpression of Bcl-2 has also been observed in human prostate cancers (71). Although Bcl-2 expression in normal prostatic epithelial cells is low or absent, Bcl-2 is ultimately upregulated in ~35% of the tumors from patients after progression to androgen independence (72). The forced overexpression of Bcl-2 protein in LNCaP prostate cancer cells increased their in vivo tumorigenic potential and resistance to apoptosis (73). Treatment of LNCaP (74) and Shionogi tumor cells (75) in vitro with G3139

inhibited Bcl-2 expression via a dose-dependent and sequence-specific manner.

The authors used the cationic lipid Lipofectin'™ for delivery, and measured Bcl-2 protein and mRNA levels. Unfortunately, they used only a single control oligonucleotide, which diminishes the credibility of the experiments. In other experiments, antisense bcl-2 oligonucleotide treatment substantially enhanced paclitaxel and docetaxel chemosensitivity in a dose-dependent manner. However, characteristic apoptotic changes were demonstrated only after combined treatment (22, 76) and not after treatment with the Bcl-2 oligomer alone. Miyake et al have tested the efficacy of G3139, administered as an adjuvant after castration, to delay the time to androgen-independent recurrence in the androgendependent mouse Shionogi tumor model (22, 74-76). Systemic administration of G3139 beginning | day postcastration in mice bearing the Shionogi (75) or LNCaP (74) tumors resulted in a more rapid regression of tumors and a significant delay in the emergence of androgen-independent recurrent tumors than in animals not exposed to the oligonucleotide. In addition, despite significant reductions of Bcl-2

expression in tumor tissues, G3139 had no effect on Bcl-2 expression in normal mouse organs (75). Adjuvant in vivo administration of G3139 and micellar paclitaxel after castration resulted in a statistically significant delay in the appearance of androgen-independent, recurrent tumors compared with administration of either agent alone (22). When this regimen was given to mice bearing established Shionogi tumors, tumor regression was far more dramatic when compared with treatment with either agent alone (76). These findings illustrated the potential utility of an antisense bel-2 approach for prostate cancer in the adjuvant setting when combined with androgen ablation and taxane treatment.

Antisense bcl-2 Oligodeoxynucleotides In Vivo Phosphorothioate oligonucleotides (including G3139) have been infused intra-

venously or intraperitoneally into patients taking part in clinical trials, and their pharmacokinetics have been extensively studied (77,78). These oligonucleotides

ANTISENSE OLIGONUCLEOTIDES

413

are stable for 48 h (15-50%), at which time levels may still be detected in tissues. Plasma clearance is biphasic, with half-lives on the order of 15—25 min and 20-40 h.

For G3139, after an intravenous bolus dose of ~5 mg/kg, the plasma elimination half-life is 22 h (79). Cotter et al (80) obtained cells from a patient with B-cell lymphoma bearing the t(from 14 to 18) translocation and, after treatment with G3139 in tissue cul-

ture, demonstrated downregulation of Bcl-2 protein and induction of apoptosis. They then xenografted these cells into a severe combined immunodeficient (SCID) mouse model (80). Preinoculation treatment of the cells with G3139 ablated the ability of the cells to grow in the host. Congruently, G3139 almost completely

abolished lymphoma growth in SO of 60 treated mice (83%) after 2 weeks. Extension of treatment to 3 weeks completely eradicated lymphoma in all animals, even at the PCR level (81). The effect appeared specific, but one confounding fact is that G3139 contains two CpsG sequences, which can activate residual natural killer (NK)-cell activation (82). However, the experiments also were repeated in NOD/SCID mice which lack NK-, B-, or T-cell activity, and similar efficacy was observed (83).

A phase I study in non-Hodgkin lymphoma patients with high Bcl-2 expression (84, 85) was then undertaken.

No treatment-related toxicities were observed at

doses of 5 mg/kg/day (84). One patient remained in remission for 3 years after starting treatment in the absence of additional therapy. No other patients achieved a complete response. In two patients, computer tomography scans demonstrated some reduction in tumor size (84), but partial response status was not achieved. Nine patients had disease stabilization and at least two of them had symptomatic improvement. Nine additional patients had progressive disease (85). To use a surrogate pharmacodynamic measurement, levels of Bcl-2 protein were measured by flow cytometry in peripheral blood samples, and these levels were found to be reduced in 7 of 16 assessable patients. Additional trials in this disease are currently in progress. G3139 also appears at this time to be able to chemosensitize melanoma cells. Treatment of melanoma cells in vitro with G3139 (200 nM) in complex with a cationic lipid led to what was proposed to be a sequencespecific and dose-dependent downregulation of the Bcl-2 mRNA (20). Two control

oligonucleotides were ineffective. Chemosensitization of human melanoma cells also occurred after treatment of melanoma xenografts with G3139 in SCID mice, and a combination of G3139 and dacarbazine resulted in complete ablation of the tumor in three of six animals. In a currently accruing phase I-II study, dacarbazine and G3139 are being combined in patients with advanced disease, including patients resistant to prior single-agent dacarbazine (86). Initial results from this study indicate that G3139 can reduce Bcl-2 expression in the xenografted melanoma deposits and that combined therapy with dacarbazine is well tolerated. G3139 in combination with docetaxel is also in phase J trials for patients with advanced breast cancer and other solid tumors, and at this point the regimen’s toxicity has been tolerable (87). Tumor response was observed in two patients with breast cancer.

414

LEBEDEVA

® STEIN

The safety data from clinical trials support further clinical development of G3139, both as a single agent and in combination with cytotoxic agents. G3139 now is also in phase I/IIA trial for patients with androgen-independent prostate cancer, in which setting the G3139 will also be combined with docetaxel. Trials in other advanced solid-tumor malignancies have been planned (88). The relative clinical success with G3139 to date is tremendously encouraging: Because its toxicity appears to be relatively low, consisting mainly of fatigue and thrombocytopenia, this molecule may represent a truly novel advance in the treatment of human solid tumors, irrespective of its mechanism of action, which may or may not be related to inhibition of bcl-2 expression. However, from the point of view of the patient whose tumor may be responding to therapy, questions of mechanism are wholly moot points.

ACKNOWLEDGMENT C.A. Stein is a member of the Scientific Advisory Board of Genta, Inc. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Stein CA, Cheng Y-A. 1997. Antisense inhi6. bition of gene expression. In Principles and Practice of Oncology, ed. VT DeVita, SA Rosenberg, S Hellman, pp. 3059-74. New York: Lippincott 2. Eder PS, DeVine RJ, Dagle JM, Walder JA. 1991. Substrate specificity and kinetics of | 7. degradation of antisense oligonucleotides by a 3’ exonuclease in plasma. Antisense Res. Dey. 1:141-51 3. Vaerman JL, Moureau P, Deldime F, Lewalle 8. P, Lammineur

C, et al. 1997.

Antisense

oligodeoxyribonucleotides suppress hematologic cell growth through stepwise release of deoxyribonucleotides. Blood 90:331-39 4. Miller P. 1992. Antisense oligonucleotide methylphosphonates. In Antisense RNA and DNA, ed. JAH Murray, pp. 241-53. New York: Wiley-Liss 5. Miller PS, Ts’o PO. 1987. A new approach to chemotherapy based on molecular biology and nucleic acid chemistry: Matagen (masking tape for gene expression). Anticancer Drug Des. 2:117—28

Crooke ST, Lemonidis KM, Neilson L, Griffey R, Lesnik EA, et al. 1995. Kinetic characteristics of Escherichia coli RNase H1: cleavage of various antisense oligonucleotide-RNA duplexes. Biochem. J. 312:599-608 SummertonJ, Weller D. 1997. Morpholino antisense oligomers: design, preparation, and properties. Antisense Nucleic Acid Drug Dev. 7:187-95 Akhtar S, Basu S, Wickstrom E, Juliano RL.

1991. Interactions of antisense

DNA oligonucleotide analogs with phospholipid membranes (liposomes). Nucleic Acids Res. 19:5551-59 9. Tonkinson JL, Stein CA. 1994. Patterns of intracellular compartmentalization, trafficking and acidification of 5’-fluorescein labeled phosphodiester and phosphorothioate oligodeoxynucleotides in HL60 cells. Nucleic Acids Res. 22:4268-75 10. Stec WJ, Zon G, Egan W, Stec B. 1984. Automated solid-phase synthesis, separation, and stereochemistry of phosphorothioate

ANTISENSE OLIGONUCLEOTIDES

analogues of oligodeoxyribonucleotides. J. Am. Chem. Soc. 106:6077—80

ke Koziolkiewicz ska

A,

M, Wojcik M, KobylanKarwowski B, Rebowska B,

et al. 1997. Stability of stereoregular oligo(nucleoside phosphorothioate)s in human plasma: diastereoselectivity of plasma 3’-exonuclease. Antisense Nucleic Acid Drug Dev. 7:43-48 12: Eckstein F. 1985. Nucleoside phosphorothioates. Annu. Rev. Biochem. 54:367-402 3% Stein CA, Subasinghe C, Shinozuka K, Cohen JS. 1988. Physicochemical properties of phosphorothioate oligodeoxynucleotides. Nucleic Acids Res. 16:320921 14. Geiger T, Muller M, Dean NM, Fabbro D. 1998. Antitumor activity of a PKCalpha antisense oligonucleotide in combination with standard chemotherapeutic agents against various human tumors transplanted into nude mice. Anticancer Drug Des. 13:35-45 . Dean NM, McKay R. 1994. Inhibition of protein kinase C-alpha expression in mice after systemic administration of phosphorothioate antisense oligodeoxynucleotides. Proc. Natl. Acad. Sci. USA 91:11762-66 16. Monia BP. 1997. First- and secondgeneration antisense inhibitors targeted to human c-raf kinase: in vitro and in vivo studies. Anticancer Drug Des. 12:327-39 lee Cowsert LM. 1997. In vitro and in vivo activity of antisense inhibitors of ras: potential for clinical development. Anticancer Drug Des. 12:359-71 18. Chiang MY, Chan H, Zounes MA, Freier SM,

Lima

WF,

et al.

1991.

Antisense

oligonucleotides inhibit intercellular adhesion molecule 1 expression by two distinct mechanisms. J. Biol. Chem. 266:18162-71 1), Ziegler A, Luedke GH, Fabbro D, Altmann KH, Stahel RA, et al. 1997. Induction of

apoptosis in small-cell lung cancer cells by an antisense oligodeoxynucleotide targeting the Bcl-2 coding sequence. J. Natl. Cancer Inst. 89:1027-—36

415

20. Jansen B, Schlagbauer-Wadl H, Brown BD, Bryan RN, van Elsas A, et al. 1998. bel-2 antisense therapy chemosensitizes human melanoma in SCID mice. Nat. Med. 4:232-34 2. Cotter FE. 1999. Antisense therapy of hematologic malignancies. Semin. Hematol. 36:9-14 22. Miayake H, Tolcher A, Gleave ME. 2000. Chemosensitization and delayed androgen-independent recurrence of prostate cancer with the use of antisense Bcl-2 oligodeoxynucleotides. J. Natl. Cancer Inst. 92:34-41 Mah Leech SH, Olie RA, Gautschi O, SimoesWust AP, Tschopp S, et al. 2000. Induction of apoptosis in lung-cancer cells following bcl-xL anti-sense treatment. Int. J. Cancer 86:570-76 24, Taylor JK, Zhang QQ, Monia BP, Marcusson

EG,

Dean

NM.

1999.

Inhibition

of Bcl-xL expression sensitizes normal human keratinocytes and epithelial cells to apoptotic stimuli. Oncogene 18:4495— 504 pay, Lebedeva I, Stein C. 1999. Bel-xL-directed oligonucleotides in prostate and bladder carcinoma cell lines. Antisense Nucleic Acid Drug Dev. 9:381-82 26. Stein CA. 1999. Two problems in antisense biotechnology: in vitro delivery and the design of antisense experiments. Biochim. Biophys. Acta 1489:45—52 Pde Guvakova MA, Yakubov LA, Vlodavsky I, Tonkinson JL, Stein CA. 1995. Phosphorothioate oligodeoxynucleotides bind to basic fibroblast growth factor, inhibit its

binding to cell surface receptors, and remove it from low affinity binding sites on extracellular matrix. J. Biol. Chem. 270:2620-27 28. Fennewald SM, Rando RF. 1995. Inhibition of high affinity basic fibroblast growth factor binding by oligonucleotides. J. Biol. Chem. 270:21718-21 DS), Rockwell P, O’Connor WJ, King K, Goldstein NI, Zhang LM, et al. 1997.

416

LEBEDEVA

® STEIN

Cell-surface perturbations of the epidermal growth factor and vascular endothelial growth factor receptors by phosphorothioate oligodeoxynucleotides. Proc. Natl. Acad. Sci. USA 94:6523-28 30. Khaled

Z, Benimetskaya

L, Zeltser

R,

Khan T, Sharma HW, et al. 1996. Multiple

mechanisms may contribute to the cellular anti-adhesive effects of phosphorothioate oligodeoxynucleotides. Nucleic Acids Res. 24:737-45 Sil Benimetskaya L, Loike JD, Khaled Z, Loike G, Silverstein SC, et al. 1997. Mac-1

(CD11b/CD18)

is an oligodeoxy-

nucleotide-binding protein. Nat. Med. 3:414-20 Bye, Yakubov L, Khaled Z, Zhang LM, Truneh A, Vlassov V, et al. 1993. Oligodeoxynucleotides interact with recombinant CD4 at multiple sites. J. Biol. Chem. 268:18818— 23 33% Stein CA, Cleary AM, Yakubov L, Lederman S. 1993. Phosphorothioate oligodeoxynucleotides bind to the third variable loop domain

(v3) of human

38.

oo;

40.

41.

42.

.

im-

munodeficiency virus type 1 gp120. Antisense Res. Dev. 3:19-31

34, Majumdar C, Stein CA, Cohen JS, Broder S, Wilson SH. 1989. Stepwise mechanism of HIV reverse transcriptase: primer function of phosphorothioate oligodeoxynucleotide. Biochemistry 28:1340-46

3h). Stein CA, Tonkinson JL, Zhang Yakubov L, Gervasoni J, et al. Dynamics of the internalization of phodiester oligodeoxynucleotides in cells. Biochemistry 32:4855—61 36. Gao WY, Han FS, Storm C, W,

Cheng

YC.

1992.

LM, 1993. phosHL60 Egan

Phosphorothioate

oligonucleotides are inhibitors of human DNA polymerases and RNase H: implications for antisense technology. Mol. Pharmacol. 41:223-29 Sik Perez JR, Li Y, Stein CA, Majumder S, van Oorschot A, et al. 1994. Sequenceindependent induction of Spl transcription factor activity by phosphorothioate

.

oligodeoxynucleotides. Proc. Natl. Acad. Sci. USA 91:5957-61 Bock LC, Griffin LC, Latham JA, Vermaas EH, Toole JJ. 1992. Selection of singlestranded DNA molecules that bind and inhibit human thrombin. Nature 355:564—66 Rando RF, Hogan ME. 1998. Biological activity of guanosine quartet-forming oligonucleotides. See Ref. 89, pp. 33552 Wang Y, Patel DJ. 1993. Solution structure of a parallel-stranded G-quadruplex DNA. J. Mol. Biol. 234:1171-83 Wyatt JR, Stein CA. 1999. G-quartet inhibitory effects of phosphorothioate oligonucleotides. In Applications of Antisense Therapies to Restenosis, ed. L Rabbani, pp. 99-118. Norwell, MA: Kluwer Acad. Gewirtz AM, Calabretta B. 1988. A c-myb antisense oligodeoxynucleotide inhibits normal human hematopoiesis in vitro. Science 242:1303-6 Basu S, Wickstrom E. 1997. Temperature and salt dependence of higher order structure formation by antisense c-myc and c-myb phosphorothioate oligodeoxyribonucleotides containing tetraguanylate tracts. Nucleic Acids Res. 25:1327-32 Mir KU, Southern EM. 1999. Determining the influence of structure on hybridization using oligonucleotide arrays. Nat. Biotechnol. 17:788—92

. Patzel V, Steidl U, Kronenwett R, Haas R,

Sczakiel G. 1999. A theoretical approach to select effective antisense oligodeoxyribonucleotides at high statistical probabil- ity. Nucleic Acids Res. 27:4328-34 46. Bennett CF. 1998. Use of cationic lipid complexes for antisense oligonucleotide delivery. See Ref. 89, pp. 129-46 47. Tu GC, Cao QN, Zhou F, Israel Y. 1998. Tetranucleotide GGGA motif in primary RNA transcripts: novel target site for antisense design. J. Biol. Chem. 273:25125-31 48. Smetsers TF, Boezeman JB, Mensink EJ. 1996. Bias in nucleotide composition of

ANTISENSE OLIGONUCLEOTIDES

antisense oligonucleotides. Antisense Nucleic Acid Drug Dev. 6:63-67 49, Walder RY, Walder JA. 1988. Role of RNase H in hybrid-arrested translation by antisense oligonucleotides. Proc. Natl. Acad. Sci. USA 85:5011-15

50. Minshull J, Hunt T. 1986. The use ofsinglestranded DNA and RNase H to promote quantitative ‘hybrid arrest of translation’ of mRNA/DNA hybrids in reticulocyte lysate cell-free translations. Nucleic Acids Res. 14:6433-51

Bile Cazenave C, Loreau N, Thuong NT, Toulme JJ, Helene C. 1987. Enzymatic amplification of translation inhibition of rabbit beta-globin mRNA mediated by anti-messenger oligodeoxynucleotides covalently linked to intercalating agents. Nucleic Acids Res. 15:4717-—36 52) Shuttleworth J, Colman A. 1988. Antisense oligonucleotide-directed cleavage of mRNA in Xenopus oocytes and eggs. EMBO J. 7:427-34 a5: Giles RV, Ruddell CJ, Spiller DG, Green JA, Tidd DM. 1995. Single base discrimination for ribonuclease H-dependent antisense effects within intact human leukaemia cells. Nucleic Acids Res. 23:954-61 54, Furdon PJ, Dominski Z, Kole R. 1989. RNase H cleavage of RNA hybridized to oligonucleotides containing methylphosphonate, phosphorothioate and phosphodiester bonds. Nucleic Acids Res. 17:9193204 35 Monia BP, Lesnik EA, Gonzalez C, Lima WF,

McGee

D, et al. 1993.

Evaluation

of 2’-modified oligonucleotides containing 2’-deoxy gaps as antisense inhibitors of gene expression. J. Biol. Chem. 268: 14514—22

56. Giles

RV,

Tidd

DM.

1992.

Increased

specificity for antisense oligodeoxynucleotide targeting of RNA cleavage by RNase H using chimeric methylphosphonodiester/phosphodiester structures. Nucleic Acids Res. 20:763-70

417

>I] . Giles RV, Tidd DM. 1992. Enhanced RNase H activity with methylphosphonodiester/phosphodiester chimeric antisense oligodeoxynucleotides. Anticancer Drug Des. 7:37-48 58. Dean N, McKay R, Miraglia L, Howard R, Cooper S, et al. 1996. Inhibition of growth of human tumor cell lines in nude mice by an antisense of oligonucleotide inhibitor of protein kinase C-alpha expression. Cancer Res. 56:3499-507 Sy), Benimetskaya L, Takle GB, Vilenchik M, Lebedeva I, Miller P, et al. 1998. Cationic

porphyrins: novel delivery vehicles for antisense oligodeoxynucleotides. Nucleic Acids Res. 26:5310-17

60. Liebhaber SA, Cash FE, Shakin SH. 1984. Translationally associated helixdestabilizing activity in rabbit reticulocyte lysate. J. Biol. Chem. 259:15597-602 61. Shakin SH, Liebhaber SA. 1986. Destabilization of messenger RNA/complementary DNA duplexes by the elongating 80 S ribosome. J. Biol. Chem. 261:16018DS 62. Altman S. 1993. RNA enzyme-directed gene therapy. Proc. Natl. Acad. Sci. USA 90: 10898900 63. Forster AC, Altman S. 1990. External guide sequences for an RNA enzyme. Science 249:783—86 64. Ma M, Benimetskaya L, Lebedeva I, Dignam

J, Takle

G, et al. 2000.

Intracellu-

lar mRNA cleavage induced through activation of RNase P by nuclease-resistant external guide sequences. Nat. Biotechnol. 18:58-61

65. Reed JC, Stein C, Subasinghe C, Haldar S, Croce CM, et al. 1990. Antisense-mediated

inhibition of BCL2 protooncogene expression and leukemic cell growth and survival: comparisons of phosphodiester and phosphorothioate oligodeoxynucleotides. Cancer Res. 50:6565—70

66. Campos L, Sabido O, Rouault JP, Guyotat D. 1994. Effects of BCL-2 antisense oligodeoxynucleotides on in vitro

418

LEBEDEVA

® STEIN

proliferation and survival of normal marrow progenitors and leukemic cells. Blood 84:595—600

67. Durrieu F, Belaud-Rotureau MA, Lacombe F, Dumain

P, Reiffers

J, et al.

1999.

Synthesis of Bcl-2 in response to anthracycline treatment may contribute to an apoptosis-resistant phenotype in leukemic cell lines. Cytometry 36:140-49

68. Keith FJ, Bradbury DA, Zhu YM, Russell NH. 1995. Inhibition of bel-2 with antisense oligonucleotides induces apoptosis and increases the sensitivity of AML blasts to Ara-C. Leukemia 9:131—38 69. Kitada S, Miyashita T, Tanaka S, Reed JC. 1993. Investigations of antisense oligonucleotides targeted against bcl-2 RNAs. Antisense Res. Dev. 3:157-69 70. Kitada S, Takayama S, De Riel K, Tanaka S, Reed JC. 1994. Reversal of chemoresistance of lymphoma cells by antisense-mediated reduction of bel-2 gene expression. Antisense Res. Dev. 4:71— 79 Cis McDonnell TJ, Navone NM, Troncoso P, Pisters LL, Conti C, et al. 1997. Expression of bel-2 oncoprotein and p53 protein accumulation in bone marrow metastases of androgen independent prostate cancer. J. Urol. 157:569-74 Ws Bauer JJ, Sesterhenn IA, Mostofi FK, McLeod DG, Srivastava S, et al. 1996. Elevated levels of apoptosis regulator proteins p53 and bel-2 are independent prognostic biomarkers in surgically treated clinically localized prostate cancer. J. Urol. 156:1511-16 Tey. Raffo AJ, Perlman H, Chen MW, Day ML, Streitman JS, et al. 1995. Overexpression of bel-2 protects prostate cancer cells from apoptosis in vitro and confers resistance to androgen depletion in vivo. Cancer Res. 55:4438-45 74. Gleave M, Tolcher A, Miyake H, Nelson C, Brown B, et al. 1999. Progression to androgen independence is delayed by adjuvant treatment with antisense Bcl-2

oligodeoxynucleotides after castration in the LNCaP prostate tumor model. Clin. Cancer Res. 5:2891-98 123 Miyake H, Tolcher A, Gleave ME. 1999. Antisense Bcl-2 oligodeoxynucleotides inhibit progression to androgen-independence after castration in the Shionogi tumor model. Cancer Res. 59:4030-34 76. Gleave ME, Miyake H, Goldie J, Nelson C, Tolcher A. 1999. Targeting bcl2 gene to delay androgen-independent progression and enhance chemosensitivity in prostate cancer using antisense bcl2 oligodeoxynucleotides. Urology 54:36— 46 W744. Agrawal S, Temsamani J, Tang JY. 1991. Pharmacokinetics, biodistribution, and stability of oligodeoxynucleotide phosphorothioates in mice. Proc. Natl. Acad. Sci. USA 88:7595—99 78. Iversen P. 1991. In vivo studies with phosphorothioate oligonucleotides: pharmacokinetics prologue. Anticancer Drug Des. 6:531-38 To Raynaud FI, Orr RM, Goddard PM, Lacey HA, Lancashire H, et al. 1997. Pharmacokinetics of G3139, a phosphorothioate oligodeoxynucleotide antisense to bel-2, after intravenous administration or continuous subcutaneous infusion to mice. J. Pharmacol. Exp. Ther. 281:420-27 . Cotter FE, Johnson P, Hall P. Pocock C,

al Mahdi N, et al. 1994. Antisense oligonucleotides suppress B-cell lymphoma growth in a SCID-hu mouse model. Oncogene 9:3049-55 81. Cotter FE, Corbo M, Raynaud F, Orr RM, Pocock C, et al. 1996. Bcl-2-antisense therapy in lymphoma: in vitro and in vivo mechanisms, efficacy, pharmacokinetics and toxicity studies. Ann. Oncol. 7:32 82. Wooldridge JE, Ballas Z, Krieg AM, Weiner GJ. 1997. Immunostimulatory oligodeoxynucleotides containing CpG motifs enhance the efficacy of monoclonal antibody therapy of lymphoma. Blood 89:2994_98

ANTISENSE OLIGONUCLEOTIDES

83. Cotter FE, Waters J, Cunningham D. 1999. Human Bcl-2 antisense therapy for lymphomas. Biochim. Biophys. Acta 1489:97— 106 84. Webb A, Cunningham D, Cotter F, Clarke PA, di Stefano F, et al. 1997. BCL-2 antisense therapy in patients with nonHodgkin lymphoma. Lancet 349:11374i 85. Waters JS, Webb A, Cunningham D, Clarke PA, Raynaud F, et al. 2000. Phase I clinical and pharmacokinetic study of bcl-2 antisense oligonucleotide therapy in patients with non-Hodgkin’s lymphoma. J. Clin. Oncol. 18:1812—23 86. Jansen B, Wacheck V, Heere-Ress E, Schlagbauer-Wadl H, Hollenstein U, et al.

1999. A phase I-II study with dacarbazine and BCL-2 antisense oligonucleotide G3139

(GENTA) as a chemosensitizer in patients with advanced malignant

419

melanoma. Proc. Am. Soc. Clin. Oncol. 18:2047 (Abstr.)

87. Chen HX, Marshall JL, Trocky N, Ling Y, Baidas S, et al. 2000. A phase I study of BCL-2 antisense G3139 (GENTA) and

weekly docetaxel in patients with advanced breast cancer and other solid tumors. Proc. Am. Soc. Clin. Oncol. 19:692 (Abstr. )

88. Morris MJ, Tong W, Osman

I, Maslak P, Kelly WK, et al. 1999. A phase I/IIA

dose-escalating trial of bcl-2 antisense (G3139) treatment by 14-day continuous intravenous infusion (CI) for patients with androgen-independent prostate cancer or other advanced solid tumor malignancies. Proc. Am. Soc. Clin. Oncol. 18:1241 (Abstr.)

89. Stein CA, Krieg AM, eds. 1998. Applied Antisense Oligonucleotide Technology. New York: Wiley-Liss

= |

—

-

2

1

; wi

5

=

-

-

7

’ > >

:

f

‘y

a

co

eA

“i

|

:

| 9)

-

ei oy

a _

_

a

y

>

Te

pe

hed A 9)

as

D

iy

:

;

~—

St wae

sth. »

&

we

. ei

pyow ~~

;

yaa

TR reer

70 years), respectively. Similarly, the clearance of omeprazole (a substrate of CYP2C19) was decreased in the elderly (250 ml/min) compared with that in the young (594 ml/min) individuals (117). Similar to the controversy over the agerelated decline in enzyme activity, there are conflicting reports regarding enzyme induction in the elderly. Using rifampicin as an inducing agent, Twum-Barima et al (118) found that the inducing effect of rifampicin on antipyrine metabolism was greater in young than in old subjects. Similarly, impaired enzyme induction in the elderly was reported by Salem et al (119). However, other investigators found similar degrees of induction in old and young subjects. Treatment with rifampicin resulted in the same degree of induction, and a 30-fold decrease in oral AUCs of verapamil in both old and young volunteers (82). The induction of theophylline metabolism by phenytoin has also been found to be unaffected by age (120). The. reason for these discrepancies is unknown. However, given the conflicting data from different pharmacokinetic studies, it is not possible to generalize about an effect of aging on the induction response. Genetic polymorphism can also influence enzyme induction. Depending on whether the individual has or does not have the functioning enzyme, there are interindividual differences in response to inducing agents. For example, induction

of polymorphic 4-hydroxylation of S-mephenytoin by rifampicin occurred only

with EMs of CYP2C19, but not with PMs (121). Mephenytoin is a 1:1 racemic

mixture of R- and S-enantiomers.

The S-mephenytoin is completely and rapidly

VARIABILITY IN DRUG-DRUG INTERACTIONS

557

metabolized by CYP2C19 to form 4-OH-mephenytoin, whereas R-mephenytoin is metabolized very slowly to form 5-phenyl-5-ethylhydantoin. The stereoselective differences in the excretion of unchanged enantiomers in urine (urinary R/S ratio) have been used as an index of metabolic capacity. Daily dosing with rifampicin for 20 days resulted in a three- to eightfold increase in the 0- to 8-h urinary R/S ratio of mephenytoin after oral administration of racemic drugs to EMs of CYP2C19, although the urinary ratio was unaffected in PMs after rifampicin treatment. In another clinical study, rifampicin treatment caused a twofold increase in systemic clearance of propafenone in PMs of CYP2D6, but it had little effect on the clearance in EMs (122). CYP2D6-mediated polymorphic hydroxylation is the major pathway of propafenone metabolism in EMs, whereas CYP3A4-CYP1A2mediated N-dealkylation is the main metabolic pathway in PMs. Thus, induction of propafenone N-dealkylation (CYP3A4-CYP1A2) by rifampicin resulted in more pronounced increases in the clearance of propafenone in PMs, but not in EMs. Omeprazole is known to induce CYP1A enzymes. A clinical study was conducted to evaluate the effect of genetic factors on enzyme induction by omeprazole. This study included 18 volunteers, 12 EMs, 5 PMs, and one intermediate metabolizer

(IM) of S-mephenytoin. The CYP1A enzyme activity was measured by a !°C-[N3-methyl]-caffeine breath test. Omeprazole treatment caused a significant increase in CYPIA activity in all PMs and the IM, whereas omeprazole had little inductive effect in EMs (123). Data analysis of all 18 subjects revealed that there was a good correlation between the percent increase in enzyme activity (measured by cumulative ‘CO, exhalation) and the plasma AUC of omeprazole. It is believed that the lack of omeprazole-inductive effect in EMs was due to less systemic exposure of omeprazole, as a result of rapid metabolism of omeprazole in EMs. CYP3A5, a member of CYP3A

subfamily, demonstrates 84% amino acid se-

quence similarity with CYP3A4. Because of their similarity, there is a large overlap in substrate specificity for both CYP isoforms. However, unlike CYP3A4, CYP3A5 is polymorphically expressed in human liver, appearing in only 25% of the human population, and CYP3AS is reportedly not inducible by rifampicin (124-126). It is conceivable that the response of CYP3A substrates to rifampicin induction will be highly dependent on the molar ratio of CYP3A4 to CYP3A5. A greater increase in the metabolism of CYP3A substrates by rifampicin induction is expected for individuals with CYP3A4 enzyme only, compared with those with CYP3A4 and CYP3AS. In addition to the environmental and genetic factors discussed above, kinetic properties of substrates can also influence the interpretation of enzyme induction when the changes in plasma AUCs of substrates are used. As shown in Figure 3 (parallel-tube model), the magnitude of changes in oral AUCs by enzyme induction is highly dependent on the kinetic properties of drugs; high-clearance drugs are more sensitive to the changes in the CL,,,. Although an increase in the CL;,,, results in an almost proportional decrease in the AUCs for low-clearance drugs, for highand intermediate-clearance drugs, the changes in the AUCs are more than proportional to the changes in the CL;,,. Therefore, it is expected that enzyme induction

558

LIN #® LU

has less effect on low-clearance drugs than on high-clearance drugs. O'Reilly (127) reported that treatment with rifampicin caused a significant reduction of warfarin plasma concentration in patients. As expected, the extent of decreases in the AUCs of warfarin [a low-clearance drug (clearance = 0.05 ml/min/kg)] in response to rifampicin induction was twofold after oral administration of warfarin. Similarly, a twofold decrease in warfarin AUCs caused by induction was observed during phenobarbital treatment (128). In contrast to warfarin, verapamil [a highclearance drug (clearance= 12 ml/min/kg)] shows profound inductive response to rifampicin treatment. Rifampicin increased the oral AUCs of S-verapamil by 32-fold (31). In another study, after 12 days of rifampicin treatment a marked reduction (12-fold) in the oral AUCs of verapamil was observed (129). Similar observations were obtained after barbiturate treatment. The extent of decrease in the oral AUC for alprenolol (clearance = 15 ml/min/kg) was more dramatic than the extent for warfarin after barbiturate induction (130-132). For drugs with intermediate clearance, the extent of decrease in oral AUC caused by enzyme induction would be greater than that for low-clearance drugs, but less than that for high-clearance drugs. Quinidine is an intermediate-clearance drug with a systemic clearance of 5 ml/min/kg. The oral AUC of quinidine was reduced by four- to sixfold with concurrent oral administration of rifampic:n (133). From these examples, it is clear that the extent of decrease in oral AUCs of drugs highly depends on their clearance. Because of interindividual variability in the basal level of enzymes, a drug could be defined as a high-clearance compound in some individuals and a low-clearance compound in others. Thus, the decrease in the oral AUC of a given drug is expected to be greater for individuals with higher basal clearance than for those with lower basal clearance. The magnitude of decrease in plasma AUCs of drugs by enzyme induction is also dependent on the route of drug administration. Although for low-clearance drugs, an increase in the CL;,, owing to enzyme induction causes a proportional decrease in AUC regardless of route of drug administration, for high- and intermediateclearance drugs, the change in AUC is much greater after oral administration than after intravenous administration (Figures 3and 4). Enzyme induction has little effect on the AUCs of high-clearance drugs after intravenous administration because the hepatic clearance of high-clearance drugs is limited by hepatic blood flow (Equations 10 and 11). Therefore, an increase in the intrinsic clearance caused by enzyme induction has little effect on the hepatic clearance of high-clearance drugs, unless inducers also alter hepatic blood flow. As shown in the examples cited above, the extent of reduction in the AUC of warfarin (a low-clearance drug) by rifampicin was almost identical, approximately twofold, after either oral or intravenous administration of warfarin (127). On the other hand, treatment of rifampicin caused a 15-fold decrease

in the AUC

of vera-

pamil (a high-clearance drug) after a single oral dose, whereas rifampicin treatment had little effect on the intravenous AUC of verapamil (129). Similar results were reported by Fromm et al (31), who found that rifampicin decreased the

VARIABILITY IN DRUG-DRUG INTERACTIONS

559

AUC of S-verapamil by 32-fold after oral administration of racemic verapamil, but it decreased the AUC of S-verapamil by only 1.3-fold after intravenous administration.

CONCLUSIONS Undoubtedly, drug interactions involving enzyme inhibition and induction will continue to be a major medical concern for clinicians and patients, and pharmaceutical companies will continue to put their efforts into predicting the potential of drug interactions of new drug candidates during drug development. When predicting drug-drug interactions, the following questions are often addressed. Can in vivo drug-drug interactions be predicted accurately from in vitro studies? Should such predictions be qualitative or quantitative? Some scientists believe that a quantitative prediction of drug interaction is possible (134-136), whereas others are less optimistic and believe that quantitative predictions would be extremely difficult, if not impossible (8). There are many factors that contribute to our inability to quantitatively predict drug interactions. One of the major contributing factors is the large interindividual variability in response in enzyme inhibition and induction. In addition, the difficulty in predicting drug interactions caused by enzyme inhibition stems from the inaccuracy of K; estimation from in vitro studies and insufficient methods for direct measurement of inhibitor concentration at the sites of metabolism (8). It is even more difficult to quantitatively predict in vivo interactions involving enzyme induction from in vitro induction studies, because many underlying mechanisms for enzyme induction remain largely unknown. Until optimal experimental conditions are established for accurate estimation of K; values and inhibitor concentrations and the underlying mechanisms of enzyme induction and the sources of interindividual variability in enzyme inhibition and induction are completely understood, information obtained from in vitro metabolic studies will go only so far in predicting whether there will be a lack of an interaction or a probability of one. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Lin JH, Lu AYH. 1998. Inhibition and induction of cytochrome P450 and the clinical implications. Clin. Pharmacokinet. 35:361oe

amitriptyline Forensic Sct. 3. Ferslew KE, McCormick

and other psychotropic drugs. Int. 86:35—41 Hagardorn AN, Harlan GC, WF. 1998. A fatal drug inter-

T, Jitsufuchi N, Zhang

action between clozapine and fluoxetine. J.

XX, Tokunaga H, Nagata T. 1997. Death attributed to the toxic interaction of triazolam,

Forensic Sci. 43:1082—85 4. Lazarou J, Pomeraanz BH, Corey PN. 1998.

2. Kudo

K, Imamura

560

LIN # LU

Incidence of adverse drug interaction in hospitalized patients: a meta-analysis of prospective studies. JAMA 279:1200-5 . Duchateau A. 1998. Posicor: veni, vidi and

gone. Pharm. Weekbl. 133:1294—95

nois JP, Guenerich FP, Guillouzo A. 1991.

. Park BK, Kitteringham NR, Pirmohamed

M, Tuker GT. 1996. Relevance of induc-

tion of human drug-metabolizing enzymes: pharmacological and toxicological implications. Br. J. Clin. Pharmacol. 41:477-91 . Ronis MJJ, Ingelmam-Sundberg M. 1999.

Induction of human drug-metabolizing enzymes: mechanisms and implications. In Handbook

of Drug

Metabolism,

Woolf, pp. 239-62. New Dekker

York:

ed. TF

Marcel

. Lin JH. 2000. Sense and nonsense in the

prediction of drug-drug interactions. Current Drug Metab. In press . Bailey DG, Spence JD, Munoz C, Amold

JMO. 1991. Interaction of citrus juices with felodopine and nifedipine. Lancet 337:268-69 . Edgar B, Bailey DG, Bergstrand R, Johns-

Mle

We

.

.

son G, Regardh CG. 1992. Acute effect of drinking grapefruit juice on the pharmacokinetics and dynamics of felodopine and its potential clinical relevance. Eur. J. Clin. Pharmacol. 42:313-17 Bailey DG, Bend JR, Amold JMO, Tran LT, Spence JD. 1996. Erythromycin felodipine interaction: magnitude, mechanism, and comparison with grapefruit juice. Clin. Pharmacol. Ther. 60:25-33. Nelson DR, Koymans L, Kamataki T, Stegeman J, FeyereisenR, et al. 1996. P450 superfamily: update on new sequences, gene mapping, accession numbers and nomenclature. Pharmacogenetics 6:1—42 Nelson DR. 1999. Cytochrome P450 and the individuality ofspecies. Arch. Biochem. Biophys. 369:1-10 Gonzalez FJ, Kasper CB. 1980. Phenobarbital induction of NADPH cytochrome c(P450) oxidoreductase messenger ribonucleic acid. Biochemistry 19:1790-96

. Sabatini

DD,

Kreibach

Adesnik M. 1982. Mechanism for incorporation of proteins in membranes and organelles. J. Cell Biol. 92:1—22 16. Ratanasavanh D, Beaune P, Morel F, Fli-

G, Morimoto

T.’

Intralobular distribution and quantitation of cytochrome P450 enzymes in human liver as a function of age. Hepatology 13:1142-51 £7. Watkins PB, Murrary SA, Thomas PE, Wrighton SA. 1990. Distribution of cytochromes P450, cytochrome bS5, and NADPH-cytochrome P450 reductase in an entire human liver. Biochem. Pharmacol. 39:471-76 18. Debri K, Boobis AR, Davis DS, Edwards RJ. 1995. Distribution and induction of CYP3A1 and CYP3A2 in rat liver and extrahepatic tissues. Biochem. Pharmacol. 50:2047-56 . Shimada T, Yamazaki H, Mimura M, Inui

Y, Guengerich FP. 1994. Interindividual variations in human liver cytochrome P450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J. Pharmacol. Exp. Ther. 270:414—23 . Knshna DR, Klotz U. 1994. Extrahepatic metabolism of drugs in humans. Clin. Pharmacokinet. 26:144—60. . de Waziers I, Cugence PH, Yang CS, Lerous JP Beaune PH. 1990. Cytochrome P450 isozymes, epoxide hydrolase and glutathione transferase in rat and human hepatic and extrahepatic tissues. J. Pharmacol. Exp. Ther. 253:387-94 . Back DJ, Rogers SM. 1987. First-pass. metabolism by gastrointestinal mucosa. Aliment Pharmacol. Ther. 227:339-57 . Pacifici GM, Franchi M, Bencini F. 1988.

Tissue distribution of drug-metabolizing enzymes in human. Xenobiotics 18:849-56 24. Paine MF, Khalighi M, Fisher JM, Shen DD, Kunze KL, et al. 1997. Characterization of interintestinal and intraintestinal variations in human CYP3A-dependent

VARIABILITY IN DRUG-DRUG INTERACTIONS

metabolism.

J. Pharmacol.

Exp.

Ther.

283:1552-62 DUSy. Murray GI, Barnes TS, Swell HE, Ewen SW, Melvin WT, Burk MD. 1988. The immunochemical localization and distribution of cytochrome P450 in normal hepatic and extrahepatic tissues with a monoclonal antibody to human cytochrome P450. Br. J. Clin. Pharmacol. 25:465—75 26. Thummel KE, O’Shea D, Paine MF, Shen DD,

Kunze

KL,

Perkins

JD, Wilkinson

GR. 1996. Oral first-pass elimination of midazolam involves both gastrointestinal and hepatic CYP3A-mediated metabolism. Clin. Pharmacol. Ther. 59:491—502

27, Paine MF, Shen DD, Kunze KL, Perkins JD,

Marsh

CL,

et al.

1996.

First-pass

metabolism of midazolam by the human intestine. Clin. Pharmacol. Ther. 60:14-24

28. Prueksaritanont T, Dwyer LM, Cribb AE. 1995. (+) Bufuraloll’-hydroxylation activity in human nad rhesus monkey intestine and liver. Biochem. Pharmacol. 50:1521-25 wg). Wu CY, Benet LZ, Hebert MF, Gupta SK, Rowland

M, et al. 1995.

Differen-

tiation absorption and first-pass gut and hepatic metabolism in humans: _ studies with cyclosporine. Clin. Pharmacol. Ther. 58:492-97

30. Holtbecker N, Fromm MF, Kroemer HK, Ohnhms

EF, Heidemann

H.

1996.

The

nifedipine-rifampin interaction: evidence for induction of gut wall metabolism. Drug. Metab. Dispos. 24:112 1-23 31; Fromm MF, Busse D, Kroemer HK, Eichelbaum M. 1996. Differential induction of prehepatic and hepatic metabolism of verapamil by rifampin. Hepatology 24:796-801 Bs Lin JH, Chiba M, Baillie TA. 1999. Is the role of the small intestine in first-pass metabolism overemphasized? Pharmacol. Rev, 51:135-57 33, Higashikawa F, Murakami T, Kaneda T, Kato A, Takano M. 1999. Dose-dependent

intestinal and hepatic first-pass metabolism

561

of midazolam, a cytochromeP450 3A substrate with differently modulated enzyme activity in rats. J. Pharm. Pharmacol. 51:67-72

34. Lin JH, Chiba M, Chen J-W, Nishime JA, deLuna FA, et al. 1999. Effect of dexame-

thasone on the intestinal first-pass metabolism of indinavir in rats: evidence of cytochrome P4503A and _p-glycoprotein induction. Drug Metab. Dispos. 27:118793

35: Thummel KE, Shen DD, Podoll TD, Kunze KL, Trager WF, et al. 1994. Use of midazo-

lam as ahuman cytochrome P4503A probe: IJ. Characterization of inter- and intraindividual hepatic CYP3A variability after liver transplantation. J. Pharmacol. Exp. Ther. 271:557-66

36. Lown KS, Kolars JC, Thummel KE, Barnett JL, Kunze KL, et al. 1994. Interpatient

heterogeneity in expression of CYP3A4 and CYP3AS5 in small bowel: lack of prediction by the erythromycin breath test. Drug Metab. Dispos. 22:947-S5 . Lampen

A, Chritians

U, Guenerich

FP,

Watkins PB, Kolars JC, et al. 1995. Metabolism of the immunosuppressant tacrolimus in the small intestine: cytochrome P450, drug interactions, and interindividual variability. Drug Metab. Dispos. 23: 1315-24

38. Thummel KE, Shen DD, Podoll TD, Kunze KL, Trager WF, et al. 1994. Use of midazo-

lam as ahuman cytochrome P4503A probe: I. In vitro- in vivo correlations in liver transplant patients. J. Pharmacol. Exp. Ther. 271:549-56 39), Ozdemir V, Kalow W, Tang BK, Paterson AD, Walker SE, et al. 2000. Evaluation of the genetic component of variability in CYP3A4 activity: a repeated drug administration method. Pharmacogenetics 10:373-88 40. Lin JH, Chen I-W, Chiba M, Nishime JA, deLuna FA. 2000. Route-dependent nonlinear pharmacokinetics of a novel HIV protease inhibitor: Involvement of enzyme

562

LIN # LU

inactivation. Drug Metab. Dispos. 28:460— 66

41. Lin JH, Chiba M, Chen I-W, Vastag KJ, Nishime JA, Dorsey BD. 1995. Time- and

dose-dependent

pharmacokinetics

of L-

754,394, an HIV protease inhibitor, in rats,

dogs and monkeys. J. Pharmacol. Exp. Ther. 274:264-69 42. Mico BA, Federowicz DA, Ripple MG, Kerns W. 1988. In vivo inhibition of oxidative drug metabolism by, and acute toxicity of 1-aminobenzotriazole (ABT): a tool for

biochemical toxicology. Biochem. Pharmacol. 37:2515-19 43. Segel TH. 1975. Enzyme Kinetics: Behavior and Analysis of Rapid Equilibrium and Steady-State Enzyme Systems. New York: Wiley & Sons 44. Okey AB. 1992. Enzyme induction in the cytochrome P450 system. In Pharmacogenetics of Drug Metabolism, ed. W Kalow , pp. 549-608, New York: Pergamon 45. Honkakoski P, Negishi M. 2000. Regulation of cytochrome P450 (CYP) genes by nuclear receptors. Biochem. J. 347:321-37 46. Lehmann JM, McKee DD, Watson MA, Wilson TM, Moore JT, Kliewer SA. 1998.

The human orphan nuclear receptor PXR is activated by compounds that regulate CYP3A4 gene expression and cause drug interactions. J. Clin. Invest. 102:1016—23 47. Sueyoshi T, Kawamoto T, Zelko I, Honkakoski P, Negishi M. 1999. The repressed nuclear receptor CAR responds to phenobarbital in activating the human CYP2B6 gene. J. Biol. Chem. 274:6043-46 48, Rettie AE, Wienkers LC, Gonzalez FJ, Trager WF, Korzekwa KR. 1994. Impaired (s)-warfarin metabolism catalyzed by the R144C allelic variant of CYP2C9. Pharmacogenetics 4:39-46 49. Tracy TS, Rosenbluth BW, Wrighton SA, Gonzalez FJ, Korzekwa KR. 1995. Role of cytochrome P450 2C9 and an allelic

variant in the 4/-hydroxylation of (R)and (S)-flurbiprofen. Biochem. Pharmacol. 49:1229-35

50. Wilkinson GR. 1987. Clearance approaches in pharmacology. Pharmacol. Rev. 39:1-47 SL. Honig PK, Wortham DC, Zamani K, Conner DP, Mullin JC, Cantilena LR. 1993. Terfenadine-ketoconazole interaction— pharmacokinetic and electrocardiographic consequences. JAMA 269:1513-18 D2: Funck-Brentano C, Turgeon J, Woosley RL, Roden DM. 1989. Effect of low dose quinidine on encainide pharmacokinetics and pharmacodynamics: influence of genetic polymorphism. J. Pharmacol. Exp. Ther. 249:13442 53: Lessard E, Yessine MA, Hamelin BA, O’ Hara G, LeBlanc J, Turgeon J. 1999. Influence of CYP2D6 activity on the disposition and cardiovascular toxicity of the antidepressant agent venlafaxine in humans. Pharmacogenetics 9:435—43 54. Gomez DY, Wacher VJ, Tomlanovich SJ, Hebert MF, Benet LZ. 1995. The effect of ketoconazole on the intestinal metabolism and bioavailability of cyclosporine. Clin. Pharmacol. Ther. 58:15—19 nNNn

. Varhe A, Olkkola KT, Neuvonen PJ. 1994.

Oral triazolam is potentially hazardous to patients receiving systemic antimycotics ketoconazole or itraconazole. Clin. Pharmacol. Ther. 56:601—7 . Greenblatt DJ, Wright CE, von Moltke LL, Harmatz JS, Ehrenberg BL, et al. 1998. Ketoconazole inhibition of triazolam and alprazolam clearance: differential kinetic and dynamic consequences. Clin. Pharmcol. Ther. 64:237-47 . Bailey DG, Kreeft JH, Munoz C, Freeman DJ, Bend JR. 1998. Grapefruit juicefelodipine interaction: effect of naringin

and 6',7’- dihydroxybergamottin

in hu-

mans. Clin. Pharmacol. Ther. 64:248-56 . Yee GC, Stanley DL, Pessa LJ, Dalla Costa T, Beltz SE, et al. 1995. Effect of grapefruit

juice on blood cyclosporin concentration. Lancet 345:955-—56

ah Kupferschmidt HHT, Ha HR, Ziegler WH, Meier PJ, Krahenbuhl S. 1995. Interaction

VARIABILITY IN DRUG-DRUG INTERACTIONS

563

between grapefruit juice and midazolam in humans. Clin. Pharmacol. Ther. 58:20-28

tion in Chinese. Pharmacogenetics 5:3742

60. Lown SK, Bailey DG, Fontana RJ, Janar-

68. Nasu K, Kubota T, Ishizaki T. 1997. Genetic analysis of CYP2C9 polymorphism in a Japanese population. Pharmacogenetics 7:405—9 69. Bhasker CR, Miners JO, Coulter S, Birkett DJ. 1997. Allelic and functional variability of cytochrome P450 2C9. Pharmacogenetics 7:51-58 70. Sullivan-Klose TH, Ghanayem BI, Bell DA, Zhang ZY, Kaminsky LS, et al. 1996.

dan SK, Adair CH, et al. 1997. Grapefruit juice increases felodipine oral availability in humans by decreasing intestinal CYP3A protein expression. J. Clin. Invest. 99:2545-53 61. Eichelbaum M, Dengler HJ, Somogyi A, Van Unruh GE. 1981. Superiority of stable isotope techniques in the assessment of bioavailability of drugs undergoing extensive first-pass elimination: studies of the relative bioavailability of verapamil tablets. Eur. J. Clin. Pharmacol. 19:127Bit 62. Eichelbaum M, Somogyi A, Van Unruh GE, Dengler HJ. 1981. Simultaneous deter-

mination of the intravenous and oral pharmacokinetic parameters of D,L-verapamil using isotope-labeling verapamil. Eur. J. Clin. Pharmacol. 19:133-37

63. Tarrus E, Cami J, Roberts DJ, Spickett RGW,

Celdran E, Segura J. 1987. Accu-

mulation of caffeine in healthy volunteers treated with furafylline. Br. J. Clin. Pharmacol. 23:9-18 64. Gibbs MA, Thummel KE, Shen DD, Kunze KL. 1999. Inhibition of cytochrome P450 3A (CYP3A) in human intestinal and

liver microsomes: comparison of K; values and impact of CYP3A5 expression. Drug Metab. Dispos. 27:180—87

65. Shin J-G, Soukhova N, Flockhart DA. 1999. Effect of antipsychotic drugs on human liver cytochrome P450 (CYP) isoforms in vitro: preferential inhibition of CYP2D6. Drug Metab. Dispos. 27:107884 66. Miners JO, Coulter S, Birkett DJ, Goldstein JA. 2000. Torsemide metabolism by CYP2C9 variants and other human CYP2C subfamily enzymes. Pharmacokineties 10:267—70 67. Wang SL, Huang JD, Lai MD, Tsai JJ. 1995. Detection of CYP2C9 polymor-

phism based on the polymerase chain reac-

The role of the CYP2C9-Leu?*? allelic variant in the tolbutamide polymorphism. Pharmacogenetics 6:341-49 il. Takahasi Muramoto

H, Kashima N, Shimizu

T,

Nomizo Y, 1998.

T, et al.

Metabolism of warfarin enantiomers in Japanese patients with heart disease having different CYP2C9 and CYP2C19 genotypes. Clin. Pharmacol. Ther. 63:519-28

(2s Tsunoda SM, Velez RL, von Mpltke LL, Greenblatt DJ. 1999. Differentiation of intestinal and hepatic cytochrome P450 3A activity with the use of midazolam as an in vivo probe: effect of ketoconazole. Clin. Pharmacol. Ther. 66:461—71 IB. Kharasch ED, Russell M, Mautz D, Thummel KE, Kunze KL, et al. 1997. The role

of cytochrome P450 3A4 in alfentanil clearance: implications for interindividual variability in disposition and perioperative drug interactions. Anesthesiology 87:36— 50 74, Helleriegel ET, Bjornsson TD, Hauck WW. 1996. Interpatient variability in bioavailability is related to the extent of absorption: implications for bioavailability and bioequivalence studies. Clin. Pharmacol. Ther. 60:601-7 TDs Andersson T, Cederberg C, Edvardsson G, Heggelund A, Lundborg P. 1990. Effect of omeprazole treatment on diazepam plasma levels in slow versus normal rapid metabolizers of omeperazole. Clin. Pharmacol. Ther. 47:79-85 76. Bottiger Y, Tybring G, Gotharson E,

564

LIN #® LU

Berttilsson L. 1997. Inhibition of the sulfoxidation of omeprazole by ketoconazole in poor and extensive metabolizers of S-mephenytoin. Clin. Pharmacol. Ther. 62:284-91 alk Pelkonen O, Maenpaa J, Taavitsainen P, Rautio

A,

Raunio

H.

1998.

86.

Inhibition

and induction of human cytochrome P450

87.

(CYP) enzymes. Xenobiotica 28:1203-53

78. Ged C, Rouillon JM, Pichard L, CombalbertJ,Bressot N, et al. 1989. The increase

in urinary excretion of 6(-hydroxycortisol) as a maker of human hepatic cytochrome P450IIJA induction. Br J. Clin. Pharmacol. 28:373-87 WY, Kolars JC, Schmiedlin-Ren P, Sceutz JD, Fang C, Watkins PB. 1992. Identification of rifampicin-inducible P4S50II[A4

88.

89.

(CYP3A4) in human small bowel entero-

cytes. J. Clin. Invest. 90:1871-78

80. McDonnell WM, Scheiman JM, Traber PG. 1992. Induction of cytochrome P4SOIA genes (CYP1A) by omeprazole in

90.

human alimentary tract. Gastroenterology

103:1509-16 81. Rost KL, Brosicke H, Heinemeyer G, Roots I. 1994. Specific and dose-dependent enzyme induction by omeprazole in human beings. Hepatology 20:1204-12 82. Fromm MF, Dilger K, Busse D, Kroemer

Ot.

H, Eichelbaum M, Klotz U. 1998. Gut wall

metabolism of verapamil in older people: effects of rifampicin-mediated enzyme induction. Br. J. Clin. Pharmacol. 45:247a5 83. Backman JT, Olkkola KT, Neuvonen PJ. 1996. Rifampin drastically reduces plasma concentrations and effects of oral midazolam. Clin. Pharmacol. Ther. 59:7-13 84. Villikka K, Kivisto KT, Backman JT, Olkkola KT, Neuvonen PJ. 1997. Triazolam is ineffective in patients taking rifampin. Clin. Pharmacol. Ther. 61:8-14 85. Hebert MF, Roberts JP, Prueksaritanont T, Benet LZ. 1992. Bioavailability of cyclosporin with concomitant rifampin administration is markedly less than pre-

.

dicted by hepatic enzyme induction. Clin. Pharmacol. Ther. 52:453—57 Robson RA, Miners JO, Wing LMH, Birkett DJ. 1984. Theophylline-rifampicin interaction: non-selective induction of theophylline metabolic pathways. Br. J. Clin. Pharmacol. 18:445-48 Von Bahr C, Steiner E, Koike Y, Gabrielsson J. 1998. Time course of enzyme induction in humans: effect of pentobarbital on nortriptyline metabolism. Clin. Pharmacol. Ther. 64:18—26 Wilkinson GR. 1997. The effects of diet, aging and disease-states on presystemic elimination and oral drug bioavailability in humans. Adv. Drug Deliv. Rev. 27:129-59 Juan D, Worwag DM, Schoeller DA, Kotake AN, Hughes RL, Frederiksen MC. 1986. Effects of dietary protein on theophylline pharmacokinetics and caffeine and aminopyrine breath test. Clin. Pharmacol. Ther. 40:187—94 Kappas A, Anderson KE, Conney AH, Alvares AP. 1976. Influence ofdietary protein and carbohydrate on antipyrine and theophylline metabolism in man. Clin. Pharmacol. Ther. 20:643-—53 Conney AH, Pantuck EJ, Hsiao KC, Garland WA, Anderson KE, Alvares AP, Kappas A. 1976. Enhanced phenacetin metabolism in human subjects fed charcoal-broiled beef. Clin. Pharmacol. Ther. 20:633—42 Kappas A, Alvares AP, Anderson KE, Pantuck EJ, Pantuck CB, et al. 1978. Effects of charcoal-broiled beef on antipyrine and theophylline metabolism. Clin. Pharmacol. Ther. 23:445—50

. Kall MA, Clausen J. 1995. Dietary effect on mixed function P4501A2 activity assayed by estimation of caffeine metabolism in human. Hum. Exp. Toxicol. 14:801-7

94, Kall MA, Vang O, Clausen J. 1996. Effects of dietary broccoli on human in vivo drug metabolizing enzymes: evaluation of caffeine,

oestrone

and chlorzoxazone

metabolism. Carcinogenesis 17:793-99

VARIABILITY IN DRUG-DRUG INTERACTIONS

O51 Pantuck EJ, Pantuck CB, Min BH, Wattenberg LW, Anderson

KE, et al. 1979.

Stimulatory effect of brussel sprouts and cabbage on human drug metabolism. Clin. Pharmacol. Ther. 25:88—95 96. Zevin S, Benowitz NL. 1999. Drug interactions with tobacco smoking. Clin. Pharmacokinet. 36:425-38 Oe Pelkonen O, Pasanen M, Kuha H, Gachalyi B, Kaivraluoma

M, et al. 1986. The

effect of smoking on 7-ethoxyresorfin O-deethylase and other monooxygenase activities in human liver: analysis with monoclonal antibodies. Br. J. Clin. Pharmacol. 22:125-134

98. Sesardic D, Boobis AR, Edwards RJ, Davis DS. 1988. A form of cytochrome P450 in man, orthologus to form d in the rat, catalyzes the O-deethylation of phenacetin and is inducible by cigarette smoking. Br. J. Clin. Pharmacol. 26:363— 2)

99. Buchthal

J, Grund

KE,

Buchman

A,

Schrenk D, Beaune P, Bock KW. 1995. In-

duction of cytochrome P4501A by smoking or omeprazole in comparsion with UDP-giucurosyl transferase in biopsies of human duodenal mucosa. Eur. J. Clin. Pharmacol. 47:431-—35

100. Hunt SN, Jusko WJ, Yurchak AM. 1976. Effect of smoking on theophylline disposition. Clin. Pharmacol. Ther. 24:546-S1

101. Jusko WJ, Schentag JJ, Clark JH, Gardner

M,

Yurchak

AM.

1978.

Enhanced

biotransformation of theophylline in marihuana and tobacco smokers. Clin. Pharmacol. Ther. 24:405-10

102. Vistisen K, Loft S, Poulsen HE. 1991. Cytochrome P450 1A2 activity in man measured by caffeine metabolism: effects of smoking, broccoli and exercise. Adv. Exp. Med. Biol. 283:407-11 103. Spigset O, Carleborg L. Hedenmalm K, Dahlqvist R. 1995. Effect of cigarette smoking on fluvoxamine pharmacokinetics in humans. Clin. Pharmacol. Ther. 58:399-403

565

104. Fulton B, Goa KL. 1997. Olanzapine: a review of its pharmacological properties and therapeutic efficacy in the management of schizophrenia and related psychoses. Drugs 53:281—98 105. Hasegawa M, Gutierrez-Esteinou R, Way L. 1993. Relationship between clinical efficacy and clozapine concentrations in plasma in schizophrenia: effect of smoking. J. Clin. Psychopharmacol. 13:38390 106. Ochs HR, Greenblatt DJ, Burstein Es. 1987. Lack of influence of cigarette smoking on triazolam pharmacokinetics. Br. J. Clin. Pharmacol. 23:759-63 107. Madan A, Dehaan R, Mudra D, Carroll K, Lecluyse E, Parkinson A. 1999. Ef-

fect of cryopreservation on cytochrome P-450 enzyme induction in cultured hepatocytes. Drug Metab. Dispos. 27:327-35

108. Kostrubsky

VE,

Ramachandran _ V,

Venkataramanan R, Dorko K, Esplen JE,

et al. 1999. The use of human hepatocyte cultures to study the induction of cytochrome P-450. Drug Metab. Dispos. 27:887-94 109. Chang TK, Yu L, Maurel P, Waxman DJ. 1997. Enhanced cyclophosphamide and ifosfamide activation in primary human hepatocyte cultures: response to cytochrome P-450 inducers and autoinduction by oxazaphosphorines. Cancer Res. 57:1946—54 110. Kato R, Takanaka A. 1968. Effect of phenobarbital on electron transport system, oxidation and reduction of drugs in liver microsomes of rats of different age. J. Biochem. (Tokyo) 63:406-8 AIDE Kamataki T, Maeda K, Shimada M, Kitani K, Nagai T, Kato R. 1985. Agerelated alteration in the activities of drugmetabolizing enzymes and contents of sex-specific forms of cytochrome P-450 in liver microsomes from male and female rats. J. Pharmacol. Exp. Ther. 233:222-28 1 Schmucker DL, Vessey DA, Wang RK, Maloney AG. 1986. Aging, sex and the

566

LIN # LU

liver microsomal drug-metabolizing system in rodents and primates. In Liver and Aging-1986: Liver and Brain, ed. K Kitani, pp. 3-12. Amsterdam: Elsevier 1S: Schmucker DL, Woodhouse KW, Wang RK, Wynne

H, James

LZ,

OF, et al. 1990.

Effects of age and gender on in vitro properties of human liver microsomal monooxygenase. Clin. Pharmacol. Ther. 48:365-74

114. Hunt CM, Westerkam WR, Stave GM, Wilson JAP. 1992. Hepatic cytochrome P4503A (CYP3A) activity in the elderly. Mech. Ageing Dev. 64:189-99 MS). Hunt CM, Westerkam WR, Stave GM. 1992. Effects of age and gender on the activity of human hepatic CYP3A. Biochem. Pharmacol. 44:275—83 116. Sotaniemi EA, Arranto AJ, Pelkonen O, Pasanen M. 1997. Age and cytochrome P450-linked drug metabolism in humans: an analysis of 226 subjects with equal histopathologic conditions. Clin. Pharmacol. Ther. 61:331—39 Wa Landahl S, Andersson T, Larsson M,

123:

124.

.

Lernfeldt B, Lundborg P, et al. 1992. Phar-

macokinetic study of omeprazole in elderly healthy volunteers. Clin. Pharmacokinet. 23:469-76 118. Twum-Barima Y, Finnigan T, Habash AI, Cape RDT, Carruthers SG. 1984. Impaired enzyme induction by rifampicin in the elderly. Br J. Clin. Pharmacol. 17:595—97 ILI). Salem SAM, Rajjayabun P, Shepherd AMM, Stevenson IH. 1978. Reduced induction of drug metabolism in the elderly. Age Ageing 7:68-73 120. Crowley JJ, Cusack BJ, Jue SG, Koup JP, Park BK, Vestal RE. 1988. Ageing and drug interactions. II. Effect of phenytoin and smoking on the oxidation of theophylline and cortisol in healthy man. J. Pharmacol. Exp. Ther. 245:513-23 al Zhou HH, Anthony L, Wood AJJ, Wilkinson GR. 1990. Induction of polymorphic 4-hydroxylation of S-mephenytoin by ri-

126.

127.

.

fampicin. Br. J. Clin. Pharmacol. 30:471— 75 Dilger K, Greiner B, Fromm MF, Hofmann U, Kroemer HK, Eichelbaum M. 1999. Consequences of rifampicin treatment on propafenone disposition in extensive and poor metabolizers of CYP2D6. Pharmacogenetics 9:551—5S9 Rost KL, Brosicke H, Brockmoller J, Scheffler M, Helge H, Roots I. 1992. Increase of cytochrome P450IA2 activity by omeprazole: evidence by the C-[N-3methyl]-caffeine breath test in poor and extensive metabolizers of S-mephenytoin. Clin. Pharmacol. Ther. 52:170—80 Aoyama T, Yamano S, Waxman DJ, Lapenson DP, Meyer VA, et al. 1989. Cytochrome P-450 hPCN3, a novel cytochrome P-450III gene product that is differentially expressed in adult human liver. J. Biol. Chem 264:10388—95 Wnighton SA, Ring BJ, Watkins PB, Vanden Branden M. 1989. Identification of a polymorphically expressed member of human cytochrome P-450III family. Moi. Pharmacol. 86:97—105 Wrighton SA, Brian WR, Sari MA, Iwasaki M, Guengerich FP, et al. 1990. Studies on the expression and metabolic capabilities of human _ cytochrome P4S50INAS (HLp3). Mol. Pharmacol. 38:207-13 O'Reilly RA. 1974. Interaction of sodium warfarin and rifampicin. Ann. Intern. Med. 81:337-40 MacDonald MG, Robison DS, Sylwester D. 1969. The effects of phenobarbital, chloral betaine and glutethimide admin-_ istration on warfarin plasma levels and hypoprothrombinemic responses in man. Clin. Pharmacol. Ther. 10:80-84

. Barbarash RA, Bauman JL, Rischer JH,

Kondos GR, Batenhorst RL. 1987. The effect of enzyme induction on verapamil pharmacokinetics: verapamil-rifampicin interaction. Drug Intell. Clin. Pharm. 21:11A (Abstr.)

VARIABILITY IN DRUG-DRUG INTERACTIONS

130. Breckenridge

A, Orme M, Davies L, Thorgeirsson S, Davies D. 1973. Dosedependent enzyme induction. Clin. Pharmacol. Ther. 14:514-20

rie

Alvan G, Piafsky K, Lind M, von Bahr C.

132

1977. Effect of pentobarbital on the disposition of alprenolol. Clin. Pharmacol. Ther. 22:316-21 Rutledge DR, Pieper JA, Mirvis DM. 1988. Effects of chronic phenobarbital disposition in humans. J. Pharmacol. Exp. Ther. 246:7-13

133) Twum-Barima Y, Carruthers SG. 1981. Quinidine-rifampicin interaction. N. Engl. J. Med. 304:1466—-69 134. Ito K, Iwatsubo T, Kanamitsu S, Ueda K,

567

Suzuki H, Sugiyama Y. 1998. Prediction of pharmacokinetic alterations caused by drug-drug interactions: metabolic interaction in the liver. Pharmacol. Rev. 50:387-11 33%

Sugiyama

Y, Iwatsubo

T, Ueda

K, Ito

K. 1996. Strategic proposals for avoiding toxic interactions with drugs for clinical use during development and after marketing of a new drug: pharmacokinetic considerations. J. Toxicol. Sci. 21:309-16 136. Von Moltke LL, Greenblatt DJ, Schmider J, Wright CE, Harmatz JS, Shader RI. 1998. In vitro approaches to predicting drug interactions in vivo. Biochem. Pharmacol. 55:113—22

Tas

UI

Annu. Rey. Pharmacol. Toxicol. 2001. 41:569-91 Copyright © 2001 by Annual Reviews. All rights reserved

NEUROTROPHIC AND NEUROPROTECTIVE ACTIONS OF ESTROGENS AND THEIR THERAPEUTIC

IMPLICATIONS

Susan J Lee and Bruce S McEwen Harold and Margaret Milliken Hatch Laboratory of Neuroendocrinology, The Rockefeller University, New York, New York 10021; e-mail: McEwen @mail.rockefeller.edu, leesus @mail.rockefeller.edu

Key Words estrogen receptor, ERa, ER£, membrane-associated ER, estradiol 17a, estradiol 178, second messengers, neurotrophins, oxidative damage, NMDA, f-amyloid @ Abstract Originally known for its regulation of reproductive functions, estradiol, a lipophilic hormone that can easily cross plasma membranes as well as the blood-brain barrier, maintains brain systems subserving arousal, attention, mood, and cognition. In addition, both synthetic and natural estrogens exert neurotrophic and neuroprotective effects. There is increasing evidence that estrogen actions are mediated by nongenomic as well as direct and indirect genomic pathways. Although in vitro models have provided the most extensive evidence for neurotrophic and neuroprotective actions to date, there are also in vivo studies that support these actions.

INTRODUCTION Since its discovery and recognition as a “female” sex hormone (1), estradiol has

been studied for its effects on the female reproductive tissues as well as its actions in the reproductive neuroendocrine system, and most recently for its ability to affect other aspects of brain function. The pioneering work of Jensen and others, demonstrating intracellular estrogen receptors (ERs) (2), introduced the use of tritium-labeled steroid hormones, and this led quickly to the discovery of estrogen-concentrating cells in the pituitary gland, hypothalamus, and other brain regions (3-5). Only much more recently has interest in estrogen actions on the brain shifted toward brain areas associated with cognitive and other functions not

directly connected to reproduction (6). These studies have revealed actions of estrogens on many major neurotransmitter systems and brain regions subserving cognitive, emotional, and vegetative functions, as well as an important role in brain development and in the emerging area of protecting nerve cells from damage.

0362-1642/01/0421-0569$14.00

569

570

LEE ® McEWEN

Estradiol 17£ is a steroid hormone with many diverse cellular effects in neural tissue. It is produced by the ovaries and brain, as well as fat tissue by the aromatization of testosterone. Thus, estradiol is a hormone in males as well as in females.

The neural effects of estrogens include a neurotrophic role in such processes as cell proliferation and differentiation, neuronal survival, and synaptogenesis during a sensitive developmental period, which leads to the establishment of the sexually differentiated brain (7-9). Estrogens also have neurotrophic effects on the adult brain, promoting collateral axonal sprouting in the deafferented hypothalamus (10) and promoting synaptogenesis in the female rat hippocampus (11, 12). Estradiol also has neuroprotective effects in the aged or injured brain, which are evident in relation to Alzheimer’s disease (13—15) and damage from ischemia (16-19). Estradiol has been shown to (a) activate nuclear ERs, (b) interact with

a form of ER that activates second messenger systems, (c) induce anti-apoptotic gene expression, (d) maintain intracellular calcium homeostasis, (e) promote antioxidant activity, and (f) modulate actions of neurotrophins. Our aim in this article is to highlight the neuromodulatory role of estrogen as a neurotrophic and neuroprotective agent.

SITES OF ESTROGEN ACTION IN BRAIN Intracellular ERs are found in both sexes starting early in development, along with aromatizing enzymes that convert testosterone and androstenedione to estradiol and estrone, respectively (7,20). In rodent brains ERs are expressed along with aromatase in the hypothalamus starting around day 14 of gestation. ERs are later expressed in other brain areas such as the amygdala, midbrain, hippocampus, and spinal cord. In the rodent hippocampus and cerebral cortex there is a transient peak of ER expression in the first week after birth (7, 21). In general, besides the hypothalamus and the amygdala in the adult brain, ER is expressed in cholinergic, 5-serotonergic, noradrenergic, and dopaminergic neural systems as well as in specific GABA interneurons of the hippocampus and cerebral cortex (6). There are two currently known primary ERs, referred to as ERa and ERB; knockout mice for both ER isoforms have been generated (for reviews see 22, 23).

ERa-deficient (WERKO) mice are infertile and show major deficits in sexual behavior, whereas ERf-deficient mice (BERKO) appear to be less profoundly affected. From both immunocytochemical and mRNA studies, ERa is localized to the hypothalamus, amygdala, and scattered neurons in other brain regions such as the midbrain, hippocampus, and cerebral cortex. ERB mRNA is found in the hypothalamus and is even more evident in other brain areas such as the midbrain,

cerebellum, hippocampus, and cerebral cortex, but the antibodies for ERB have been less reliable, and localization of ERB immunoreactivity remains problematic (6). Both ER isoforms are expressed in the hypothalamus and midbrain and to some extent in the hippocampus and cerebral cortex, whereas ERB appears to be the exclusive isoform in the substantia nigra and cerebellum (24, 25).

ACTIONS OF ESTROGENS

571

Complementary to the in situ and immunocytochemical findings, in vivo ligandbinding studies originally performed with *H-estradiol (4) and recently performed

by Shughrue & Merchenthaler, using '*I-estrogen (26), show binding to all highaffinity ER isoforms. The '*I-estrogen uptake method appears to be more sensitive than *H-estradiol uptake and reveals some ERs where previously none had been detected. For example, in the original *H-estradiol uptake study only hippocam-

pal interneurons were labeled (4, 27), whereas in the recent !*>I-estrogen binding studies, cell nuclear estrogen binding sites were reported in pyramidal cells of the ventral CA1—3 of the hippocampus and laminae II-VI of the isocortex. Although both ER isoforms are found primarily in cell nuclei of target cells (25, 28), there are also indications of the presence of ER in other parts of the cell (29-33). Recent ultrastructural evidence showed the presence of ERa-

immunoreactivity (Ir) in extranuclear sites in the hippocampus, namely in some axon terminals, dendritic spines, and glial processes (32). Although the functional significance of these extranuclear receptors is not yet known, the discussion below points to possible intracellular signaling mechanisms for the extranuclear forms of ER. Additionally, the enzyme that converts testosterone to estradiol, the cytochrome p450 aromatase complex, is localized throughout the neuronal perikarya, including dendrites, axon processes, and synaptic vesicles in the hypothalamic and limbic cell groups in Japanese quail, rat, monkey, and human (30). Taken together, these findings suggest the possibility that some brain cells express a local mechanism to regulate rapid estrogen synthesis and action at the synaptic level. This is a dramatic change from the dogma of nuclear ER and, as such, requires that

substantial additional data be considered as fact. This new view of estrogen action may be relevant to the fact that estrogen regulates synaptic connectivity in the hypothalamus as well as in the hippocampus in vivo (12, 34, 35). Similarly, in vitro studies on primary cultures of hippocampal neurons have revealed that estrogen induction of new dendritic spines and synapses involves the interactions of GABAergic interneurons and brain-derived neurotrophic factor (BDNF) (36). Estradiol treatment induces biphasic responses in GABAergic cells, initially decreasing GABA-Ir and GAD-Ir (synthesizing enzyme for GABA) and subsequently increasing both GABA and GAD (37). Hippocampal interneurons contain cell nuclear ERa-Ir (37, 38). Reminiscent of the develop-

mental expression in vivo, ERa-Ir increases in primary hippocampal culture for 7 days, during which time neurons undergo proliferation and differentiation, and then slowly decreases by day 28 in vitro (38). Others have shown that at 3-5 days in vitro, hippocampal neurons exhibit low levels of nuclear staining for ERa but exhibit abundant nonnuclear, peri-plasma membrane neurite labeling (33). It is conceivable that both nuclear and membrane-associated forms of ERa and ERB may be involved in mediating neurotrophic and neuroprotective estrogen effects. As shown below, nonnuclear estrogen effects are likely to involve activation of second messenger systems and transcription factors that are regulated via phosphorylation and dephosphorylation, some of which are linked to cell survival. Other nonnuclear estrogen actions are likely to occur independently of

572

LEE ® McEWEN

either nuclear or nonnuclear ER, based upon their structure-activity profile and concentration dependence in the micromolar range.

MECHANISMS OF ESTROGEN ACTION The actions of estrogens will therefore be categorized as (a) direct genomic actions, (b) indirect genomic actions via second messenger cascades, (c) nongenomic effects at low estrogen levels, and (d) nongenomic effects at high estrogen levels

involving antioxidant actions. It is important to keep in mind that there is a lack of information in some cases that would clearly identify which of these mechanisms may be involved in particular cellular processes or effects of estrogens (see Figure 1).

Direct Genomic Mechanism The classical direct genomic mechanism involves the nuclear form of ERa or ERB. Once bound by estrogen, the nuclear ligand-receptor complex acts as a transcription factor by binding either to an estrogen response element (ERE) or to fos-jun heterodimers, which in turn bind to an activation protein-1 (AP-1) response element (39-41). This direct genomic mechanism requires >45 min for new protein synthesis and probably much longer to alter cellular response (42).

Indirect Genomic Mechanism The newly discovered indirect genomic actions of estradiol are postulated to occur when activation of a form of ER, possibly associated with cell membranes, stimulates a second messenger system such as adenylyl cyclase (AC), protein kinase A (PKA), protein kinase B (PKB), also known as Akt, protein kinase C (PKC),

and mitogen-activated protein kinase (MAPK), also known as extracellular signalrelated kinase (ERK) (43-48). This results in phosphorylation of various cellular substrates ranging from membrane and cytoplasmic proteins to transcriptional regulators, such as cAMP-response element-binding protein (CREB by PKA) (49-52) or serum response factor (SRF)-Elk-1 complex (by MAPK/ERK) (for review see 48). These specific gene regulatory proteins (CREB, SRF-EIk-1) act at the DNA regulatory regions cAMP responsive element (CRE) and serum response element(SRE), respectively. The resulting cascades are capable of regulating non-EREcontaining genes such as MAP2, £-tubulin (53, 54), and GAP43 (55), presumably

by activating the MAPK/ERK pathway. Although steroids have long been suspected of affecting second messenger systems, based upon studies of oocytes (56) and cancer cells (46), recent evidence

reinforces a direct coupling of estrogen effects to a G-protein-linked second messenger cascade. The most recent, direct evidence comes from work in the hypothalamus (43), where estrogen has been shown to modulate receptor/effector coupling and/or expression of the genes encoding all major classes of G-protein-coupled

ACTIONS OF ESTROGENS

— ie)

E,/ Neurotrophins Indirect genomic —>

> Gus / Gaq 2°4 messenger regulation

| Non-genomic

*

©

0 &

v mene

xe

spe

hea

Ras/Raf

PKA, Akt, PKC, JNK

CREB, NF-«B, Jun, Rsk, p38 Other effects

ea ERE

CRE} SRE bom

Transcriptional effects

9 Direct genomic

Neuroprotection / cel! survival cell growth / plasticity

Figure 1 Putative mechanisms of estrogen action. In the direct genomic mechanism the nuclear form of ERa@ or ERf associates with either the ERE or fos/jun heterodimers that bind, in turn, to AP1 sites. Indirect genomic mechanisms include the activation of an ER linked to second messenger systems such as AC/PKC, cAMP/PKA and MAPK/ERK, converging with the genomic pathway. In one of these pathways, Ras activates Raf, which leads to sequential phosphorylation and activation of MAPK/ERK. Activated ERK then translocates into the nucleus to interaci directly with nuclear transcription factors (e.g. CREB, cfos/cjun), and indirectly through the activation of intermediary signaling proteins (e.g. Rsk, p38, JNK) to bind to the DNA regulatory regions CRE and SRE. Neurotrophins and estrogens may influence each other’s actions by regulating receptors and/or ligand availability through reciprocal regulation at the genomic level. Nongenomic estrogen effects at high concentrations involve antioxidant effects not mediated by known intracellular ERs. ERE,

AP-1, SRE, and CRE are regulatory regions in DNA sequences that are recognized by specific generegulatory proteins. ERE is recognized by estrogen-ER complexes; AP-1 is recognized by fos/jun heterodimers; CRE is recognized by phospho-CREB (phosphorylated by PKA in response to a rise in CAMP levels); SRE is recognized by SRF-Elk-1 complex phosphorylated by MAPK/ERK. The MAPK/ERK migrates from the cytoplasm to the nucleus and phosphorylates Elk-1, thereby activating it to turn on the transcription of the fos gene. MAPK/ERK and PKC can phosphorylate jun protein, which combines with the newly formed fos to form heterodimers that ultimately bind to AP-1. x, estriol;

, 17a-estradiol; e, 178 estradiol; ERE, estrogen response element; CRE,

cAMP response element; SRE, serum response element. See text for other abbreviations.

574

LEE ® McEWEN

receptors. Specifically, estrogen activates a membrane-associated ER that stimulates PKC, which can rapidly uncouple the jz (opioid) and GABAg receptors via a G-protein to K+ channel. This, in turn, reduces 6-endorphin and GABA binding or increases B-endorphin and GABA release, implying that estrogen modulates signaling pathways shared by both y-opioid and GABAg receptor

systems. Furthermore, the activated PKC can also stimulate AC to produce cAMP, which in turn activates the PKA. The rapid ER-mediated activation of PKA leads to the phosphorylation of CREB (pCREB), which can then alter gene transcription through interaction with the regulatory region on a DNA sequence, CRE, altering gene expression. Thus, estrogen can rapidly increase PKA activity, which has the ability to modulate the coupling of G-protein-coupled receptors to their effector systems (43). A selective PKA antagonist (KT5720 and Rp-cAMP) can block the effects of estrogen. Whether this general mechanism exists throughout the central

nervous system remains to be seen. Second, a critical study by Levin (44) has provided further information about the source of the estrogen receptor involved in coupling to second messenger systems. This study demonstrated that a single transcript arising from transfection of either ERa or ERB cDNA into Chinese hamster ovarian cells leads to expression of around 3% of the receptor as a membrane-associated form of ER and 97% as the nuclear ER. Both forms of ER show similar estrogen affinities and specificities. The membrane ER activates the a subunits of several GTP-binding proteins (e.g. Gag and Gas) and rapidly stimulates inositol phosphate and adenylyl cyclase activity. This leads to downstream signal transduction cascades, such that Gag activates PLC and increases IP3 and Ca2+ and then activates ERK activity. On the other hand, Gas stimulates AC, increases cAMP

generation and activates PKA. This

study further demonstrated that estradiol promotes and acts broadly through a G-protein-coupled receptor—dependent mechanism in a wide variety of cellular responses throughout the CNS. These second messenger pathways (AC — PKA; PKC -> AC + PKA; MAPK/ ERK; Ras + ERK —> Rsk) may be activated in response to an influx of extracellular Ca’*, the binding of neurotrophins (nerve growth factor or BDNF) to their cognate receptors, or the binding of estrogens to a membrane-associated ER. The MAPK/ERK cascade is a major signal transduction pathway, which can be turned on by a wide range of extracellular proliferation- and differentiation-inducing sig-nals. Activated MAPK/ERK can translocate to the nucleus and interact directly with the nuclear regulatory sequences: CRE, SRE, and AP-1 (57). These are the DNA sites to which the factors bind, which can subsequently lead to the phosphorylation and activation of transcription factors in the nucleus, such as CREB and immediate early genes such as c-fos/c-jun and AP-]. Ras (a member of a large family of GTP-binding proteins that help relay signals from membrane receptors to the nucleus) activates Raf, which sequentially phosphorylates and activates MAPK kinase (MEK) and MAPK/ERK. MAPK/ERK can phosphorylate another kinase, Rsk, which can then also translocate to the nucleus to interact indirectly with

ACTIONS OF ESTROGENS

575

nuclear response elements (57). Figure 1 shows a schematic overview of known ER signaling events, genomic and nongenomic, as well as those linked to the form of ER that activates second messengers.

The MAPK pathway is thought to play an important role in the action of neurotrophins and in synaptic plasticity (58). Moreover, as discussed below, estrogenmediated neuroprotection may involve the tyrosine kinase/MAPK signal transduction cascade, because estrogen can rapidly activate tyrosine kinase and MAPK activity and because its effects are blocked by inhibitors of tyrosine kinase and MAPK (59). In summary, although further studies must demonstrate that membraneassociated ER actually activates second messenger functions in vivo, the existence of ER in dendritic spines and some presynaptic endings in the hippocampus (32) suggest possible signaling from the synapse to the cell nucleus.

Nongenomic Effects at Low Estrogen Concentrations In addition to the above mechanisms, estradiol has rapid nongenomic effects that enhance the amplitude of kainate-induced currents of CA1 neurons (60, 61) or in-

hibit calcium currents in striatal neurons (62). Although these effects are produced by nanomolar or lower levels of estradiol, the pharmacology of these estrogen effects is somewhat different from those likely to involve intracellular ER, based upon the lack of effects of nonsteroidal estrogen antagonists. Thus, a different estrogen receptor mechanism may be involved (6). It should also be noted that all of the intracellular ER—-related second messenger pathways depicted in Figure | and discussed above can result in the phosphorylation of membrane and cytoplasmic proteins and that, in principle, estradiol is thus capable of affecting a host of nongenomic events throughout the cell. However, based upon the report of Razandi et al (45), it would be expected that these pathways should show inhibition of estrogen action by nonsteroidal estrogen antagonists.

Nongenomic Effects at High Estrogen Concentrations Involving Antioxidant Effects Other nongenomic effects clearly involve a receptor system other than the intracellular ER. This is so because of the structure-activity profile in which 17@ and 176 estradiol are equally potent and because, generally speaking, micromolar, rather than nanomolar, concentrations of estradiol are required for this class of effects,

which are largely neuroprotective in various cell culture models. Neuroprotective actions of estradiol in vitro were first described using SK-N-SH human neuroblastoma cells under serum deprivation (63). Estradiol 176 (2 4M) increased total live cell number for up to 48 h without increasing thymidine incorporation, indicating an effect on cell survival and not cell division. Since then there have been numerous studies, all reporting the neuroprotective effects against various neurotoxic conditions (see Table 2).

576

LEE ® McEWEN

At varying concentrations of 17a or 17£ estradiol (from pM-/4M) with different cell types, the responses range from enhancement of survival, to facilitation of neurite outgrowth, to prevention of cell death. These various effects were blocked by an NMDA receptor antagonist (APS) but not by an estrogen antagonist (ICI 182,780), again confirming that nuclear ERs are not involved. Because both 17a and 17 estradiol have been shown to be effective, with 17a being more effective than 17f in many cases, this may reflect a novel estrogen receptor distinct from the intracellular ER a and £. However, there are also numerous protective effects that involve second messenger activation, and these may involve intracellular ER @ or B, which are discussed below, along with other examples of the neuroprotective actions that involve micromolar levels of various estrogens.

NEUROTROPHIC EFFECTS OF ESTROGENS Neurotrophic factors promote the growth, survival, and maintenance of neurons. Since the discovery of nerve growth factor, a large number of neurotrophic factors have been discovered. Cell culture techniques have become the standard methods for assaying putative neurotrophic factors and identifying responsive cell types. Estrogens, which have neurotrophic effects, are also able to interact with the neu-

rotrophins. The neurotrophic effects of estrogen are summarized in Table 1. The first neurotrophic actions of estrogens on cell survival/growth were demonstrated using organotypic slice cultures of the developing hypothalamus, preoptic area, and cerebral cortex (48,64). At about the same time, estrogens came to be recognized for their role in sexual differentiation of the brain as metabolites of testosterone (see above). These estrogenic actions bear a general similarity to the neurotrophins, a family of growth factors, and their cognate receptors, tropomyosin-related kinase receptors (trk): (a) nerve growth factor (nerve growth factor binds trkA), (b) brain-derived neurotrophic factor (BDNF binds trkB), (c) neurotrophin-3 (NT-3 binds trkC), and (d) neurotrophin 4/5 (NT4/5 binds trkB).

Widespread colocalizations of ER and neurotrophin receptors (trks) are found mainly in neurons of the cerebral cortex, hypothalamus, hippocampus, and sensory ganglia. This receptor co-expression helps to explain the fact that estradiol and neurotrophins generally appear to exert reciprocal regulation upon each other’s actions at the level of gene transcription. For example, EREs are present in the low-affinity;

neurotrophin receptor and the pan-neurotrophin receptor, p75‘™®, as well as in trkA. As a result, following ovariectomy, ER mRNA trkA mRNA

expression peaks while

expression decreases, or vice versa (48, 65). In addition, it has been

reported that estrogen treatment alters the neurotrophin receptor (p75‘'®/trkA) ratio, which can subsequently alter ligand autophosphorylation as well as neurotrophin binding affinities. Thus, the intricate reciprocal regulation of trks and ERs emphasizes the advantages of having multiple “fail-safe” neurotrophic mechanisms that converge upon and cross-couple with known MAPK signaling pathways

(48).

ACTIONS OF ESTROGENS TABLE 1

S77

Neurotrophic effects of estrogen

Brain Regions/ Neuronal/ Cell Types

Steroid

Modulation

Effective Concentration

Hypothalamic explant

17 B-estradiol

Neurite growth

100 ng/ml

Hypothalamic neurons

17 B-estradiol

Up-regulates IGF

10-°M

100

Hypothalamic neurons

17 B-estradiol

Neurite outgrowth

0.1 uM

101

Hypothalamic neurons

17 B-estradiol

Neuronal survival

107! M

Amygdala neurons

17 B-estradiol

Neurite outgrowth

100 nM

102

Neuronal differentiation Dendritogenic growth

100nM 100 nM

102 102

Neuronal differentiation

10-100 nM

103

No proliferation

10-100 nM

103

| Up-regulates Bcl-x

1 nM

104

Neuronal survival

1 nM

104

Cortical neurons Dorsal root ganglion

17 B-estradio!

17 a- and B-estradiol

Ref. 48

Hippocampal neuronal

17 B-estradiol

Up-regulates BDNF

105

Hippocampal neuronal

17 B-estradiol

Down-regulates BDNF

Dopaminergic neurons

17 6-estradiol

Neurite outgrowth

1 pM-10 aM

106

17 f-estradiol

Increase TH mRNA

1 pM-10 nM

106

Dopaminergic neurons

17 B-estradiol

Increase DA uptake

10-'4*M

107

Neocortical cells

17 B-estradiol

Neurite outgrowth

1 nM

108

37

An additional way in which estrogens seem to fine-tune neurotrophin responses is seen in the hippocampus in vivo and in vitro. Ovarian hormones regulate hippocampal BDNF mRNA levels in vivo, showing fluctuations across the estrous cycle (66). Moreover, whereas estrogen treatment has been reported to increase BDNF mRNA levels in the whole hippocampus of ovariectomized rats (67), estradiol treatment has also been reported to transiently decrease BDNF expression in inhibitory hippocampal interneurons; this decrease, along with transiently

578

LEE ® McEWEN

decreased GABAergic activity in the same cells, is hypothesized to result in disinhibition of pyramidal cells, allowing the activity-dependent increase in dendritic spine density (37). BDNF treatment can block the estradiol-induced increase in spine density, whereas blocking BDNF with neutralizing antibodies mimics the effects of estradiol. It is noteworthy that the BDNF gene contains a putative estrogen response element (68).

Estrogens and neurotrophins can rapidly phosphorylate the estrogen receptor (48). In neural and nonneural tumor cell lines, estrogen effects that are presumably plasma membrane—mediated are very rapid, occurring in seconds to minutes; this may involve a shared estrogen and growth factor signaling pathway for cell proliferation and ER phosphorylation. For example, in mammary tumor cells (MCF-7) estradiol can elicit maximal phosphorylation of Sre within 10 seconds (46). It is still not clear whether the estrogen receptors that mediate the growthpromoting properties of estrogen in the developing brain are the classical receptors, ERa and ER, or perhaps another yet unidentified receptor subtype. However, there is growing evidence showing that estrogen receptors can mediate extracellular signaling in both an estrogen-dependent and estrogen-independent manner through growth factor signaling pathways (MAPK/ERK) (69). Toran-Allerand and colleagues (48) have suggested another potential pathway for estrogen-induced MAPK/ERK activation. This involves a multimeric caveolar-associated complex consisting of RAF, hsp 90, ER, Src, and ERK (58). This pathway has been implicated in signal transduction and vesicular trafficking (for review see 48). Interestingly, like many of the neurotrophic factors, estrogen stimulation of neurite growth is developmentally regulated and hence is not seen in normal adults. However, after neural damage or loss of estrogens, responsiveness to estrogen can once again be demonstrated. Manifestations of these effects include reinnervation of deafferented neurons in the surgically isolated hypothalamus (10) and the induction of new dendritic spines and synapse in the undamaged hippocampus of ovariectomized female rats (11, 12,70).

Although the maintenance of the capacity for growth and regeneration is a component of neuroprotection, the protection from damage is another aspect. As noted above and discussed in detail below, there are estrogen effects that reduce free radical production that are not stereospecific for 178 estradiol and may therefore involve receptors other than ERa or ERB.

NEUROPROTECTIVE EFFECTS OF ESTROGENS The decline of estrogen levels, whether natural or surgically induced, has been implicated in the etiology or the progression of age-associated neurodegenerative States. Estrogen replacement therapy has been helpful for some postmenopausal women. The benefits include improved cognitive function/mental performance, particularly verbal memory (6), or delayed onset of neurodegenerative disorders, such as Alzheimer’s (13, 15) and possibly also Parkinson’s diseases (71), as well as

ACTIONS OF ESTROGENS

579

reduced incidence of osteoporosis and cardiovascular diseases (72). There are also

reports of estrogen neuroprotection from ischemic damage (16-19). However, much of the progress regarding mechanism has involved the use of cell culture (see Table 2).

Anti-Apoptotic Effects Another mechanism by which estrogens might exert their neuroprotective effects is via modulation of molecules involved in the programmed cell-death pathway. The negative regulators of cell death are anti-apoptotic proteins, Bcl-2 and Bcl-xl, whereas the positive regulators of apoptosis are Bax, Bad, and Bid. In cultured

hippocampal neurons, estrogen increases Bcl-xl expression and decreases both caspase-mediated proteolysis and cell death induced by £-amyloid (73). In fact, Pike (73) showed colocalization of ER and Bcl-x in human neuronal populations that exhibit relative resistance to Alzheimer’s disease neurodegeneration. It has been suggested that estrogen treatment can enhance neuronal resilience to apoptotic insult via an ER-dependent genomic pathway that involves increased anti-apoptotic protein Bcl-xl, with subsequent inhibition of pro-apoptotic steps and capase-mediated proteolysis. In substantia nigra dopaminergic neurons, estrogens provide neuroprotection against apoptosis via an ER mechanism that involves the AP-1 response element and the ERB receptor, which predominates in this brain region (74). It appears likely that regulation of Bcl-2 expression is involved in the dopaminergic neurons (74). An estrogen-mediated increase in Bcl-2 expression, associated with neuroprotection, has been reported in the cerebral cortex after ischemia (16). This study also showed that other members of the apoptotic family (bax, bcl-xl, bcl-xs, and bad) were not changed by estrogen treatment. In addition, estradiol increased the ratio of ERB/ERa in the ischemic model, and it was suggested that ERAdependent signaling may be linked with neuroprotection (40), although the data for this suggestion are correlative. In relation to Bcl-2 regulation, recent studies (75-77) demonstrated that in the neuroblastoma cell line (SK-ER3), the mRNA of

the pro-apoptotic gene Nip2, which interacts with Bcl-2, is significantly decreased following estradiol treatment (75, 77). The exact role of Nip2-Bcl-2 in neural cell apoptosis is not known. However, Nip2 has been shown to influence brain maturation such that Nip2 mRNA transiently increases (embryonic day 15—20) and then

decreases by embryonic day 21 (77). Thus, the Nip2 gene, a target for estrogenic activity, may be involved in the complex apoptotic-anti-apoptotic mechanism, and estrogens may interact to shift this balance. Further studies are needed to show specific localization of ER and Nip2 and the exact interaction of Nip2 with Bcl-2.

Oxidative Stress Neurotoxins, amyloid f-protein, hydrogen peroxide, glutamate, and NMDA

can

cause oxidative stresses, which induce cell death. These stresses have been linked to

LEE ® McEWEN

580

Ell Srl

+ WU | = OI-WU 00!

osurl JU

WH 0S-S1

€ll ell 6S 801

Ww QI-WU 001

WE OI

16 WwW 0I-WU 001

6S TIL 16

asuel wd

BE Ill

dad ‘ouy

é ON

SOX ON ON

jorpenso-J/ |

[Olpenso-g pur -nL|

AYLOIXOJOUIKXY

AWIIXOJOWOXA

oz1ds jorpesso-f L] [oIpenso-g pur-o,| — apnidad-¢ projAury

SSd1JS DADEPIXO

— Aqorxo} oyeureyn]H uolPRAudap wines AVLOIXO} OJBUIEINIH

aprxosadng AYLOIXOJOWOXA

jeduesoddiy Arewug

yeormsos ABUL

oreydaouasow Areullid

“OH

ON

[oOIpenso-¢ pur -v/|

orureeyjodAy yer APU

joipensa-g7]

ON

SOA

uoneAudap wines

prepsAwre yer ATPWLUg

jorpenso-J 1]

uoyeAudap winiasg

SOK

jorpensq

—s1weyeyjodAy asnow Arewg é

uoneaudep uniag

Oll

jeotio0o/eduresoddry oAtA uy

601

/Suolsay uleig

sadk] [PPO/MeuoINaNY

=SNayonu ayenose I[Npe—OAIA U]

jo adky,

7 ATAVL

aseuleg

dUON Pl0.1235

apndad-g projAury

10jd399y

= PAA OAT

Ty-[og ‘uy

«=—§._, UNJRNPoyY

€L

aAIIIJY

Z-[og ‘ouy uoneyusUu0D

OL ‘Py

udsoNsa Jo sjoayja aayoajoidomeN

“asraloap “daq ‘asraioul “Suy,

OWVCNEAS

veuny

projAury opuidad-g

AVIOTXO}OPOX|

AVLOTXO}OPOXY

AVIOTXO}OPIOXY

—-AJIOTX0} aJeureIN[H

HS-N-MS

Jorpense-g pue -0;]

‘On

opydad-g projAuy

ueuny

[orpeasa-g pur -0/]

Jorpense-g pue -0/,

“OCH

C1Od HA

Urwin} CUa-AS

QIVUILINIH AVLOTXO}

Ol W”

opiidad-¢ projxury

‘OCH

BUlol[s-9D dz

W7 OI W7 OL

GON GON

S20 CGLH

18a CLN

Ueuny cold

BULOISP[QOINSN]

6ll

We

611 6ll

C8

ON ON

[orpense-J/]

projAury “OCHepnded-¢ pur JoIpenso-¢ uimnias -20/ winiasg jorpenss-J// uoneAlidap |“< |uoneAudap

sil sil

C8

LL €9

1] jorpenso-

ON ON ON ON

OTT OHI

aa!

7-70 WU WU 7-70 WZW 100 |

Lil

SII

$8 98

“99q ZdIN [[99 yleop “99d

581 ACTIONS OF ESTROGENS

582

LEE

® McEWEN

diseases such as atherosclerosis and neurodegenerative disorders (both Alzheimer’s and Parkinson’s disease). They stimulate free radical accumulation induced by lipid peroxidation in neurons. As noted above, both 178 and 17a@ estradiol promote neu-

roprotection in these models via an antioxidant effect (78). In addition, glutamate causes a significant drop in glutathione levels, which also leads to an increase in intracellular peroxides and ultimately causes cell death (79). Estrogens, which possess intrinsic antioxidant activity, particularly in the 4M range, may serve as scavengers for free radicals (80). Estriol, 17@-estradiol, and 176-estradiol all have

significant antioxidant properties, with estriol (a weak estrogen) being the most potent antioxidant (81). The differences in antioxidant potency are due to minor variations in the phenolic structure (82), and the structure-activity profile as well as the 4M concentration range indicate that these effects are independent of the activation of the classical intracellular ERs.

Glutamate and NMDA Glutamate (range 1—3 44M) is required for optimal survival of normal hippocampal neurons in culture, but exposure to very high concentrations of glutamate (3-10 jzm) is toxic to neurons (83). Estrogen pretreatment produced an effective

neuroprotection from glutamate excitotoxicity, whereas other nonestrogenic steroids did not show protection (80, 84). Whether or not ER is present, many factors, such as neuronal cell types, nature of excitotoxic agent, and length of treatment, contribute to substantial differences in neuroprotective estrogen concentrations. The effective concentration of estrogen-mediated neuroprotection ranges from a low 0.1 nM (63,85, 86) to 10-50 nM (38, 84) to 50 uM (82, 87). At low levels of estrogen, it is likely that the direct genomic or indirect genomic cascades (see above) are involved, because antagonists of the intracellular ER block the effects, whereas at high levels of estrogen, the antioxidant-nongenomic pathway has been implicated (80, 88). In the primary hippocampal cultures, excitotoxic damage produced by NMDA (0.1—-1 zm) can be prevented by preincubation (24-48 h) with estrogens as determined by reduced cellular lactate dehydrogenase release and by an increased phosphorylation of CREB (pCREB) (38), which suggests that this neuroprotective effect may involve changes in transcriptional regulation of CREB.

B-AMYLOID PEPTIDES It is well known that accumulation of 6-amyloid protein is thought to be a causal factor in the development of cerebral plaques and tangles associated with the neuropathology of Alzheimer’s disease. Estrogen treatment is reported to inhibit the formation of toxic 8 amyloid and favor the production of the naturally secreted form of the amyloid precursor protein (89, 90). At the same time, estrogens exert protective effects against 6-amyloid toxicity in both primary and slice cultures of

ACTIONS OF ESTROGENS

583

the hippocampus (80, 91), as well as in the HT22 neuronal cell line that reportedly lacks functional ER (78), and in a neuroblastoma cell line (90). Consistent with the antioxidant effects described above, the protective effects of estrogen against B-amyloid toxicity show equal potency between 176 estradiol (active at the ER) and 17q@-estradiol (inactive at the intracellular ER) (88). Furthermore, the fact that estrogens protected HT22 neurons that lack functional ER suggests a non-ERmediated effect against B-amyloid toxicity.

Role of Astroglial Cells and Aromatase Astroglial cells may play an important role in neuroprotection, at least in part via their ability to express aromatase. This may be part of a response pattern in which glial cells are activated to produce specific growth factors, including the production of estradiol from testosterone to prevent further cell damage (92). Although estrogens increase the activity of brain aromatase by acting through the genome, it has been shown that aromatase activity decreases in the presence of

ATP, Ca**, and Mg?* in brain homogenates, a condition that promotes protein phosphorylation (93). This suggests that the local estrogen synthesis in the brain can be changed rapidly when needed and act in a paracrine fashion on neighboring cells (30, 94).

Role of ER in Neuroprotection and the Usefulness and Limitations of SERMs As noted above, not all neuroprotective effects of estrogens involve such high levels of estrogen or lack of stereospecificity for estradiol 178. Also noted above, a number of studies have reported that activation of ERK/MAPK by estradiol, possibly via an intracellular ER, mediates some forms of neuroprotection (58, 59, 74). These may be manifestations of the direct coupling of an intracellular ER to a second messenger system, as discussed earlier. Moreover, as noted above, estradiol

can also directly regulate gene expression at the nuclear level through binding of ER to ERE or to fos-jun heterodimers that bind to the AP-1 site. However, the bind-

ing of ERB to fos/jun heterodimers suppresses gene transcription from the AP-1 element, whereas nonsteroidal anti-estrogens such as tamoxifen have the opposite effect (40). Thus, tamoxifen blocks ERa and ER

actions via the ERE but acts as

an agonist when ERf is acting via the AP-1 site. Therefore, it is conceivable that anti-apoptotic effects of estradiol may be mediated by activation of anti-apoptotic genes through ERE or by the suppression of pro-apoptotic genes through the AP-1 enhancer element. Tamoxifen and raloxifene are examples of partial agonists for ER. These are also known as selective estrogen receptor modulators (SERM). SERMs have the advantage of providing beneficial estrogen-like effects on some tissues (bone, heart, and perhaps neurons) while acting as anti-estrogens in other tissue (breast and uterus). Furthermore, as noted above, SERMs have differential effects through

584

LEE ® McEWEN

ERa and ER and their interactions with ERE and AP-1 sites. The problem in the brain is that there are many actions of estrogens that are mediated via different intracellular mechanisms, some in which a SERM may be an agonist and others in which a SERM may be an antagonist (for review see 6). Agonistic actions of SERMs are known for the induction of choline acetyltransferase in the basal forebrain and hippocampus (95, 96), whereas antagonistic actions of SERMs are evident in blocking the estrogen induction of dendrite spines in the hippocampus (6). Moreover, a recent study shows somewhat different regional patterns of agonistlike effects of both tamoxifen and raloxifene in regulating SHT-2A receptors in cerebral cortex and striatum of ovariectomized, adult rats (97). In addition, SERMs appear to block estrogen effects through membraneassociated ER to the extent to which they can stimulate second messenger sys- ~ tems (45). Yet they are ineffective in blocking estrogen effects on calcium currents in striatal neurons (62) and they mimic estrogen in affecting kainate currents in hippocampal neurons (98, 99). Thus, the usefulness of SERMs as substitutes for estrogen replacement in postmenopausal women must be evaluated with great

care.

CONCLUSIONS The study of neurotrophic and neuroprotective actions of estrogens has emphasized the fact, evident from the investigation of other actions of estrogens in the nervous system, that estrogen effects in neural tissue are highly diverse and that some of these effects involve mechanisms of action that are not mediated by estrogen binding to a nuclear ER acting via the classical ERE. In particular, the possibility of second messenger pathways triggered by estrogens and the identification of estrogen receptors and aromatizing enzymes in some, but not all, nerve terminals, dendritic spines, and glial cell processes raise the possibility of anatomically specific intracellular signaling processes. Furthermore, the cross talk between estrogens and neurotrophins at the receptor and second messenger levels indicates that growth-promoting effects and neuroprotective actions have some degree of overlap in terms of receptors and initial pathways. Finally, these findings, together with the development of SERMs for mimicking or blocking intracellular ER, as well as the increasing evidence of antioxidant actions of estrogens, albeit at micromolar concentrations, make the study of estrogen actions in neural tissue an exciting and challenging topic and a target for pharmaceutical development. ACKNOWLEDGMENTS

The authors are indebted to those whose work is mentioned in this review and whose names appear in the references and to investigators who have contributed in this area but whose work may not have been referenced because of the strict space limitations. We would also like to acknowledge support from Grants NS07080-33

ACTIONS OF ESTROGENS

585

(National Science Foundation) and 1PO1AG16765 (National Institute on Aging)

for research from this laboratory that is described in this review. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED ile Allen E, Doisy E. 1924. The induction of a sexually mature condition in immature females by injection of the ovarian follicular hormone. Am. J. Physiol. 69:577—

88 . Jensen E, Jacobson H. 1962. Basic guides to the mechanism of estrogen action. Recent Prog. Horm. Res. 18:387-408 . Pfaff DW. 1980. Estrogens and Brain Function. New York: Springer-Verlag . Pfaff DW,

Keiner

M.

1973.

Atlas

of

estradiol-concentrating cells in the central nervous system of the female rat. J. Comp. Neurol. 151:121-58 . Stumpf W. 1971. Autoradiographic techniques and the localization of estrogen, androgen and glucocorticoid in the pituitary and brain. Am. Zool. 11:725-39 . McEwen BS, Alves SE. 1999. Estrogen actions in the central nervous system. Endocrine Rey. 20:279-307 . McEwen BS. 1983. Gonadal steroid influ-

ences on brain development and sexual differentiation. In Reproductive Physiology IV, ed. R Greep, pp. 99-145. Baltimore, MD: Univ. Park Press Naftolin F, . MacLusky NJ, Clark AS, Goldman-Rakic PC. 1987. Estrogen formation in the mammalian brain: possible role of aromatase in sexual differentiation of hippocampus and neocortex. Steroids 50:459-74 1991 Organotypic . Toran-Allerand CD. culture of the developing cerebral cortex and hypothalamus: relevance to sexual differentiation. _Psychoneuroendocrinology 16:7—24 10. Matsumoto A, Arai Y. 1981. Neuronal plasticity in the deafferented hypothalamic arcuate nucleus of adult female rats and its

enhancement by treatment with estrogen. J. Comp. Neurol. 197:197-205 . Gould

E,

Woolley

CS,

Frankfurt

M,

McEwen BS. 1990. Gonadal steroids regulate dendritic spine density in hippocampal pyramidal cells in adulthood. J. Neurosci. 10:1286-91

1D Woolley C, Gould E, Frankfurt M, McEwen BS. 1990. Naturally occurring fluctuation in dendritic spine density on adult hippocampal pyramidal neurons. J. Neurosci. 10:4035-39 Henderson VW. . Paganini-Hill A, 1994. Estrogen deficiency and risk of Alzheimer’s disease in women. Am. J. Epidemiol. 140:256-61 VW, Paganini-Hill A, . Henderson Emanuel CK, Dunn ME, Buckwalter JG.

1994. Estrogen replacement therapy in older women. Arch. Neurol. 51:896—900 . Tang MX,

Jacobs D, Stern Y, Marder K,

Schofield P, et al. 1996. Effect of estrogen during menopause on risk and age at onset of Alzheimer’s disease. Lancet 348:42932 DB, Shughrue PJ, Wilson ME, Merchenthaler I, Wise PM. 1999. Estra-

. Dubal

diol modulates bel-2 in cerebral ischemia: a potential role for estrogen receptors. J. Neurosci. 19:6385—93 . Hawk T, Zhang Y-Q, Rajakumar G, Day AL, Simpkins JW. 1998. Testosterone increases and estradiol decreases middle

cerebral artery occlusion lesion male rats. Brain Res. 796:296—98

size in

18. Kume-Kick J, Rice ME. 1998. Estrogendependent modulation of rat brain ascorbate levels and ischemia-induced ascorbate loss. Brain Res. 803:105—-13

iQ: ShiJ, Panickar KS, Yang S-H, Rabbani O,

586

LEE

® McEWEN

Day AL, Simpkins JW. 1998. Estrogen attenuates over-expression of B-amyloid precursor protein messenger RNA in an animal model of focal ischemia. Brain Res. 810:87—92 20. Naftolin F. 1994. Brain aromatization of androgens. J. Reprod. Med. 39:257-61 Die O’ Keefe JA, Handa RJ. 1990. Transient elevation of estrogen receptors in the neonatal hippocampus. Dev. Brain Res. 57:11927 ID Couse JF, Korach KS. 1999. Estrogen receptor null mice: What have we learned and where will they lead us? Endocr. Rev. 20(3):358-417 . Kuiper GGJM,

M,

Nilsson

Enmark E, Pelto-Huikko

S,

Gustafsson

JA.

1996.

Cloning of a novel estrogen receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. USA 93:5925-30

24. Shughrue PJ, Lane MV, Merchenthaler I. 1997. Comparative distribution of estrogen receptor a and 8 mRNA in the rat central nervous system. J. Comp. Neurol. 388:507—-25 23. Kuiper GGJM, Carlsson B, Grandien K, Enmark E, Haggblad J, et al. 1997. Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors. Endocrinology 138:863— 70 26. Shughrue PJ, Merchenthaler I. 2000. Evidence for novel estrogen binding sites in the hippocampus. Neuroscience 99(4):605—12 27. Loy R, Gerlach J, McEwen BS. 1988. Autoradiographic localization of estradiolbinding neurons in rat hippocampal formation and entorhinal cortex. Dev. Brain Res. 39:245-51 28. Green S, Walter P, Kumar V, Krust A, Bornet JM, et al. 1986. Human estrogen receptor CDNA: sequence, expression and homology to v-erb-A. Nature 320:134—-39 PS). Blaustein JD. 1992. Cytoplasmic estrogen receptors in the rat brain: immunochemical evidence using three antibodies with dis-

tinct epitopes. Enodocrinology 131:1336— 42 30. Naftolin F, Horvath TL, Jakab RL, Leranth C, Harada N, Balthazart J. 1996. Aromatase immunoreactivity in axon terminals of the vertebrate brain. An immunocytochemical study on quail, rat, monkey and human tissues. Neuroendocrinology 63:149-55 Suh Wagner CK, Silverman AJ, Morrell JI. 1998. Evidence for estrogen receptor in cell nuclei and axon terminal within the lateral habenula of rat: regulation during pregnancy. J. Comp. Neurol. 392:330-42 32. Milner T, McEwen BS, Hayashi S, Li CJ, Reagan LP, Alves SEA. 1999. Ultrastructural evidence that hippocampal alpha estrogen receptors are located at extranuclear sites. J. Comp. Neurol. 429:355-71

53° Clarke CH, Norfleet AM, Clarke MSF, Watson CS, Cunningham KA, Thomas ML. 2000. Perimembrane localization of the estrogen receptor @ protein in neuronal processes of cultured hippocampal neurons. Newroendocrinology 71:34-42 . Carrer H,

Aoki A. 1982. Ultrastructural

changes in the hypothalamic ventromedial nucleus of ovariectomized rats after estrogen treatment. Brain Res. 240:221—33 . Frankfurt M,

36.

Se

.

39.

McEwen

BS. 1991. Estro-

gen increases axodendritic synapses in the VMN of rats after ovariectomy. NeuroReport 2:380-82 Murphy DD, Segal M. 1996. Regulation of dendritic spine density in cultured rat hippocampal neurons by steroid hormones. J. Neurosci. 16:4059-68 Murphy DD, Cole NB, Segal M. 1998.. Brain-derived neurotrophic factor mediates estradiol-induced dendritic spine formation I hippocampal neurons. Proc. Natl. Acad. Sci. USA. 95:11415-17 Lee SJ, Dunlop JD, Li CJ, McEwen BS. 1999. Neuroprotective effects of estrogen following NMDA-induced neurotoxicity. Soc. Neurosci. Abstr. 25:581.1 Umayahara Y. 1994, Estrogen regulation

ACTIONS OF ESTROGENS of the insulin-like growth factor I gene transcription involves an AP-1 enhancer. J. Biol. Chem. 269:16433-43 40. Paech K, Webb P, Kuiper GGJM, Nilsson S, Gustafsson J-A, Kushner PJ, Scanlan TS. 1997. Differential ligand activation of estrogen receptors ERa and ERB at AP1 sites. Science 277:1508-10 41. Webb P, Nguyen P, Valentine C, Lopez GN, Kwok GR, et al. 1999. The estrogen receptor enhances AP-1 activity by two distinct mechanisms with different requirements for receptor transactivation functions. Mol. Endocrinol. 13(10):1672-85 42. Barnea A, Gorski J. 1970. Estrogen induced protein. Time course of synthesis. Biochemistry 9:1899-904 43. Kelly MJ, Wagner EJ. 1999. Estrogen modulation of G-protein-coupled receptors. Trends Endocrinol. Metab. 10:369-74 . Levin

ER.

1999.

Cellular

functions

of

the plasma membrane estrogen receptor. Trends Endocrinol. Metab. 10(9):374-77

45. Razandi M, Pedram A, Greene GL, Levin ER. 1999. Cell membrane and nuclear estrogen receptors (ERs) originate from a single transcript: studies of ERa and ERB expressed in Chinese hamster ovary cells. Mol. Endocrinol. 13:307-19 46. Migliaccio A, DiDomenico M, Castoria G, deFalco A, BontempoP, etal. 1996. Tyrosine kinase/p2i ras/MAP kinases pathway activation of estradiol-receptor complex in MCF-7 cells. EMBO J. 15:1292300 47. Watters JJ, Campbell JS, Cunningham MJ, Krebs EG, Dorsa DM. 1997. Rapid mem-

brane effects of steroids in neuroblastoma cells: effects of estradiol on mitogen activated protein kinase signaling cascade and c-fos immediate early gene transcription. Endocrinology 139:4030-33 48. Toran-Allerand CD, Singh M, Sétal6 G Jr. 1999. Novel mechanisms of estrogen action in the brain: new players in an old story. Front. Neuroendocrinol. 20:97-121 Klaus WL, Katzenel49. Aronica SM,

587

lenbogen BS. 1994. Estrogen action via cAMP signaling pathways: stimulation of adenylate cyclase, cAMP-regulated gene transcription. Proc. Natl. Acad. Sci. USA 91:8517-21 . Gu Q, Moss RL. 1996. 176-estradiol potentiates kainate-induced currents via activation of cAMP cascade. J. Neurosci. 16:3620-29 Si Murphy DD, Segal M. 1997. Morphological plasticity of dendritic spines in central neurons is mediated by activation of cAMP response element binding protein. Proc. Natl. Acad. Sci. USA. 94:1482— 87 23 Zakon HH. 1998. The effects of steroid hormones on electrical activity of excitable cells. Trends Neurosci. 21:202-7 53. Chiaia N, Foy M, Teyer TJ. 1983. The hamster hippocampal slice. II. Neuroendocrine modulation. Behav. Neurosci. 97:839-43 54. Garcia-Segura LM, Olmos G, Tranque P,

Naftolin

F

1987.

Rapid

effects

of

gonadal steroids upon hypothalamic neuronal membrane ultrastructure. J. Steroid. Biochem. 27:615—23 2p) Costello B, Meymandi A, Freeman J. 1990. Factors influencing GAP43 gene expression in PC12 pheochromoctoma cells. J. Neurosci. 10:1398—406 56. Finidori-Lepicard J, Schorderet-Slatkine S, Hanoune J, Baulieu E. 1981. Progesterone inhibits membrane bound adenylate cyclase in Xenopus laevis oocytes. Nature 292:255-57 Sih, Curtis J, Finkbeiner S. 1999. Sending signals from synapse to the nucleus: possible roles for CAMK, Ras/ERK and SAPK pathways in the regulation of synaptic plasticity and neuronal growth. J. Neurosci. Res. 58:88—95

58. Singh M, Sétal6 G Jr, Guan X, Warren M, Toran-Allerand CD. 1999. Estradiol in-

duced activation of mitogen activated protein kinase in cerebral cortical explants: convergence of estradiol and neurotrophin

588

LEE ® McEWEN

signaling pathways. J. Neurosci. 19:117988 59) Singer CA, Figueroa-Mascot XA, Batchelor RH,

Dorsa DM.

1999. The mitogen-

activated protein kinase pathway mediates estrogen neuroprotection after glutamate toxicity in primary cortical neurons. J. Neurosci. 196:2455—63 60. Moss RL, Gu Q. 1999. Estrogen: mechanisms for a rapid action in CA1 hippocampal neurons. Steroids 64:14—21 61. Gu Q, Korach KS, Moss RL. 1999. Rapid action of 17B-estradiol on kainate-induced currents in hippocampal neurons lacking intracellular estrogen receptors. Endocrinology 140:660—66 62. Mermelstein PG, Becker JB, Surmeier DJ. 1996. Estradiol reduces calcium currents in rat neostriatal neurons via a membrane receptor. J. Neurosci. 16:595—604 63. Bishop J, Simpkins JW. 1994. Estradiol treatment increases viability of glioma and neuroblastoma cells in vitro. Mol. Cell Neurosci. 5:303-8 64. Toran-Allerand CD. 1976. Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: implications for sexual differentiation. Brain Res, 106:407-—12 65. Bothwell M. 1996. p75N?®: a receptor after all. Science 272:506-7 66. Gibbs RB. 1998. Levels of trkA and BDNF mRNA, but not NGF mRNA, fluctuate across the estrous cycle and increase in response to acute hormone replacement. Brain Res. 787:259-68 67. Singh M, Meyer EM, Simpkins JW. 1995. The effect of ovariectomy and estradiol replacement on brain-derived neurotrophic factor messenger ribonucleic acid expression in cortical and hippocampal brain regions of female Sprague-Dawley rats. Endocrinology 136:2320 68. Sohrabji F, Miranda RC, Toran-Allerand CD. 1995. Identification of a putative estrogen response element in the gene encoding brain-derived neurotrophic factor.

Proc. Natl. Acad. Sci. USA. 92:1111014 WT, ManiS, Smith 69. O’ Malley B, Schrader C, Weigel NL, et al. 1995. An alternative ligand-independent pathway for activation of steroid receptors. Recent Prog. Horm. Res. 50:333-47 70. Woolley CS, McEwen BS. 1992. Estradiol mediated fluctuation in hippocampal synapse density during estrous cycle in the adult rat. J. Neurosci. 12:254954 ale Sandyk R. 1989. Estrogens and pathophysiology of Parkinson’s disease. Int. J. Neurosct. 45:119-22 W2: Grady D, Rubin SM, Petitti DB, Fox CS, Black D, et al. 1992. Hormone therapy to prevent disease and prolong life in post menopausal women. Am. Intern. Med. 117:1016—37 TBE Pike C. 1999. Estrogen modulates neuronal Bcl-xl expression and B-amyloid-induced apoptosis: relevance to Alzheimer’s disease. J. Neurochem. 72:1552-63 74. Sawada H, Ibi M, Kihara T, Urushitani M, Honda K, et al. 2000. Mechanisms of antiapoptotic effects of estrogens in nigral dopaminergic neurons. FASEB J. 14:120214 — nN . Gamier M, DiLorenzo D, Albertini A, Maggi A. 1997. Identification of estrogenresponsive gene in neuroblastoma SK-ER3 cells. J. Neurosci. 17:4591-99 76. Garcia-Segura LM, Cardona-Gomez P, Naftolin F, Chowen JA. 1998. Estradiol upregulates Bcl-2 expression in adult neurons. NeuroReport 9:593—97 is Brusadelli A, Sialino H, Piepoli T, Pol-lio G, Maggi A. 2000. Expression of the estrogen-regulated Nip2 during rat brain maturation. Int. J. Dev. Neurosci. 18:317— 20 ass Green PS, Gridley KE, Simpkins JW. 1998. Nuclear estrogen receptor independent neuroprotection by estratrienes: a novel interaction with glutathione. Neuroscience 84:7-10

ACTIONS OF ESTROGENS 49: Behl C. 1999. Vitamin E and other antioxidants in neuroprotection. Int. J. Vitam. Nutr. Res. 69(3):213-19

80. Behl C, Skutella T, Lezoulach F, Post A, Widemann M, et al. 1997. Neuroprotection against oxidative stress by estrogens: structure-activity relationship. Mol. Pharmacol. 51:535-41 81. Mooradian AD. 1993. Antioxidant properties of steroids. Steroid Biochem. Mol. Biol. 45(6):5S09-11 82. Moosmann B, Behl C. 1999. The antioxidant neuroprotective effects of estrogens and phenolic compounds are independent from their estrogenic properties. Proc. Natl. Acad. Sci. USA 96:8867— 72

589

cursor protein. J. Biol. Chem. 269:13065— 68

90. Xu H, Gouras GK,

Greenfield JP, Vincent B, Naslund J, et al. 1998. Estrogen

reduces neuronal generation of Alzheimer B-amyloid peptides. Nat. Med. 4:447-51

il. Goodman Y, Bruce AJ, Cheng B, Mattson MP. 1996. Estrogens attenuate and corticosterone exacerbates excitotoxicity, oxidative injury, and amyloid B-peptide toxicity in hippocampal neurons. J. Neurochem. 66:1836—-44 OD Garcia-Segura LM, Naftolin F, Hutchinson JB, AzcoitiaI, Chowen JA. 1999. Role of astroglia in estrogen regulation of synaptic plasticity and brain repair. J. Neurobiol. 40(4):574-84

83. Bambrick LL, Yarowsky PJ, Krueger BK.

. Balthazart J, Ball GF. 1998. New insights

1995. Glutamate as a hippocampal neuron survival factor: an inherited defect in the trisomy 16 mouse. Proc. Natl. Acad. Sci. USA 92:9692-96 84. Singer CA, Rogers KL, Stickland TM, Dorsa DM. 1996. Estrogen protects primary cortical neurons from glutamate toxicity. Neurosci. Lett. 212:13-16 Bishop J, Simpkins JW. 85. Green S, 1997. 17B-estradiol exerts neuroprotective effects on SK-N-SH cell. J. Neurosci. 17:511-15 86. Green PS, Gridley KE, Simpkins JW. 1996. Estradiol protects against B-amyloid (25-25)-induced toxicity in SK-N-SH human neuroblastoma cells. Neurosci. Lett. 218:165-68

into the regulation and function of brain estrogen synthase. Trends Neurosci. 21:243— 48 94. Schlinger BA, Callard GV. 1989. Localization of aromatase in synaptosomal and microsomal subfraction of quail brain. Neuroendocrinology 49:434-41 Oy. Luine V, McEwen BS. 1977. Effects of an estrogen antagonist on enzyme activities and 7H estradiol nuclear binding in uterus, pituitary and brain. Endocrinology 100:903-10 96. Wu X, Glinn MA, Ostrowski NL, Su Y,

87. Weaver CE TT,

Farb DH.

Jr, Park-Chung M, Gibbs 1997. 17-f-estradiol pro-

tects against NMDA-induced excitotoxicity by direct inhibition of NMDA receptors. Brain Res. 761:338-41 88. Green PS, Simpkins JW. 2000. Neuroprotective effects of estrogens: potential mechanism of action. Jnt. J. Dev. Neurosci. 18:347-58 89. Jaffe AB, Toran-Allerand CD, Greengard P, Gandy SE. 1994. Estrogen regulates metabolism of Alzheimer amyloid £ pre-

Ni B,

et al. 1999. Raloxifene and estra-

diol benzoate both fully restore hippocampal choline acetyltransferase activity in ovariectomized rats. Brain Res. 847:98— 104 Oy. Cyr M, Landry M, DiPaolo T. 2000. Modulation by estrogen receptor directed drugs of 5-hydroxytryptamine-2A receptors in rat brain. Biol. Psychiatry 23:69-78 98. Wong M, Moss RL. 1991. Electrophysiological evidence for a rapid membrane action of the gonadal steroid, 17-estradiol,

on CAI pyramidal neurons of the rat hippocampus. Brain Res. 543:148-52 OY). Wong M, Moss RL. 1992. Long-term and short-term electrophysiological effects

590

LEE ® McEWEN

of estrogen on the synaptic properties of hippocampal CA1 neurons. J. Neurosci. 12:3217-25

100. Garcia-Segura LM, Duenas M, Fernandez-Galaz MC, Chowen JA, Argent J, et al. 1996. Interaction of signaling pathways of insulin-like growth factor I and sex steroids in the neuroendocrine hypothalamus. Horm. Res. 46:160— 64 101. Ferreira A, Caceres A. 1991. Estrogenenhanced neurite growth: evidence for selective induction of Tau and _stable microtubules. J. Neurosci. 11:392-

400

102. Lorenzo A, Diaz H, Carrer H, Caceres A. 1992. Amygdala neurons in vitro: neurite growth and effects of estradiol. J. Neurosci. Res. 33:418-35 103. Zhang L, Chang YH, Barker J, Hu Q, Zhang L, et al. 2000. Testosterone and es-

104.

105.

106.

107.

trogen affect neuronal differentiation but no proliferation in early embryonic cortex of the rat: the possible roles of androgen and estrogen receptors. Neurosci. Lett. 281:57-60 Patrone C, Ma ZQ, Pollio G, Agrati P, Parker MG, Maggi A. 1996. A crosscoupling between insulin and estrogen receptor in human neuroblastoma cells. Mol. Endocrinol. 10:499-S07 Dittrich F, Feng Y, Metzdorf R, Gahr M. 1999. Estrogen-inducible sex specific expression of brain derived neurotrophic factor mRNA in a forebrain song control nucleus of juvenile zebra finch. Proc. Natl. Acad. Sci. USA 96:824146 Beyer C, Raab H. 1998. Non-genomic effects of estrogen: embryonic mouse midbrain respond with a rapid release of calcium from intracellular stores. Eur J. Neursoci. 10:255—62 Behl C. 1999. Alzheimer’s disease and oxidative stress: implications for novel therapeutic approaches. Prog. Neurobiol. 57:301-23

108. Diaz-Brinton R, Tran J, Proffitt P, Montoya M. 1997. 17f-estradiol enhances the growth and survival of neocortical neurons in culture. Neurochem. Res. 22:1339-51 A, Rosenbaum E, 109. Faivre-Bauman Puymirat J, Grouselle D, Tixier-Vidal A. 1981. Differentiation of fetal mouse hypothalamic cells in serum-free medium. Dev. Neurosci. 4:118—29 110. Arimatsu Y, Hatanaka H. 1986. Estrogen treatment enhances survival of cultured fetal rat amygdala neurons in a defined medium. Dev. Brain Res. 26:151a9 11h. Chowen JA, Torres-Aleman I, GarciaSegura LM. 1992. Trophic effects of estradiol on fetal rat hypothalamic neurons. Neuroendocrinology 56:895— 901 . Brook S, Chan R, Howard S, Sapolsky R. 1997. Endocrine modulation of the neurotoxicity of gp120: implications for AIDS related dementia complex. Proc. Natl. Acad. Sci. USA 94:9457-62 i lse Sawada H, Ibi M, Kihara T, Urushitani M, Honda K, et al. 1998. Estradiol protects mesencephalic dopaminergic neurons from oxidative stress-induced neuronal death. J. Neurosci. 54:70719 114. Shy H, Malaiyandi L, Timiras PS. 2000. Protective action of 17f-estradiol and tamoxifen on glutamate toxicity in glial cells. Int. J. Dev. Neurosci. 18:289— 97 Lis. Mook-Jung I, Joo I, Sohn S, Kwon HJ, Huh K, Jung MW. 1997. Estrogen blocks neurotoxic effects of beta-amyloid (1-42) and induces neurite extension of B103

cells.

Neurosci.

Lett.

235:101—

4 116. Bonnefont AB, Munoz FJ, Inestrosa NC. 1998. Estrogen protects neuronal cells from the cytotoxicity induced by acetyl-cholinesterase-amyloid complexes. FEBS Lett. 411:220-24

ACTIONS OF ESTROGENS ie

Mattson MP, Robinson N, Guo Q. 1997. Estrogens stabilize mitochondrial function and protect neural cells against the pro-apoptotic action of mutant presenilin-1. NeuroReport 8:3812-17 118. Singer CA, Rogers KL, Dorsa DM. 1998. Modulation of bel-2 expression: a poten-

I).

591

tial component of estrogen protection in NT2 neurons. NeuroReport 9:2565-68 Behl C, Widemann M, Trapp T, Holsboer F. 1998. 17b-estradiol protects neurons from oxidative stress-induced cell death in vitro. Biochem. Biophys. Res. Comm. 216:473-82

Annu. Rey. Pharmacol. Toxicol. 2001. 41:593-624 Copyright © 2001 by Annual Reviews. All rights reserved

GENETIC VARIATIONS AND POLYMORPHISMS OF G PROTEIN-COUPLED RECEPTORS: Functional

and Therapeutic Implications Brinda K Rana, Tetsuo Shiina, and Paul A Insel Department of Pharmacology, University of California at San Diego, La Jolla, California 92093-0636; e-mail: [email protected], [email protected], [email protected]

Key Words

SNPs, variants, mutants, structure-function, complex diseases

@ Abstract G protein-coupled receptors (GPCRs) represent a major class of proteins in the genome of many species, including humans. In addition to the mapping of a number of human disorders to regions of the genome containing GPCRs, a growing body of literature has documented frequently occurring variations (i.e. polymorphisms) in GPCR loci. In this article, we use a domain-based approach to systematically examine examples of genetic variation in the coding and noncoding regions of GPCR loci. Data to date indicate that residues in GPCRs are involved in ligand binding and coupling to G proteins and that regulation can be altered by polymorphisms. Studies of GPCR polymorphisms have also uncovered the functional importance of residues not previously implicated from other approaches that are involved in the function of GPCRs. We predict that studies of GPCR polymorphisms will have a significant impact on medicine and pharmacology, in particular, by providing new means to subclassify patients in terms of both diagnosis and treatment.

INTRODUCTION The heterotrimeric GTP-binding protein (G protein)-coupled receptors (GPCRs) are members of a large family of proteins found in eukaryotes and certain prokaryotes. The GPCR family is the third most abundant family in Caenorhabditis elegans, comprising 5% of its genome with approximately 1100 members (1). The Drosophila genome has at least 160 GPCRs whereas it is estimated that there are at least 700 GPCRs in the human genome (2). These plasma membrane-bound receptors have evolved to recognize a diversity of extracellular physical and chemical signals, such as nucleotides, peptides, amines, Ca**, and photons. On recognition of such signals, the GPCRs act via one or more heterotrimeric G proteins to alter

the level of intracellular messengers as a proximal event in signaling pathways that influence a wide variety of metabolic and differentiated functions. Because of

0362-1642/01/0421-0593$14.00

593

594

RANA

® SHIINA

® INSEL

their ubiquitous expression and multiple actions, it is not surprising that variation at GPCR loci can be associated with disease states. In this article, we review recent observations regarding genetic polymorphisms in human GPCRs, in particular therapeutic implications of such polymorphisms. The recent progress in identifying GPCRs, sequencing the genomes of humans and other species, mutational analyses of GPCRs, and discovery of coding region and single nucleotide polymorphisms (SNPs) at human GPCR loci make attempts to review this field a daunting task, at best akin to trying to take a snapshot of a rapidly moving train of information. We have terminated our literature search as of March 2000; information is evolving so quickly as to supercede data available then. We have recently posted on our laboratory web site (www-insel.ucsd.edu) the addresses of useful GPCR-related web sites/databases. It is our intention to update this list with links to other sites containing human GPCR polymorphism data, some of which are still under development. Our emphasis is on naturally occuring GPCR variants and their possible impact on drug therapy. Thus, we have organized the review to focus on structural domains of GPCRs and their role in signal transduction, discussing GPCR mutations that affect cellular response to hormones, neurotransmitters, and other pharmacological agents. We have highlighted data reported for rhodopsin, thyrotropin stimulating hormone (TSH) receptors, and vasopressin 2 (V,) receptors, in part because of the large number of mutations found in their structural domains. Studies on those mutations allow one the opportunity to draw inferences regarding structure-function relationships of GPCRs and, in some cases, on response to agonists and antagonists. Because our principal emphasis is on the therapeutic impact of genetic variation at GPCR loci, we do not discuss structure-function relationships in detail, but interested readers may wish to consult other recent reviews that emphasize

such relationships (e.g. 3-9).

GPCR Classification GPCRs have been divided into three main classes. These contain the known human GPCRs. Two additional classes encompass fungal pheromone and Dictyostelium GPCRs. Each of these classes has been further divided into subclasses based on a combination of sequence similarity, functional domains, and ligand binding properties (10-11). Though comparison of GPCRs from different classes: results in little or no sequence similarity, all GPCRs share a common overall structure (Figure la—c; see color insert). Hydropathy analysis predicts that GPCRs have seven hydrophobic transmembrane (TM) helices connected by three intracellular loops (IL) and three extracellular loops (EL). GPCRs also contain an N-terminal extracellular region with N-glycosylation sites and a C-terminal intracellular domain, generally with sites for phosphorylation. Each of these domains has distinct properties depending on the class and subclass of receptors; these properties confer on individual receptors their ligand binding and G protein specificity.

GPCR POLYMORPHISMS

595

According to their class, GPCRs share additional structural and functional properties, as reviewed by Bockaert & Pin (11), Schoneberg et al (8), and Wess (3), and summarized in Figure | (see color insert). Class A contains most known mem-

bers of the GPCR family and includes three subclasses. Sequence alignment of receptors reveals approximately 20 amino acid residues conserved by most class A GPCRs, which include the following: two Cys residues in EL1 and EL2 that form a disulfide bridge necessary for maintaining proper receptor conformation; the AspArg-Tyr sequence (DRY motif) in the proximal region of IL2; proline residues in the TM regions; an Asn-Pro-X-X-Tyr motif in TM7; and a Cys residue in the C-terminal domain, which can be palmitoylated, thereby forming a fourth IL (Figure 1a). Within class A GPCRs are receptors for photons of light (opsins and rhodopsin), biogenic amines, and other small physiological ligands. The binding site for these molecules is located within the seven TM helices. A second subclass within class A GPCRs includes receptors that respond to chemokines, thrombin, and other small peptides. These ligands bind to the N terminus, the ELs, and regions of the TMs that are close to the ELs. Receptors of the third subclass of class A GPCRs have a relatively large N-terminal domain and bind glycoprotein hormones, such as luteinizing hormone (LH), follicle-stimulating hormone (FSH), and TSH. The class

B GPCRs,

the second

largest class, contains

receptors that bind

to higher-molecular-weight hormones, such as glucagon and calcitonin. Class B GPCRs are characterized by long N-terminal regions (>100 residues) involved in ligand binding, six well-conserved Cys residues in the N-terminal region, two wellconserved Cys residues in EL1 and EL2, and approximately 15 other residues that are identical in all members of this class of GPCRs (12) (Figure 1b). Finally, class C, the smallest class of GPCRs, contains the metabotropic glutamate receptors, the GABAsg receptor, and the calcium sensing receptor (CaR). These receptors possess extremely long extracellular N-terminal domains that are involved in ligand binding and contain several conserved Cys residues in the TM and extracellular regions (Figure Ic).

Variation at GPCR Loci As readily accessible, plasma membrane-bound molecular entities that regulate a wide variety of physiological and metabolic processes, GPCRs are commonly used as therapeutic drug targets. To date, numerous studies have reported genetic variation at GPCR loci and their role in human disease and traits (Tables 1-4). A

genetic “polymorphism” is defined here as one of at least two commonly occurring (i.e. at least 1% of the population) genetic variants at a locus whereas the term mutation is used for a genetic variant that occurs in an isolated individual or pedigree for germline mutations and for less frequently occurring somatic mutations identified in isolated tissues. Several hundred mutations have been reported at GPCR loci, most within the coding and 5’ untranslated regions. In rhodopsin alone, 63 variants of the receptor have been identified in families with autosomal dominant forms of retinitis

596

RANA

® SHIINA

® INSEL

pigmentosa (RP) (13). At least 36 of the 70 reported variants of the V, are believed to cause X-linked nephrogenic diabetes insipidus (NDI), a with renal resistance to vasopressin (14). Studies on variation in GPCRs been limited to germline mutations. In addition to germline variants of

receptor disorder have not the TSH

receptor identified in individuals with, for example, congenital nonautoimmune

hyperthyroidism, a large number of somatic variants of the TSH receptor have been identified in toxic adenomas and nodules (15—16). From these examples it is

obvious that both somatic and germline mutations at GPCR loci have the potential to contribute to variation in GPCR signaling and function. In addition to the large number of GPCR variants associated with monogenic diseases, there is growing evidence for GPCR polymorphisms in complex human diseases and traits. For example, two major polymorphisms of the 65-adrenergic receptor (B,-AR) have been examined for their association with asthma (17) and hypertension (18). Such GPCR polymorphisms contribute to a complex phenotype in combination with polymorphisms at other genetic loci as well as with environmental factors. The majority of naturally occurring variants at GPCR loci have been reported only within the past several years. In parallel with these reports, pharmacological studies have primarily involved mutational analyses designed to assess the role of domains in GPCR function. Most of these latter studies have been conducted by using GPCR variants generated through site-directed mutagenesis. Thus, the role of the naturally occurring variants of most GPCR variants in human disease and traits still needs to be elucidated. In the remainder of this review, we discuss GPCR polymorphisms by taking a “sequential domain approach.” We begin at the amino terminus, citing mutagenesis studies that elucidate roles of this domain on signal transduction, followed by studies identifying naturally occurring genetic variants. Because the number of these variants is very large, we limit our examples to those with functional consequences, associated with disease, or possible therapeutic impact. We continue along the receptor peptide to discuss the TM domains, the intracellular loop domains, and the C-terminal domain. In addition, many pharmacologically altered natural GPCR variants have been reported as a consequence of variation in noncoding regions, expression mechanisms, or dimerization. Thus, we conclude with sections on such variants.

THE N-TERMINAL DOMAIN The N terminus of GPCRs is an extracellular domain of variable size, ranging from 154 residues in the calcitonin receptor to 36 residues in the rhodopsin receptor. This domain has several features important for GPCR function, including asparagine residues and motifs for N-glycosylation, which influence intracellular trafficking of receptors to the plasma membrane (19), and cysteine residues that can influence

protein folding critical for trafficking of a functional receptor to the cell surface

GPCR POLYMORPHISMS TABLE 1 a

Selected genetic variants of the N-terminal domain® ee ee a Functional consequence

ee

ee

Receptor

Variant

(clinical consequence)

Reference

B>-AR

Argl6Gly

Enhanced agonist-promoted down-regulation Blunted agonist-promoted down-regulation

20

Gln27Glu

597

20

5-HT\,

Gly22Ser

Attenuated receptor downregulation and desensitization

30

5-HT5¢

Cys23Ser

Decreased agonist binding affinity

32

j4-Opioid

Asn40Asp

Increased agonist binding affinity

36

Rhodopsin

Thr4Lys Asn1I5Ser

Affects N-glycosylation (RP) Alters N-glycosylation site (RP)

SY) 38

*RP, Retinitis pigmentosa.

(20). The N terminus of some GPCRs also contains residues involved in ligand binding, activation, and down-regulation. A variety of disease-associated variants have been identified in this domain

(Table

1). For example, in rhodopsin, six

variants have been identified in patients with a dominant form of RP: Thr4Lys, Asnl5Ser, Thr17Met, Pro23His, Pro23Leu, and Gln28His. Variants in this region

that have functional consequences but lack clear association with disease are also common.

Receptor Down-Regulation B,-ARs are expressed on a wide variety of cell types, including smooth muscle cells, in which the receptors promote relaxation in response to catecholamines and synthetic agonists. Two main polymorphisms have been identified at the B,-AR loci, and they code for Arg/6Gly and the Gln27Glu variants. Green et al (20) expressed these variants in a heterologous cell system and showed that the variants did not display altered agonist affinity or coupling to G, but instead had altered agonistmediated regulation. The extent of agonist-promoted down-regulation of receptor number, which required several hours of incubation with agonist, was enhanced

for the Arg/6Gly variant but was blunted for the Gln27Glu variant. When 16Gly and 27Glu were present together, down-regulation resembled that of the Arg/6Gly variant. Thus, the change at codon16 to Gly dominates over that at codon 27. Green et al (20) suggested that the difference in down-regulation in these variants is due to altered receptor degradation after internalization. Turki et al (21) reported that the Arg/6Gly variant occurs more commonly in individuals with nocturnaltype asthma. A potentially important pharmacological implication of the Arg/6Gly

598

RANA

® SHIINA

® INSEL

variant is that children homozygous for Gly/6, compared with those homozygous for Arg16, appear less likely to show bronchodilation in response to the Badrenergic agonist albuterol (22), perhaps because of chronic down-regulation of B>-AR in response to endogenous catecholamines. Some studies have confirmed the greater extent of Arg/6 homozygotes for bronchodilation in response to B>-AR agonist (23) whereas others have not observed this (e.g. 24, 25). Although larger numbers of patients need to be examined, consideration of the codon16 polymorphism may prove useful for future therapeutic strategies for asthma that target B>-ARs. Because of the role of 6,-AR in vascular smooth muscle relaxation, the Arg/6Gly and Gln27Glu variants of 6,-AR have been examined for association with essential hypertension. Some studies suggest that either the Gly/6 (26-27) or the Arg16 allele (28) is more frequent in individuals affected with or prone to develop hypertension, but we and our colleagues fail to find such an association (29). As with most complex diseases, the precise role of a variant at an individual locus is difficult to assess, and further studies on different populations will be necessary. Another naturally occurring N-terminal variant that has been implicated in affecting agonist-induced down-regulation is the Gi/y22Ser variant of the hS-HT,, receptor. This variant displays attenuated agonist-promoted down-regulation and functional desensitization with no effect on ligand binding profiles when compared with the wild-type receptor (30). Because the effect of selective serotonin reuptake inhibitors used to treat psychiatric disorders is partially dependent on h5-HT,, receptor down-regulation, it is conceivable, but not yet well tested, that individuals carrying such variants may have altered sensitivity to antidepressant treatment.

Ligand Binding Residues of the N terminus of certain GPCRs are also involved in ligand binding. Okazaki et al (31) found a Lys47Asp variant of the calcium-sensing receptor (CaR) in an individual with an autosomal dominant form of hypocalcemia and implicated this codon as important for calcium sensing. Lappalainen et al (32) reported that a Cys23Ser variant of the 5-HT,¢ receptor, found at a frequency of 13% in unrelated | Caucasian individuals, exhibited lower binding affinity than did wild-type receptor for agonist. This change in binding affinity to 5-HT receptor agonist binding led to studies on association of response to clozapine, which is used for the treatment of schizophrenia. A study of 168 schizophrenic patients found that individuals with Cys at codon 23 responded better to clozapine than did individuals with Ser at this position (33). Further work will be needed to confirm this observation. In-

terindividual variation in clozapine response has also been associated with variants of another 5-HT receptor (discussed below).

Receptor Trafficking Glycosylation and oligosaccharide processing are important for protein trafficking of adrenergic receptors to the plasma membrane (34, 35). The importance of

RANA

® SHIINA

# INSEL C-1

(a)

(b)

(Cc)

“=8Y COOH

Figure 1 Diagram of main structural features of the three classes of GPCRs. (a) Class A GPCRs share approximately 20 amino acid residues in common, including two Cys residues (in green) which form a disulfide bridge between EL1 and EL2, a Asp-Argtyr (DRY) motif in IL2, a Asn-Pro-X-X-Tyr (NPXXY) motif (in magenta) in TMVII, and a Cys residue (in orange) that serves as a site for palmitoylation in the C-terminal domain. (/) Class B GPCRs share six conserved Cys residues (in blue) in their N-terminal domain, two highly conserved Cys residues (in green) in the EL] and EL2 domains, and several conserved amino acid residues (indicated by their one letter code). (c) Class C GPCRs are characterized by their long N-terminal domain and conserved Cys residues in the extracellular and TM regions.

GPCR POLYMORPHISMS

599

N-glycosylation has also been shown for other GPCRs. For example, Petaja-Repo et al (19) studied human 6$-opioid receptors in a heterologous expression system and found that N-glycosylation of N-terminal asparagine residues within a defined glycosylation consensus motif is initiated cotranslationally in the endoplasmic reticulum (ER) and completed in the trans-Golgi network. These fully processed receptors are then transported to the cell surface in 10 min. The authors suggested that the rate-limiting step in this process is the exit of a fully processed GPCR from the ER. Less than 50% of synthesized receptors are fully processed, of which only a fraction attain the proper conformation to exit the ER. Thus, folding and export that rely partially on glycosylation events are potentially key events in the control of GPCR expression and activity. Several polymorphisms have been identified that alter N-glycosylation in GPCRs. The Asn40Asp variant of the -opioid receptor was found in 10% of a study population of 113 heroin addicts in a methadone maintenance program and 39 controls with no history of drug or alcohol abuse (36). When the population was stratified for ethnicity, this variant was significantly more frequent in Hispanic non—opioid-dependent subjects than in other ethnic groups studied. This variant also displays a higher binding affinity than does wild-type receptor for the endogenous beta-endorphin. A point mutation rhodopsin at codon 15, which changes an Asn residue to Ser, altering a glycosylation site, was described in a five-generation Australian family with RP (37). Another rhodopsin variant, Thr4Lys, which affects glycosylation of Asn2 and probably causes rhodopsin not to be incorporated into the membrane, was found in patients with a different form of dominant RP (38).

The examples given here indicate that N-terminal variants affect a variety of functions of GPCRs and have given insight into roles of N-terminal domain residues in GPCR signal transduction. It is worth noting that several of the observed polymorphisms leading to functional consequences, such as agonist-mediated down-regulation, were not expected from prior studies of the N-terminal region. We expect that future studies of polymorphic GPCRs will reveal additional roles for N-terminal amino acid residues.

TRANSMEMBRANE DOMAINS The transmembrane (TM) domains of GPCRs

are comprised of seven a-helices

imbedded in the lipid bilayer of the plasma membrane. The seven helices are thought to be arranged as a tight, ring-shaped core (5). Similar to most TM proteins, the hydrophobic amino acid residues are presumably arranged to face the lipid bilayer, whereas the more hydrophilic amino acid residues face the core. Furthermore, helix-helix interactions contribute to the functional tertiary structure of the GPCR necessary for receptor folding and stability, ligand binding, and ligandinduced conformational changes for G protein coupling. Thus, mutations in the TM domain can have an array of deleterious effects (Table 2). For example, 11

naturally occurring mutations in the TM domains of rhodopsin of RP patients have been shown to interfere with folding or stability of the receptor (39).

600

RANA

TABLE 2

® SHIINA

® INSEL

Selected genetic variants of the TM domains*

ES

Receptor

Variant

Functional consequence (clinical consequence)

Reference

D4

Vall94Gly

Decreased agonist binding

ae

ET,

Cysl09Arg

Disruption of putative signal sequence (Hirschsprung’s disease) Altered receptor coupling to Gy

105 47

Trp276Cys FSH

Phe591Ser

Deficient in activating adenylyl cyclase via G,

45

V>

Lys44Pro

Retention in pre-Golgi compartment (NDI) Defective processing and retention in ER (NDD Lacks cell surface expression; retention in Golgi compartment Abolished agonist binding

14

Trp164Ser Serl67Lys AV278/279

Tyr128Ser

14 41 14

aTM, transmembrane; D4, dopamine 4; ETg, endothelin B; FSH, follicle-stimulating hormone; V>, vasopressin 2; NDI, nephrogenic diabetes insipidus; ER, endoplasmic reticulum.

Mutations that alter receptor folding can lead to defective posttranslational modification, intracellular transport, or cell surface expression. For example, the TM domains contain several highly conserved Pro residues important for receptor folding and expression. Wess et al (40) demonstrated the significance of such Pro residues in a site-directed mutagenesis study in which three conserved Pro residues (Pro242 of TMV, ProS05 of TMVI, and Pro540 of TMVII) of the m3 muscarinic receptor were changed to Ala. This resulted in a 35- to 100-fold lower expression level than in the wild-type receptor. A change to a Pro residue in the TM domain can also alter receptor expression. The V, receptor Lys44Pro mutant of TMI, which is found in patients with X-linked NDI, lacks glycosylation, indicating retention in the pre-Golgi compartment. The introduction of Pro is key to receptor folding and subsequent processing, because substitution of another residue, such as the Lys44Phe variant, shows correct processing of the receptor, although it also has defective binding. Other mutants of the TM domain of the V, receptor in NDI patients are the Trp164Ser, Serl67Lys, and Serl67Thr in TMIV; all three produce changes in amino acids conserved among GPCRs and show defective processing and retention in the ER, as indicated by their lack of glycosylation (14). In contrast to these mutants, the V, receptor mutant with a frame deletion in TMVI, AV278 /279, also

found in X-linked NDI patients, shows processing similar to the wild-type receptor but lacks cell surface expression because of retention within the Golgi or postGolgi compartment (41). These studies on V, receptor TM domain mutants have useful therapeutic implications. The defective folding, processing, and cell surface

GPCR POLYMORPHISMS

601

expression of mutant V, receptors that still retain some residual receptor function (ligand binding, AC coupling, and G protein stimulation) probably explains why patients with such mutations fail to respond to standard treatment for X-linked NDI. In these patients, even high doses of the synthetic V receptor agonist desmopressin fail to restore the defective renal tubular concentrating ability observed in NDI

(14). Another naturally occurring variant of the TM domain has been shown to affect translocation of the receptor into the plasma membrane. The Cys/09Arg mutant of the proximal region of TM1 of the endothelin B receptor was identified in a patient with Hirschsprung’s disease (42), a congenital disorder characterized by the absence of ganglion cells in the distal portion of the intestinal tract caused by the premature arrest of neural crest cell migration. This mutation replaces a hydrophobic residue with a charged residue, thereby disrupting the hydrophobic stretch of a putative signal sequence involved in the transport, folding, and proper targeting of the mature receptor from the ER to the plasma membrane. Thus, therapeutic strategies for this disorder (and probably for other disorders) need to consider the possibility of defects in processing, transport, and cell surface expression. Residues of the TM domain are also involved in ligand binding. Many studies have used site-directed mutagenesis to show altered ligand binding of TM domain mutants. Thus, for example, an extensive mutagenesis study of the TM helices of the melanocortin 1 receptor (MC1R) revealed the importance of several acidic residues in TMII and -III for ligand binding and of aromatic residues in TMIV, -V, and -VI in the formation of a hydrophobic pocket for ligand binding (43). Naturally occurring mutations in the TM domain have also been identified that can alter ligand binding affinity. For example, in the dopamine 4 (D4) receptor, the Vall94Gly variant of TMIV changes Val residue one position away from the Ser residue critical to dopamine binding. This variant has been identified at a frequency of about 12% in the Afro-Caribbean population and is reported as twofold less sensitive to dopamine, clozapine, and olanzapine than is the wildtype receptor (44). No definitive studies have shown an association of this variant with schizophrenia or other disease. However, TM domain variants of V, receptor have been identified as a cause of disease. The Arg//37rp mutant of TMIII of the V, receptor displays a 20-fold increase in Kp for the agonist Arg-vasopressin whereas agonist binding in the 7yr/28Ser mutants of TMIII is completely abolished (14). Both mutations are found in individuals with X-linked NDI. Such results suggest that therapeutic approaches for the treatment of NDI, and perhaps other diseases,

will need to consider the binding properties of each affected individual’s mutant receptor. Analysis of naturally occurring variants has also implicated the TM domain in G protein coupling. Variants of TMVI in two members of the glycoprotein hormone receptor subfamily of GPCRs, FSH receptor and LH receptor, result in receptors deficient in the ability to activate adenylyl cyclase via G,. The FSH receptor mutant, Phe591Ser, eliminates cAMP production when expressed in COS cells. A similar

602

RANA

® SHIINA

® INSEL

effect is seen when this mutation is introduced in a homologous position in the LH receptor (45); this mutant receptor is able to bind human chorionic gonadotropin with an affinity similar to that of the wild-type receptor but is defective in its ability to activate adenylyl cyclase (46). A variant of the endothelin B receptor, in which a highly conserved Trp residue at position 276 in TMV is substituted with a Cys, impairs ligand-mediated increases in Ca** levels in transfected cells, most likely by altering receptor coupling to G, (47). This variant was identified in a Mennonite pedigree and was found to associate with Hirschsprung’s disease such that homozygotes and heterozygotes for the variant had, respectively, a 74% and 21% risk for the disease.

Dimerization Several studies provide evidence that GPCRs can exist as homodimers or heterodimers, and receptors in this state have altered receptor function (e.g. 48, 49). Given the potentially important role of homodimerization in the function of GPCRs, identifying key domains for dimerization is important for assessing the role of polymorphisms and for developing new therapeutic agents. Hebert et al (50) showed that agonists favored 6,-AR homodimerization. Furthermore, using evidence from a sequence in the glycophorin A protein that is essential for homodimerization, [LIXXGVXXGVXXT], Hebert et al proposed that a sequence in TM6 of §,-AR, [LKTLGHMGTFTL], is responsible for 8,-AR dimerization. Consistent with this proposal, they found that addition of this peptide competed for intact GPCRs and reduced homodimerization. The functional consequences of the reduction in homodimerization was a significant reduction in agonist-stimulated adenylyl cyclase activity. Based on the model of Hebert et al (50), Bai et al (51) suggested that a similar domain plays a role in the high level of dimerization observed in the CaR and identified a putative consensus dimerization motif, [LMALGFLIGYTCL], in TMV of CaR. Although no further studies have been done to provide evidence that this domain is in fact playing a role in the dimerization of CaR, a naturally occurring frameshift mutation at codon 747, identified in an individual with familial hypocalciuric hypercalcemia syndrome, terminates the receptor in TMV (52). Pearce et al (53) showed that this truncated receptor, which is lacking TMV, cannot form dimers

under the same conditions as do wild-type CaRs; it is likely that domains lacking in the truncated CaR are responsible for the loss of dimerization properties. It is interesting that several other mutations in CaR have been isolated in this region. Watanabe et al (54) isolated a Phe788Cys substitution just downstream of the putative consensus dimerization domain in a Japanese family with severe familial hypoparathyroidism. De Luca et al (55) described a patient with sporadic hypoparathyroidism who was heterozygous for a Leu77Arg mutation. This muta-

tion lies within the putative consensus dimerization domain: [LMAL’7? GFLIG

YTCL ]. Functional studies of this mutant have not been reported, but it would be interesting to see whether there is an observed difference in dimerization and

GPCR POLYMORPHISMS

603

signaling. In receptors for which dimerization is critical for function, as with the CaRs, a mutant receptor may act as a dominant negative, greatly reducing the function of the expressed normal receptor. For example, the Arg/85Gin N-terminal mutant of the CaR, which was identified in a female infant with hyperparathyroidism, is thought to exert a strong negative effect on the function of the wild-type CaR, resulting in an unusually severe form of hypercalcemia (56).

INTRACELLULAR LOOP DOMAINS Whereas amino acid sequences in the TM domain contribute primarily to ligand binding and structural stability, the IL domains are particularly important for receptor interactions with signaling and regulatory proteins. The N- and C-terminal regions of IL3 have been implicated in G protein coupling and specificity for class A and B GPCRs (57-59); IL1 and {L2 may also be important in G protein interactions (3). The IL domains of GPCRs are also involved in interactions with other proteins, such as £-arrestins (60), and contain consensus sites for phosphorylation by G protein receptor kinase (GRK) (61) and second messenger kinases, such as protein kinase A (62) and protein kinase C (63), which mediate receptor phosphorylation and desensitization. Further, a Pro-rich motif has been identified in IL3 of some GPCRs (e.g. B-AR and D4) as a site for interaction with the Src homology domains of such proteins as Nck and Grb2, which may be involved in receptor internalization (64-66). Variants of the IL domains have been identified that can result in constitutive activity of GPCRs (Table 3), which suggests that

TABLE 3

Selected genetic variants of the IL, EL, and C-terminal domains®

Receptor

Variant (domain)

Functional consequence

Reference

B\-AR

Gly389Arg (C)

Enhanced agonist-stimulated coupling to G,

108

D2

Ser311Cys (IL) Pro310Ser (IL)

Decreased agonist binding affinity Altered receptor coupling to G;

74 74

ET,

Ser390Arg (C)

Altered G protein coupling (Hirschsprung’s disease)

105

P2NG

Arg334Cys (C)

Putative additional palmitoylation site

103

Rhodopsin

Arg/35Gly (IL)

Unable to mediate G protein release of GDP

85

TSH

Ala623Iso (IL) Ala623Gly (IL)

Constitutive activity

iS)

Unable to couple G,

84

Ala623Ser (IL)

V>

Arg137His (IL)

JL, Intracellular loop; EL, extracellular loop; AR, adrenergic receptor; D2, dopamine 2; ET, endothelin B; TSH, thyroid stimulating hormone; V2, vasopressin 2.

604

RANA

® SHIINA

® INSEL

the wild-type receptor sequence of IL2 and IL3 helps maintain the inactive conformation of GPCRs whereas constitutively active substitutions stabilize, as do agonists, the active conformation of the receptors. Dopamine receptors are the most well-documented GPCRs for which polymorphisms have been identified in IL domains. Variants of dopamine 2 (D2) receptors have been associated with a number of disorders, including substance abuse and schizophrenia (67—70). The Pro310Ser and Ser311Cys of the D2 receptors are located in IL3, with similar distribution patterns in different human populations. The Cys311 polymorphism occurs at a frequency of 3% in the Taiwanese Chinese population (71), about 2% in the Japanese population (72), and 3% in the Caucasian population (73). The Pro310Ser variant is rarer, occurring at a frequency of 0.4% in Caucasians (73). Cravchik et al (74) showed that the ligand binding affinity of the Ser311Cys variant for dopamine was approximately twofold lower than the affinities of the Pro3JOSer variant and D2 wild-type. Despite the small difference in binding affinity, there was a marked difference in the ability of variants to couple to G, and inhibit adenylyl cyclase activity. Although wild-type receptor inhibited forskolin-stimulated levels of cAMP synthesis by >90%, the Pro3]0Ser and Ser311Cys variants were able to inhibit forskolin response by only 24% and 58%, respectively. Such results suggest that these IL3 variants have a decreased ability to allow for the agonist-induced conformational changes required to activate G,. Another intriguing polymorphism of dopamine receptors is in D4 receptors, which, like D2 receptors, inhibit adenylyl cyclase activity. The antipsychotic agent clozapine has increased affinity for D4 compared with D2 receptors, perhaps in part explaining its antipsychotic activity (75). Among the numerous polymorphisms obtained in the amino-terminal and TM regions is a highly polymorphic, hypervariable 48-bp tandem repeat, termed variable number tandem repeat, in IL3 [as reviewed by Van Tol (76)]. Individuals possess two to ten tandem repeat units; at least 19 different sequence variants have been identified in the individual units, and in excess of 20 different polymorphic forms of the D4 receptor have been identified in IL3. In spite of this considerable variation in IL3, few definitive data have documented that these polymorphisms are associated with disease, although some evidence suggests an association of the repeat polymorphism with novelty seeking and attention deficit hyperactivity disorder. In addition, pharmacological. studies have not revealed major differences between the different repeat polymorphic variants of IL3 in D4. A highly conserved sequence found in members of class A GPCRs is the Glu/Asp-Arg-Tyr/His (E/DRY/H) motif located at the TM3-IL2 junction. Sitedirected mutagenesis of the Asp residue of the DRY motif results in constitutive activity of GPCRs, including 8,-AR (77), the cannabinoid receptor (78), and the histamine H, receptor (79). Mhaouty-Kodja et al (80) provide evidence that the

Asp 142 residue of the DRY motif in the w;,-AR plays a role in receptor interaction with regulatory proteins such as GRK2 and £-arrestin. The authors suggest that the negative charge of this residue may be directly involved in interacting with GRKs and -arrestins or that mutations of Asp142 may change the conformation of IL2 and IL3 such that these loops can no longer bind to activate GRK2, thereby

GPCR POLYMORPHISMS

605

reducing desensitization. The second residue of the DRY motif, Arg, is the most

conserved amino acid of this motif, as revealed through sequence alignment of 620 receptors (81). Mutations of the Arg residue impair receptor-mediated si gnal transduction in M, muscarinic receptors (82). Scheer et al (83) recently demonstrated that substitution of the Arg143 of the DRY motif in a,,-AR by several different amino acid residues results in an array of functional effects. A change to Ala or lle resulted in a complete loss of receptor-mediated response, whereas a change to Lys conferred constitutive activity to the receptor. Using computer-simulated mutagenesis and results from site-directed mutagenesis experiments, the authors conclude that Arg143 of the DRY motif in w,-AR mediates receptor activation by enabling the intracellular loops to achieve the proper conformation for interacting with the G protein. Although naturally occurring mutations of the Asp residue have not been reported, disease-associated mutations of the conserved Arg residue of the DRY sequence have been identified in the V, receptor and rhodopsin. Rosenthal et al (84) identified an Arg/37His variant of V, receptor found segregating in individuals of a family with X-linked NDI. The Arg/37His variant was able to bind Arg vasopressin with normal affinity but was unable to couple to G,. Several mutations of the Arg residue have also been identified in rhodopsin (Arg/35Gly, Arg /35Leu, Arg135Pro, Arg135Trp) in individuals suffering from dominant forms of RP. Functional studies on the Arg/35Gly variant showed that this mutant receptor was able to bind the G protein transducin but was unable to mediate the release of GDP on light stimulation (85). A mutation of the Arg of the DRY motif has also been reported in the melanocortin-1 receptor, which is thought to contribute to skin and hair pigmentation in humans. The Arg/42His variant has been identified in an individual with red hair (86). The contribution of this allele to the red hair phenotype in humans has not been determined; functiona! studies on this variant are lacking. Three variants of the TSH receptor have been found in individuals with thyroid disease at the amino acid position homologous to Ala293 of a,,-AR, a site at which mutagenesis yields constitutive activation (87). The two somatic mutations, Ala623Ile and Ala623Ser, and one germline mutation,

Ala623 Val, result in

constitutively active receptors (15), which decrease the response to endogenous or exogenous TSH. Although we are not aware of studies documenting expression in humans, it is intriguing that site-directed mutagenesis studies on amino acid residues homologous to the Ala293 of a@;,-AR in a,-AR and £,-AR also result in constitutively active receptors (88, 89). Studies on a@},-AR, @-AR, and 6,-AR showed increased constitutive activity of these mutants combined with increased agonist-independent GRK2-mediated phosphorylation. (See page 624 for Note.)

THE EXTRACELLULAR LOOP DOMAINS Of the GPCR

domains discussed in this review, a smaller number of polymor-

phisms have been reported in the EL domains compared with other GPCR regions (Table 3). The most prominent structural feature of these domains is the disulfide

606

RANA

® SHIINA

® INSEL

bridge found in most GPCRs that is formed by two conserved Cys residues of EL1 and EL2. An additional disulfide bridge between a Cys in EL3 and the N-terminal domain can also occur and has, for example, been found in the angiotensin receptor. These disulfide bridges are thought to be important for proper receptor folding, with consequences for intracellular transport (as discussed in the N-Terminal Domain section), and the formation of a ligand binding pocket (90-92). Mutations that contribute to additional Cys residues in EL domains can compete for disulfide bond formation with the conserved Cys and perturb receptor function. For example, the Gly185Cys NDI-causing mutation of EL2 of the V, receptor has impaired binding, possibly because of competition with the conserved EL2 cysteine for the formation of a disulfide bridge with the conserved EL1 Cys (93). Site-directed mutagenesis studies have shown that the EL domains play a key role in ligand binding and ligand specificity in other GPCRs (e.g. 94-96). Tsukaguchi et al (41) reported that the Arg202Cys variant of the V, receptor, found in a patient with NDI, has impaired binding and decreased activation of adenylyl cyclase by vasopressin. The binding impairment of the mutant receptor suggests that this residue participates in ligand recognition. Cotte et al (97) showed that a Leu residue at the same position in the rat V, receptor facilitates ligand binding, whereas the Arg at position 202 in the human receptor is not favorable. A change to Cys at this position may constitute an even more negative determinant restricting receptor interaction with ligand. Such effects would imply differences in the ability of the variant receptors to be activated by endogenous and exogenous agonists.

THE C-TERMINAL DOMAIN The intracellular carboxy-terminal (C-terminal) domain is involved in several aspects of GPCR signaling. All GPCRs contain Ser and/or Thr residues in this domain, which can serve as sites for G protein receptor kinase (GRK)-mediated phosphorylation and receptor desensitization (98,99). Some GPCRs contain a cysteine residue in the C-terminal domain, which can serve as a site for palmitoylation. This can create a fourth IL because of the ability of the palmitoylated cysteine to insert in the plasma membrane. Also contained in the C-terminal domain upstream of the palmitoylated cysteine residue and downstream of TMVII is a G protein-binding domain in which variants have been identified (Table 3). More recent evidence suggests that the C terminus may be involved in interactions with other proteins that mediate GPCR signaling, such as the recently identified calcyon (100), PDZ domain-containing proteins (101), and Homer/Vesl proteins (102), although GPCR variants that alter interactions with these proteins have not yet been established. Class A GPCRs contain the palmitoylated cysteine in the C terminus and this can be a site for polymorphism. Janssens et al (103) identified a polymorphism with a frequency of 0.2 in the P2Y, receptor, which produces an Arg334Cys substitution at a potential site for palmitoylation. Although there were no major pharmacological differences in response to natural agonists tested between the two polymorphic

GPCR POLYMORPHISMS

607

receptors, the authors observed slight variation in the time course of second messenger generation: InsP; accumulation was slower in the Cys334 variant than in the Arg334 variant. This site may have pharmacological relevance. P2Y, receptors are expressed in airway epithelial cells and respond to the nucleotides UTP and ATP by increasing intracellular calcium and thereby regulating chloride conductance in a manner alternative to the activity of the cystic fibrosis transmembrane regulator (CFTR). As such, P2Y, receptors are under study as therapeutic targets with nucleotide analogs for the treatment of cystic fibrosis and chronic bronchitis. Thus, expression of this polymorphism might impact therapeutic utility of nucleotides in the treatment of such diseases. The Arg492Cys variant of the a ;,-adrenergic receptor was identified in a study of patients with benign prostatic hypertrophy (104). The substitution of a Cys at residue 492 may confer a palmitoylation site in the C terminus next to the postulated palmitoylation site at 490 for this receptor. This polymorphism showed no marked pharmacological differences in binding or receptor-mediated intracellular Ca2+ levels, nor was it differently distributed between patients with benign prostatic hypertrophy and normal control subjects. However, the distribution between the Japanese and US populations was substantially different: Arg492 is the major allele in the Japanese population (90%), whereas Cys492 is the major allele in the US population (66%).

Prior to the palmitoylated cysteine in the proximal region of the C-terminal domain is a region sometimes referred to as the fourth IL domain, which contains a G protein binding site. Variants in this region that alter G protein coupling have been identified in at least two different GPCRs. Tanaka et al (105) identified a

Ser390Arg variant of the endothelin B (ET) receptor in a Japanese patient with Hirschsprung’s disease. This variant has binding affinities similar to the wildtype ET, receptor but is characterized by a decreased ligand-induced intracellular calcium level and decreased inhibition of adenylyl cyclase activity. A similar region of the related ET, receptor contains the putative palmitoylation site and forms a fourth cytoplasmic loop required in G protein coupling (106). The data suggest that the replacement of a Ser at position 390 with the positively charged Arg residue might cause a decrease in G protein coupling through conformational changes. The same region in two other class A GPCRs, the 6,-AR and a@,-AR, has been implicated in G protein binding (107); a naturally occurring polymorphism of this region has been identified in the 6,-AR. Mason et al (108) observed that the initially described “wild-type” sequence of 6,;-AR, which codes for a Gly at residue 389 in the putative G protein binding domain, actually occurs at a frequency of 0.26, whereas a “gain-of-function” Arg389 polymorphism is at a frequency of 0.74 and thus more likely to be the true wild-type. Several types of experiments, such as receptor-promoted binding of GTPyS to G, and adenylyl cyclase assays, showed that agonist-stimulated coupling to G, was enhanced in the Arg389 allele in comparison to the Gly389 allele. Given the key role of 6 ,|-AR as the predominant B-AR subtype in the heart involved in responses to B-adrenergic agonists, genetic variation at 6 ,-AR might contribute to interindividual differences in the response to B-blockers in the treatment of cardiovascular diseases. The study by Mason et al

RANA

608

® SHIINA

® INSEL

(108) emphasizes the importance of population studies to identify the expression and frequency of the genetic variants. A substitution 20 residues away from the end of the C-terminal domain of the 5-HT,, receptor, creating the His452Tyr variant, has been identified in two independent cohorts of unrelated patients treated with clozapine (109). Functional data on this variant are lacking; however, the study did find a statistically si gnificant association of the Tyr452 allele with nonresponders to clozapine. Although the precise target of clozapine’s therapeutic action is not known, studies such as these, which associate variants of a neurotransmitter receptor to which the drug binds, strengthen the candidacy of these receptors as important therapeutic targets.

NONCODING-REGION POLYMORPHISMS AND OTHER TRANSCRIPTIONAL AND TRANSLATIONAL MECHANISMS ALTERING EXPRESSION A large number of polymorphisms have been identified within the promoter region and 5’ untranslated region (UTR) of GPCRs: fewer polymorphisms have been reported thus far for other noncoding regions, such as in the 3’ UTR and introns (Table 4). Because levels of receptor expression are important in determining the magnitude and sensitivity of response to agonists (110), it is not surprising that GPCR polymorphisms in noncoding regions have been associated with disease and altered drug responses. In this section, we give examples of such polymorphisms and suggest mechanisms by which expression of GPCRs may be altered. Aside from SNPs that have been identified in the noncoding region of human GPCR loci, several other mechanisms that alter receptor expression have been identified but are rare or have only been identified in the GPCR loci of nonhumans. Such disease-related polymorphisms can show altered response to endogenous

TABLE 4

Selected genetic variants the noncoding region* Functional consequence (clinical consequence)

Receptor

Variant (region)

B>

—412C/G (5'UTR)

Destroys transcription factor binding site

122

B>o-AR

—47T/C (5'LC)

Decreased receptor expression

131

CCRS

—1454G/A (5 UTR)

Reduced promoter activity

123

D4

—521C/T (promoter)

D2

—141C Ins/Del (promoter)

Lower promoter activity

eS

ET,

Splice variant

Defective in G protein coupling

136

— Decreased transcriptional activity

Reference

114

“5'UTR, 5’ untranslated region; AR, adrenergic receptor; 5/LC, 5’ leader cistron; CC, chemokine receptors; D4, dopamine 4; D2, dopamine 2; Ins, insertion; Del, deletion; ET, endothelin B.

GPCR POLYMORPHISMS

609

hormones or neurotransmitters, and presumably to exogenous agents as well, although few studies have directly tested this inference. An example is the insertion of a 226-bp, short, interspersed nucleotide element within the 5’ flanking intronic region needed for pre-mRNA lariat formation and proper splicin gin the hypocretin receptor 2 gene (HCRTR2), which binds orexin peptides derived from the precursor peptide hypocretin. Deficiency of hypocretin can produce sleep pathologies resembling narcolepsy (111); faulty splicing of HCRTR2 yields a receptor incapable of proper signal transduction and results in narcolepsy in dogs (112). Such studies provide a hypothesis for assessing the pathogenesis of human narcolepsy and a potential drug target for the disease. RNA editing, amechanism whereby the primary nucleotide sequence of an RNA transcript is posttranscriptionally modified, provides another means to generate expression of functionally different GPCRs. Niswender et al (113) showed that the brain can express two different edited forms of the human 5-HT)¢ receptor, each with different constitutive activity from the nonedited form. A more common mechanism for altering expression of GPCRs is alternate splicing. This is discussed in more detail below.

SUE As indicated above, dopamine (D) receptors are targets of antipsychotic drugs, and polymorphisms affecting the expression of dopamine receptor genes (DRD) can contribute to individual variation in response to drug therapy. Okuyama et al (114) identified a —521C/T polymorphism in the 5’ promoter region of DRD4. Reporter assays were used to assess the effect of this polymorphism on receptor expression and revealed that the —52/T allele had 40% less transcriptional activity than did --521C. The authors observed a weak association of the more transcriptionally active polymorphism with schizophrenia. However, because an elevation of DRD4 mRNA

is found in the frontal cortex of postmortem brains from schizophrenics,

the authors proposed that this polymorphism may be a candidate for a genetic factor that influences both schizophrenia and the response to clozapine therapy, which targets D4 receptors, an idea that requires further testing. As with the D4 receptor, the D2 receptor may be involved in schizophrenia and is a major site of action of neuroleptic agents used in its treatment. Arinami et al (115) identified a —/41C Ins/Del polymorphism in D2 receptors and an association of the insertion polymorphism with schizophrenia in a case-control study in a Japanese population. In reporter assays, the —/4/C Del polymorphism had lower promoter activity. Breen et al (116) reported the reverse association in a case-control study of schizophrenia in a Caucasian population, but they attributed the results to linkage disequilibrium with a nearby polymorphism. This DRD2 polymorphism serves as an example that genetic background in different populations can contribute to a role of specific polymorphisms in disease and drug response. A number of 5’ UTR and promoter polymorphisms have been identified in other candidate genes for schizophrenia, although further investigation is required to

610

RANA

® SHIINA

® INSEL

establish their role in this or other diseases. Because a decreased density of 5-HT>, receptors has been reported in the frontal cortex of schizophrenic patients (117), many studies have been conducted to identify polymorphism in the promoter region of its gene. A —1438G/A polymorphism was found, but the results showed no clear association with schizophrenia (118). However, Collier et al (119) have reported

that this polymorphism is associated with anorexia nervosa. Polymorphisms have been identified in two other candidate genes for schizophrenia, but functional analysis and studies associating these polymorphisms with this or other diseases are lacking: —19G/A and —/8C/Tin the 5’ UTR ofthe 5-HT;, receptor gene (120), and —707C/G and —343A/G of the DRD3 promoter (121). Considerable effort has been directed at defining noncoding changes in GPCRs that may be associated with cardiovascular diseases. One example is the gene for the bradykinin B, receptor. Erdmann et al (122) assessed the role of three novel, but rare, promoter variants isolated in individuals with cardiac disease: A —4]2C/G variant, identified in a patient with dilated cardiomyopathy, destroys a binding site for the transcription factor Sp1, which affects basal gene transcription; a —704C/T mutation, identified in an individual with hypertrophic cardiomyopathy, destroys the binding site of a nuclear binding protein; and a —78C/T mutation, also isolated from an individual with dilated cardiomyopathy, reduces protein binding of an unidentified protein. It will be of interest to determine whether these polymorphisms influence efficacy or toxicity of agents such as angiotensin converting enzyme (ACE) inhibitors, which blunt the degradation of bradykinin. Perhaps the best example of promoter polymorphism affecting a noninherited disease is that of the CC chemokine receptor-5 gene (CCR5) promoter polymorphism in HIV-1 pathogenesis. McDermott et al (123) identified an A/G polymorphism at bp 59029 of the CCR5, 1454 bp upstream of the initiation site in the CCRS promoter. Both promoter alleles were common (43%—68% allelic frequency for —/454A, depending on race). In vitro promoter activity of the —]454G polymorphism was 45% lower in activity than was —/454A. In a cohort of HIV-1 seroconvertors lacking both CCRSA32 and CCR2-64] alleles, the —]454G/G individuals progressed to AIDS on average 3.8 years more slowly than did the —1454A/A individuals. CCR5 —1454G/G thus appears to be protective relative to CCRS —1454A/A, and about twice as protective relative to CCRSA32 or CCR2641. This effect may result from reduced CCR5 mRNA production. These results identify a site in the CCR5S promoter that may be a useful target for treatment of HIV-1 infection.

3’UTR The angiotensin II type 1 receptor gene (ATIR), which is expressed in vascular smooth muscle cells and regulates growth and vasoconstriction in response to angiotensin II, has been particularly well studied for a 3’UTR polymorphism, an A/C transversion found at position 1166. Bonnardeaux et al (124) and Wang etal (125) observed an increase in allelic frequency of 1/66C in populations of Caucasian individuals with essential hypertension and inferred that this polymorphism

GPCR POLYMORPHISMS

611

imparted a small effect on blood pressure. Szombathy et al (126) evaluated fewer than 100 subjects but reported that the //66C allele, although not more common in hypertension, was associated with higher values for systolic and diastolic blood pressure in overweight, hypertensive patients. These results suggest that the 1166C polymorphism is potentially involved in the regulation of blood pressure, with the effects of the genotype being more pronounced in overweight, and perhaps older, subjects. Miller et al (126a) evaluated 66 healthy Caucasian subjects and demonstrated the association of the C allele with lower baseline renal function, as assessed by several indices, and with greater hemodynamic

and renal

responses to the AT receptor antagonist losartan. The effect of this polymorphism may contribute to other cardiovascular diseases as well. Szombathy et al (127) reported an association of the 1/66C polymorphism with mitral valve prolapse syndrome in Caucasians. Although the A/C alleles at nucleotide 1166 of ATIR failed to show an association with early coronary disease in a study by Alvarez et al (128), a significant interaction between a deletion polymorphism of ACE was found. Individuals homozygous for the deletion polymorphism of ACE and the 1/66C polymorphism of AT/R had an increased risk for coronary artery disease. This suggests the importance of genetic background in the involvement of polymorphisms with disease and perhaps in drug response as well.

5’ Leader Cistron The 5S’ UTRs of eukaryotic mRNA can contain sequences that regulate the efficiency of translation of the mRNA. One example is upstream AUG sequences that code for initiation of short open reading frames (UORFs), which influence the translation of the main ORF (129). Recent data suggest that polymorphism can occur in such uORFs and that such polymorphisms can influence GPCR expression. For example, the 6,-AR contains a short ORF termed the 5’ leader cistron (5’LC). This sequence begins 102 bp upstream of the 6,-AR coding sequence and encodes a putative 19—amino acid peptide. This 5’LC inhibits 6,-AR translation; mutational inactivation of the 5’LC increases receptor expression (130). McGraw et al (131) have identified a T to C transversion polymorphism at —47, leading to a putative peptide product containing an Arg instead of a Cys at residue 19 (Arg/9Cys) in the C terminus of this ORF. Of particular importance, this 5’LC-Arg/9 variant acted to lower the basal level of receptor expression. Although this result will require confirmation from other studies because others have obtained somewhat different results (132), it could provide a potential explanation for at least part of the substantial intersubject variability in 6,-AR expression. Expression of the 5’LC polymorphism for this receptor (and perhaps other GPCRs) could be a key determinant of both “basal” and agonist-regulated levels of receptor expression in light of evidence that the 5’LC-Arg/9 variant is in linkage disequilibrium with the coding sequence polymorphisms at codons 16 and 27, which influence susceptibility of receptors to undergo agonist-promoted down-regulation, as discussed above (131). Yamada et al (133) found an increased frequency of the 5’LC-Arg/9 variant in obese individuals compared with nonobese individuals. In addition, they found

612

RANA

® SHIINA

® INSEL

a polymorphism at —20 in which the —20C allele was in linkage disequilibrium with the —47C polymorphism. Another example of a 5’LC is found in the promoter of the D3 receptor gene, which encodes a putative peptide of 36 amino acids. Sivagnanasundaram et al (121) found that a polymorphism in the 5’LC, the Lys9Glu, and two SNPs in the promoter region, —707C/G and —343 A/G, and a coding polymorphism Ser9Gly were in tight linkage disequilibrium and associated with schizophrenia. Further studies will be needed to confirm and extend these findings. Conceivably, this or other 5’LCs could provide therapeutic targets to regulate GPCR expression.

Alternate Splicing Molecular diversity in the GPCRs is also achieved through alternate splicing. For example, DRD2 generates two alternative RNA isoforms, the D,, (long) and the D>, (short), that differ by a 29—amino acid sequence in the third cytoplasmic loop of the receptor protein (134). Because such isoforms of GPCRs may differ in pharmacological properties and can be expressed differentially in tissues and in development, it is important to consider alternate splicing events in therapeutic drug design targeting GPCR isoforms. For example, the full-length CaR is expressed in undifferentiated keratinocytes, whereas its expression is decreased as keratinocytes differentiate, at which time there is an increase in an alternatively spliced “loss of function form” of CaR, which lacks exon 5 (135). A splice variant has also been identified in the ET, receptor. The binding properties of this ET, splice variant were similar to those of wild-type ET, receptors, but the splice variant is defective in G protein coupling, as implied by a lack of increase in inositol phosphate accumulation in response to agonists (136). To date, no studies have directly assessed the role of splice variants in differences in drug response, but evidence from the receptors discussed above suggest that this may be a key area for further studies. A splice variant of the 5-HT, receptors has also been identified in humans (137). This variant has a shorter C-terminal domain that contains a unique sequence with two Pro-Val repeats, which cause the receptor to have constitutive activity. This splice variant has been found only in brain, whereas 5-HT, receptors are expressed in a wide variety of tissues, including brain, colon, urinary bladder, esophagus, and heart, and 5-HT, receptor agonists, such as metoclopramide and cisapride, are used therapeutically as prokinetic drugs. Such drugs have been shown to be superagonists in colliculi neurons but only partial agonists or antagonists in other cells (138). Differentially expressed splice variants probably contribute to such tissue-specific responses.

TRUNCATED RECEPTORS Genetic variations altering multiple domains through frameshift mutations have also been identified; these frameshift mutations result in truncated receptors that can influence the pathogenesis and treatment of disease. Here, we give two examples, one that results in disease and another that confers a selective advantage

GPCR POLYMORPHISMS

613

to the carrier for protection against disease. We make mention of the possibility of truncated receptors because the latter has given insight into new therapeutic approaches to disease. In a study of families with morbid obesity, Vaisse et al (139) identified the presence of a heterozygous 4-bp insertion in the coding sequence of the melanocortin 4 receptor (MC4R). This insertion results in a truncated receptor that lacks TMVI and TMVII. MC4R is expressed in the hypothalamus, and studies of mice have implicated this receptor in the regulation of body weight. Mice that do not express MCA4R display morbid obesity with hyperinsulinemia. Another study identified a 4-bp deletion in one individual from a cohort of severely obese children (140). This deletion at codon 211 results in a frameshift that introduces a stop codon in the region encoding TMS. Although functional studies of these truncated receptors were not done, these variant receptors are likely to be nonfunctional and probably contribute to the obese phenotype in individuals carrying these variants. A deletion in the coding region of the CCRS5 can also be beneficial. The 32bp deletion of nucleotides 794 to 825 of the coding region of the CCR%5 results in a frameshift that truncates the receptor after codon 206 (141) and is common to many populations, including those of European descent, Indians, and Middle Easterners (142). Although the truncation does not seem to produce an aberrant phenotype in individuals carrying this variant, this polymorphism is a factor in HIV-1 resistance in Caucasians with complete penetrance; infected Caucasian individuals heterozygous for the polymorphism show a slower rate of disease progression. Because the individuals homozygous for the truncated receptor seem to be normal, the existence of this polymorphism suggests that CCRS may be a good candidate for antiretroviral therapy. Hall et al (143) reported that individuals carrying the deletion polymorphism are at reduced risk of developing asthma, , and this may help explain the high prevalence of this mutation in the general population. Thus, CCR5 may also be a potential therapeutic target for asthma therapy, in addition to therapy related to HIV infection, as noted above.

SUMMARY AND CONCLUSION It is apparent that numerous naturally occurring variants exist in virtually all domains of GPCRs, but thus far, definitive understanding of their impact on disease and drug therapy is incomplete, with the exception of only a small number of receptors and subsets of variants. We believe that there are three main reasons for this current gap in knowledge. (a) A comprehensive compilation of all common and important variants is not yet available for most GPCRs. Completion of the human genome project and more extensive analyses of SNPs of GPCR genes should be forthcoming (from within months to a few years) and should define the “suspects” that will need to be interrogated. (b) Detailed and precise functional data have not been available for the vast majority of GPCR variants. (c) Careful statis-

tical genetic analyses have not generally been undertaken to define relationships between GPCR variants, disease, and response to drugs largely because of the lack of sufficient data to achieve statistical significance.

614

RANA

® SHIINA

® INSEL

A starting point for the analysis of GPCR variants is a comparison of detected sequence variations with information available in public databases, such as Genbank. It is thus intriguing that as new data become available for GPCR loci, one learns that the database version of the wild-type form of the receptor is actually a rarer variant than certain other “polymorphic” forms (e.g. 108, 131). This is particularly the case when one begins to look at different populations, implying a need to consider that there may be no single wild-type form of a given receptor and that, instead, one must qualify such information as relevant for only certain

(sub)populations. Such an idea has potentially important implications in terms of disease associations, disease progression, and drug therapy, especially if different receptor alleles show important functional differences. We propose that the following set of studies be conducted to establish the role of a GPCR polymorphism in a complex disease (or trait) and in order to assist in designing appropriate pharmaceutical strategies for the treatment of the disease: (a) population studies to ascertain the distribution of the polymorphism in affected and unaffected individuals in different ethnic populations, (b) in vitro studies to assess the impact of the polymorphism on GPCR structure-function. and (c) in vivo studies to determine the physiological and pharmacological consequences of the polymorphism both in normal subjects and those with a given disease. As suggested above, efforts to decipher the role of GPCR polymorphisms in pathogenesis and therapy of complex diseases must consider the varying expression of GPCR variants among different populations. Several studies suggest that there are ethnic differences in the distribution of GPCR polymorphisms (144-146). For some GPCRs, this variation may manifest in ethnic differences in clinical phenotypes. For example, a number of studies have observed ethnic variation in the efficacy of B-AR antagonists as antihypertensive drugs: Chinese men are the most sensitive, African-American men are the least sensitive, and Caucasian-American men show intermediate sensitivity (147, 148). A thorough population study of the B-ARs, the targets of these drugs, is necessary to understand this observed difference in response to medication. Hence, to design appropriate therapeutic strategies for a complex disease, a study of the population distribution of the polymorphism must be conducted to determine the ethnic distribution of the polymorphism in affected and unaffected individuals, and to determine whether the genetic back-. ground of the individual will affect response. Second, in vitro studies are required to assess whether the polymorphism has an impact on GPCR function. Studies on polymorphic receptors commonly employ heterologous expression systems to assay for binding properties, changes in the level of second messengers, G protein coupling, and desensitization and internalization. Drawbacks of such systems include the possible loss of functional information due to receptor expression outside the genetic background of the individual carrying the polymorphism, or due to expression outside the genetic background of its endogenous cell. More informative strategies might employ isolated cells or cell lines that endogenously carry the polymorphic receptor, preferably isolated from individuals who carry the polymorphism. For example, McGraw et al (131)

GPCR POLYMORPHISMS

615

used human airway smooth muscle cells endogenously expressing various forms of the B-AR to study the impact of receptor polymorphisms on receptor binding and cyclic AMP formation. We believe that the use of cells endogenously expressing GPCR polymorphisms may provide particularly relevant systems for in vitro functional studies designed to enhance understanding of the role of GPCR polymorphisms in complex diseases and their impact on drug therapy. Ultimately, the role of a GPCR polymorphism must be related to the in vivo setting. Pinpointing the effect of a single polymorphism to a complex trait can be difficult, as the contribution of a polymorphism at an individual locus is generally small and large numbers of subjects may need to be studied to provide strong Statistical power for association and linkage. The possibility that linkage disequilibrium may account for observed patterns of statistically significant patterns of linkage to a particular locus must also be carefully considered. Pharmacologists must be alert to the usefulness of information gleaned from both in vitro and in vivo studies of endogenous agonists (e.g. hormones and neurotransmitters) acting on polymorphic receptors. Such results can yield unexpected, but potentially important, hints regarding structurally important residues involved in the regulation of receptor expression and function. In addition, information derived from studies of polymorphic receptors has the potential to reveal regions of receptors—or the receptors themselves, as in the case of HIV infection and the CC5 receptor—to which new types of therapeutic agents might be targeted. Overall, the impact of studies on GPCR polymorphisms in therapeutics is likely to be considerable as we move into an era in which we are able to subclassify groups of patients with what was previously considered a disease/syndrome, and as we recognize that pharmacodynamics depends as much, or more, on underlying genetic differences between individuals than on environmental factors. Given the key role of GPCRs in signal transduction and functional regulation in virtually every organ system, studies of polymorphisms of GPCRs seem destined to play a major role in therapeutics in this decade and beyond.

ACKNOWLEDGMENTS Work in our laboratory on this topic has been supported by grants from NIH. We thank Dr. David Weiner (Acadia Pharmaceutical, Inc., San Diego, CA) for helpful discussions and Linda Pan for her assistance in editing this manuscript. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Bargmann I. 1998. Neurobiology of the Caenorhabditis elegans genome. Science 282:2028-33 2. Rubin

GM,

Yandell

MD,

Wortman

JR,

Miklos GLG, Nelson CR, et al. 2000. Comparative genomics of the eukaryotes. Science 287:2204-15 3. Wess J. 1998. Molecular basis of receptor/G

616

RANA

#® SHIINA

protein-coupling Ther. 80:231-64

® INSEL

selectivity.

Pharmacol.

TE, Bouvier M. 1998. Structural and functional aspects of G protein-coupled receptor oligomerization. Biochem. Cell. Biol. 76:1—-11 . Ji TH, Grossmann M, Ji I. 1998. G protein-coupled receptors. I. Diversity of receptor-ligand interactions. J. Biol. Chem. 273:17299-302 . Hebert

LS.

. Gerther U, Kobilka BK. 1998. G protein-

coupled receptors. II. Mechanisms of agonist activation. J. Biol. Chem. 273:17979— 82 . Lefkowitz RJ. 1998. G protein-coupled receptors. III. New roles for receptor kinases and $-arrestins in receptor signaling and desensitization. J. Biol. Chem. 273:18677— 80 . Schéneberg T, Schultz G, Gudermann

.

18.

T.

1999. Structural basis of G protein-coupled receptor function. Mol. Cell. Endocrinol. 151:181-93 . Gether U. 2000. Uncovering molecular mechanisms involved in activation of G protein-coupled receptors. Endocrine Rev. 21:90-113

10. Horn F, Weare J, Beukers MW, Horsch S, Bairoch A, et al. 1998. GPCRDB: an information system for G protein-coupled receptors. Nucleic Acid Res. 26:277-81 iLike Bockaert J, Pin JP. 1999 Molecular tinkering ofG protein-coupled receptors: an evolutionary success. EMBO J. 18:1723-29 We Laburthe M, Couvineau A, Gaudin P, Maoret J-J, Rouyer-Fessard C, Nicole P. 1996. Receptors for VIP, PACAP, secretin, GRF, glucagons, GLP-1, and other members oftheir new family of G protein-linked receptors: structure-function relationship with special reference to the human VIP-1 receptor. Ann. NY Acad. Sci. 805:94-111 Daiger SP, Sullivan LS, Rodriguez JA. 1995. Correlation of phenotype with genotype in inherited retinal degeneration. Behav. Brain Sci. 18:452—67 Oksche A, Rosenthal W. 1998. The molec-

.

.

ular basis of nephrogenic diabetes insipidus. J. Mol. Med. 76:326—37 Holzapfel HP, Fuhrer D, Wonerow P, Weinland G, Scherbaum WA, Paschke R. 1997. Identification of constitutively activating somatic thyrotropin receptor mutations in a subset of toxic multinodular goiters. J. Clin. Endocrinol. Metab. 82:4229-33 Polak M. 1999. Hyperfunctioning thyroid adenoma and activating mutations in the TSH receptor gene. Arch. Med. Res. 30:510-13 Liggett SB. 2000. Beta(2)-adrenergic receptor pharmacogenetics. Am. J. Respir. Crit. Care Med. 161:S197-201 Buscher R, Herrmann V, Insel PA. 1999. Human adrenoceptor polymorphisms: evolving recognition of clinical importance. Trends Pharmacol. Sci. 20:94-99 Petaéja-Repo UE, Hogue M, Laperriére A, Walker P, Bouvier M. 2000. Export from the endoplasmic reticulum represents the limiting step in the maturation and cell surface expression of the human opioid receptor. J. Biol. Chem. 275:13727-36 Green SA, Turki J, Innis M, Liggett S. 1994. Amino-terminal polymorphisms of the human beta 2-adrenergic receptor impart distinct agonist-promoted regulatory properties. Biochemistry 33:9414-19

. Turki

J, Pak J, Green

S. 1995.

Genetic

polymorphisms of the beta 2-adrenergic receptor in nocturnal and nonnocturnal asthma. Evidence that Gly16 correlates. with the nocturnal phenotype. J. Clin. Invest. 1995:163541 . Martinez

in)in)

FD,

Graves

PE,

Baldini

M,

Solomon S, Erickson R. 1997. Association between genetic polymorphisms of the beta2-adrenoceptor and response to albuterol in children with and without a history of wheezing.J.Clin. Invest. 100:3184— 88

ot Lima JJ, Thomason DB, Mohamed MH, Eberle LV, Self TH, et al. 1999. Impact

GPCR POLYMORPHISMS of genetic polymorphisms of the beta2adrenergic receptor on albuterol bronchodilator pharmacodynamics. Clin. Pharmacol. Ther. 65:519-25 24. Lipworth BJ, Hall IP, Aziz I, Tan KS, Wheatley A. 1999. Beta2-adrenoceptor polymorphism and bronchoprotective sensitivity with regular short- and long-acting beta2-agonist therapy. Clin. Sci. 96:25359 Dy. Lipworth BJ, Hall IP, Tan S, Aziz I, Coutie W. 1999. Effects of genetic polymorphism on ex vivo and in vivo function of beta2adrenoceptors in asthmatic patients. Chest 115:324-28 26. Kotanko P, Binder A, Tasker J. 1997. Essential hypertension in African Caribbeans associates with a variant of the beta2adrenoceptor. Hypertension 30:773-76 Dili Gratze G, Fortin J, Labugger R. 1999. Beta2 adrenergic receptor variants affect resting blood pressure and agonist-induced vasodilation in young adult Caucasians. Hypertension 33:1425-30 28. Timmermann B, Mo R, Luft FC, Gerdts E, Busjahn A, et al. 1998. Beta-2 adreno-

ceptor genetic variation is associated with genetic predisposition to essential hypertension: the Bergen Blood Pressure Study. Kidney Int. 53:1455—60

UE). Herrmann V, Buscher R, Go MM, Ring KM, Hofer JK, et al. 2000. Beta 2-adren-

ergic receptor polymorphisms at codon 16, cardiovascular phenotypes and essential hypertension in whites and African Americans. Am. J. Hypertens. 13:1021-6

30. Rotondo

A, Nielsen DA, Nakhai Hulihan-Giblin B, Bolos A, Goldman

B, D.

1997. Agonist-promoted down-regulation and functional desensitization in two naturally occurring variants of the human serotoninlA receptor. Neuropsychopharmacology 17:18-26

Bik Okazaki R, Chikatsu N, Nakatsu M, Takeuchi Y, Ajima M, et al. 1999. A novel activating mutation in calcium-sensing receptor gene associated with a family of au-

617

tosomal dominant hypocalcemia. J. Clin. Endocrinol. Metab. 84:363-66

32. Lappalainen J, Zhang L, Dean M, Oz M, OzakiN, et al. 1995. Identification, expression, and pharmacology of a Cys73-Ser3 substitution in the human 5-HT3¢ receptor gene (HTR2C). Genomics 27:274-79 33% Sodhi MS, Arranz MJ, Curtis D, Ball DM, Sham P, et al. 1995. Association between

clozapine response and allelic variation. NeuroReport 7:169-72 34. George ST, Ruoho AE, Malbon CC. 1986. N-glycosylation in expression and function of beta-adrenergic receptors. J. Biol. Chem. 261:16559-64 Si Hughes RJ, Pasillas M, Saiz J, Jasper J, Insel PA. 1997. Decreased transcript expression coincident with impaired glycosylation in the beta2-adrenergic receptor gene does not result from differences in the primary sequence. Biochim. Biophys. Acta 1356:281-91 36. Bond C, LaForge KS, Tian M, Melia D, Zhang S, et al. 1998. Single-nucleotide polymorphism in the human mu opioid receptor gene alters beta-endorphin binding and activity: possible implications for opiate addiction. Proc. Natl. Acad. Sci. USA 95(16):9608-13 Sf Sullivan LJ, Makris GS, Dickinson P, Mulhall LEM, App D, et al. 1993. A new codon 15 rhodopsin gene mutation in autosomal dominant retinitis pigmentosa is associated with sectorial disease. Arch. Ophthalmol. 111:1512-17 38. Bunge S, Wedemann H, David D, Terwilliger DJ, van den Born LI, et al. 1993. Molecular analysis and genetic mapping of the rhodopsin gene in families with autosomal dominant retinitis pigmentosa. Genomics 17(1):230-33 39: Sung CH, Davenport CM, Nathans J. 1993. Rhodopsin mutations responsible for autosomal dominant retinitis pigmentosa. Clustering of functional classes along the polypeptide chain. J. Biol. Chem. 268:26645—49

618

RANA

#® SHIINA

® INSEL

40. Wess J, Nanavati S, Vogel Z, Maggio R. 1993. Functional role of proline and tryptophan residues highly conserved among G protein-coupled receptors studied by mutational analysis of the m3 muscarinic receptor. EMBO J. 12:331-38 41. Tsukaguchi H, Matsubara H, Taketani S,

tor gene in multigenic Hirschsprung’s disease. Cell 79:1257-66

48. Cyejic S, Devi LA. 1997. Dimerization of the delta opioid receptor: implication for a role in receptor internalization. J. Biol. Chem. 272:26959-64 49. Rocheville M, Lange DC, Kumar U, Patel

Mori Y, Seido T, Inada M. 1995. Binding-,

SC, Patel RC, et al. 2000. Receptors

intracellular transport-, and biosynthesisdefective mutants of vasopressin type 2 receptor in patients with X-linked nephrogenic diabetes insipidus. J. Clin. Invest. 96:2043-S0

dopamine and somatostatin: formation of hetero-oligomers with enhanced functional activity. Science 288:154—57 50. Hebert TE, Moffett S, Morello JP, Loisel TP, Bichet DG, et al. 1996. A peptide derived from a beta2-adrenergic receptor transmembrane domain inhibits both receptor dimerization and activation. J. Biol. Chem. 271:16384-92 SI Bai M, Trivedi S, Brown EM. 1998. Dimerization of the extracellular calcium-sensing receptor (CaR) on the cell surface of CaRtransfected HEK293 cells. J. Biol. Chem. 273:23605-10 52 Pearce SHS, Trump D, Wooding C, Besser GM, Chew SL, et al. 1995. Calciumsensing receptor mutations in familial benign hypercalcemia and neonatal hyperparathyroidism. J. Clin. Invest. 96:268392

42. Tanaka H, Moroi K, Iwai J, Takahashi H, Ohnuma

N, et al. 1998. Novel muta-

tions of the endothelin B receptor gene in patients with Hirschsprung’s disease and their characterization. J. Biol. Chem. 273(18):11378-83

43. Yang YK, Dickinson C, Haskell-Luevano C, Gantz I. 1997. Molecular basis for the

interaction of [Nle*, D-Phe’] melanocyte stimulating hormone with the human melanocortin-1 receptor. J. Biol. Chem. 272:23000-10

44. Liu C,

IS,

Seeman

P, Sanyal S, Ulpian PE, et al. 1996.

Rodgers-Johnson

Dopamine D4 receptor variant in Africans, D4valinel94glycine, is insensitive to dopamine and clozapine: report of a homozygous individual. Am. J. Med. Genet. 61:277-82 45. Kotlar TJ, Young RH, Albanese C, Crow-

. Pearce SHS, Williamson C, Kifor O, Bai

nNws)

M, Coulthard MG, et al. 1996. A familiar syndrome of hypocalcemia with hypercalciuria due to mutations in the calcium-sensing receptor. N. Engl. J. Med. 335:1115-22

ley WE Jr, Scully RE, et al. 1997. A mu-

tation in the follicle-stimulating hormone receptor occurs frequently in human ovarian sex cord tumors. J. Clin. Endocrinol. Metab. 82(4):1020—26 46. Kremer H, Kraaij R, Toledo SP, Post M, Fridman JB, et al. 1995. Male pseudohermaphroditism due to a homozygous missense mutation of the luteinizing hormone receptor gene. Nat. Genet. 9(2):160— 64 47. Puffenberger EG, Hosoda K, Washington SS, Nakao K, deWit D, et al. 1994. A missense mutation of the endothelin-B recep-

for

an-

. Watanabe T, Bai M, Lane CR, Matsumoto

S, Minamitani K, et al. 1998. Familial hypoparathyroidism: identification of a novel gain of function mutation in transmembrane domain 5 of the calcium-sensing receptor. J. Clin. Endocr. Metab. 83:2497502 555 De Luca F, Ray K, Mancilla EE, Fan G-F, Winer KK, et al. 1997. Sporadic hypoparathyroidism caused by de novo gain-of-function mutations of the Ca(2+)sensing receptor. J. Clin. Endocr. Metab. 82:2710-15

GPCR POLYMORPHISMS

56. Bai M, Pearce

SHS,

Kifor

O, Trivedi

S, Stauffer UG, et al. 1997. In vivo and in vitro characterization of neonatal hyperparathyroidism resulting from a de novo, heterozygous mutation in the Ca(2+)-sensing receptor gene: normal maternal calcium homeostasis as a cause of secondary hyperparathyroidism in familial benign hypocalciuric hypercalcemia. J. Clin. Invest. 99:88—96 Sie Cotecchia S, Ostrowski J, Kjelsberg MA, Caron MG, Lefkowitz RJ. 1992. Constitu-

tive activation of the alpha 1B-adrenergic receptor by all amino acid substitutions at a single site. Evidence for a region which constrains receptor activation. J. Biol. Chem. 267:1633-39

58. D’Angelo DD, Eubank JJ, Davis MG, Dorn GW II. 1996. Mutagenic analysis of platelet thromboxane receptor cysteines. J. Biol. Chem. 271:6233-40 ao) Chung H-O, Yang Q, Catt KJ, Arora KK. 1999. Expression and function of the gonadotropin-releasing hormone receptor are dependent on a conserved apoloar amino acid in the third intracellular loop. J. Biol. Chem. 274:35756-62 60. Mukherjee S, Palezewski K, Gurevich VV, Hunzicker-Dunn M. 1999. Beta-arrestindependent desensitization of luteinizing hormone/choriogonadotropin receptor is prevented by a synthetic peptide corresponding to the third intracellular loop of the receptor. J. Biol. Chem. 274:12984-89 61. Tsuga H, Kameyama K, Haga T, Honma T, Lameh

J, et al. 1998.

Internalization

and down-regulation of human muscarinic acetylcholine receptor m2 subtypes. J. Biol. Chem. 273:5323-30

62. Yuan N, Friedman J, Whaley BS, Clark RB. 1994, cAMP-dependent protein kinase and protein kinase C consensus site mutations of the beta-adrenergic receptor. Effect on desensitization and stimulation of adenyl-

yleyclase. J. Biol. Chem. 269:23032-38 63. Hipkin RW, Wang Y, Schonbrunn A. 2000. Protein kinase C activation stimulates the

619

phosphorylation and internalization of the sst2A somatostatin receptor. J. Biol. Chem. 275:5591-99

64. Yu H, Chen JK, Feng S, Dalgarno DC, Brauer AW, et al. 1994. Structural basis for

the binding of proline-rich peptides to SH3 domains. Cell 76:933-45

65. Oldenhof J, Vickery R, Anafi M, Oak J, Ray A, et al. 1998. SH3 binding domains in the dopamine D4 receptor. Biochemistry 37:15726-36 66. Tang Y, Hu LA, Miller WE, Ringstad N, Hall RA, et al. 1999. Identification of

the endophilins (SH3p4/p8/p13) as novel binding partners for the bl-adrenergic receptor. Proc. Natl. Acad. Sci. USA 96:12559-64

67. Blum K, Noble EP, Sheridan PJ, Montgomery A, Ritchie T, et al. 1990. Allelic association of human dopamine D2 receptor gene in alcoholism. JAMA 263:2055—60 68. Smith SS, O’Hara BF, Persico AM, Gorelick DA,

Newlin

DB,

et al. 1992.

Ge-

netic vulnerability to drug abuse. The D2 dopamine receptor Taq I B1 restriction fragment length polymorphism appears more frequently in polysubstance abusers. Arch. Gen. Psychiatry 49:723-72

69. Arinami T, Itokawa M, Enguchi H, Tagaya H, Yano

S, et al. 1994.

Association

of

dopamine D2 receptor molecular variant with schizophrenia. Lancet 343:703-4

70. Higuchi S, Muramatsu T, Murayama M, Hayashida M. 1994. Association of structural polymorphism of the dopamine D2 receptor gene and alcoholism. Biochem. Biophys. Res. Commun. 204:1199-205 TA. Chen CH, Chien SH, Hwu HG. 1996. No association of dopamine D2 receptor molecular variant Cys311 and schizophrenia in Chinese patients. Am. J. Med. Genet. 67:418-20 V2, Itokawa M, Arinami T, Futamura N, Hamaguchi H, Toru M. 1993. A structural polymorphism of human dopamine D2 receptor,

D2(Ser311—-Cys).

Biochem.

Biophys. Res. Commun. 196:1369-75

620

RANA

® SHIINA

® INSEL

3, Gejman PV, Ram A, Gelernter J, Friedman E, Cao Q, et al. 1994. No structural mutation in the dopamine D2 receptor gene in

alcoholism or schizophrenia. Analysis using denaturing gradient gel electrophoresis. JAMA 271:204-8 74. Cravchik A, Sibley DR, Gejman PV. 1996. Functional analysis of the human D> dopamine receptor missense variants. J. Biol. Chem. 271:26013-17 WS. Van Tol HHM, Bunzow JR, Guan H-

82.

83.

C, Sunahara RK, Seeman P, et al. 1991.

Cloning of a human dopamine DR receptor gene with high affinity for the antipsychotic clozapine. Nature 350:610— 14 76. Van Tol HHM. 1998. Structural and functional characteristics of the dopamine D4 receptor. In Catecholamines, ed. DS Goldstein, G Eisenhofer, R McCarty, 42:486—

84.

90. San Diego, CA: Academic

WH. Rasmussen SG, Jensen AD, Liapakis G, Ghanouni P, Javitch JA, Gether U. 1999 Mutation of a highly conserved aspartic acid in the beta2 adrenergic receptor: constitutive activation, structural instability, and conformational rearrangement of transmembrane segment 6. Mol. Pharmacol. 56(1):175—84

78. Rhee MH, Nevo I, Levy R, Vogel Z. 2000. Role of the highly conserved Asp-Arg-Tyr motif in signal transduction of the CB2 cannabinoid receptor. FEBS Lett. 466:300— 4 79% Alewijnse AE, Timmerman H, Jacobs EH, Smit MJ, Roovers E, et al. 2000. The effect of mutations in the DRY motif on the constitutive activity and structural instability of the histamine H(2) receptor. Mol. Pharmacol. 57(5):890-98

80. Mhaouty-Kodja S, Barak LS, Scheer A, Abuin L, Diviani D, et al. 1999. Constitu-

tively active alpha-\b adrenergic receptor mutants display different phosphorylation and internalization features. Mol. Pharmacol. 55:339-47 81. Seibold A, Dagarag M, Birnbaumer M.

.

86.

1998. Mutations of the DRY motif that preserve beta 2-adrenoceptor coupling. Recept. Channels 5(6):375—85 Zhu SZ, Wang SZ, Hu J, el-Fakahany EE. 1994. An arginine residue conserved in most G protein-coupled receptors is essential for the function of the ml muscarinic receptor. Mol. Pharmacol. 45:517—23 Scheer A, Costa T, Fanelli F, De Benedetti PG, Mhaouty-Kodja S, et al. 2000. Mutational analysis of the highly conserved arginine within the Glu/Asp-Arg-Tyr motif of the alpha(1b)-adrenergic receptor: effects on receptor isomerization and activation. Mol. Pharmacol. 57(2):219-31 Rosenthal W, Seibold A, Antaramian A, Gilbert S, Birmbaumer M, et al. 1994. Mutations in the vasopressin V2 receptor gene in families with nephrogenic diabetes insipidus and functional expression of the Q2 mutant. Cell Mol. Biol. 40(3):429-36 Acharya S, Karnik SS. 1996. Modulation of GDP release from transducin by the conserved Glul34-Arg135 sequence in rhodopsin. J. Biol. Chem. 271:25406— 11 Box NF, Wyeth JR, O'Gorman LE, Martin NG, Sturm RA. 1997. Characterization of melanocyte stimulating hormone receptor variant alleles in twins with red hair. Hum. Mol. Genet. 6:1891—97

. Kjelsberg MA, Cotecchia S, Ostrowski J,

Caron MG, Lefkowitz RJ. 1992. Constitutive activation of the alpha 1B-adrenergic receptor by all amino acid substitutions at a single site. Evidence for a re-. gion which constrains receptor activation. J. Biol. Chem. 267:1430-33 . Pei G, Samama P, Lohse M, Wang M, Codina J, et al. 1994. A constitutively active mutant 8>-adrenergic receptor is constitutively desensitized and phosphorylated. Proc. Natl. Acad. Sci. USA 91:2699-702

89. Ren Q, Kurose H, Lefkowitz RJ, Cotecchia S. 1993. Constitutively active mutants of the 6-adrenergic receptor. J. Biol. Chem. 268:16483-87

GPCR POLYMORPHISMS 89a. Morello JP, Salahpour A, Laperriere A, Bernier V, Arthus MEF, et al. 2000.

Pharmacological chaperones rescue cellsurface expression and function of misfolded V2 vasopressin receptor mutants. J. Clin. Invest. 105:887-95

90. Savarese TM, Wang CD, Fraser C. 1992. Site-directed mutagenesis of the rat ml muscarinic acetylcholine receptor. Role of conserved cysteines in receptor function. J. Biol. Chem. 267:11439-48

Oil. Noda K, Saad Y, Graham RM, Karnik SS. 1994. The high affinity state of the beta 2-adrenergic receptor requires unique interaction between conserved and nonconserved extraceilular loop cysteines. J. Biol. Chem. 269:6743-52 o> Cook JV, Eidne KA. 1997. An intramolecular disulfide bond between conserved extracellular cysteines in the gonadotropinreleasing hormone receptor is essential for binding and activation. Endocrinology 138:2800-6

O83; Schiilein R, Ziihlke K, Oksche A, Hermosilla

R, Furkert

J, et al. 2000.

The

role of conserved extracellular cysteine residues in vasopressin V2 receptor function and properties of two naturally occurring mutant receptors with additional extracellular cysteine residues. FEBS Lett. 466:101-6 94. Chhajlani V, Xu X, Blauw J, Sudarshi S. 1996. Identification of ligand binding residues in extracellular loops of the melanocortin 1 receptor. Biochem. Biophys. Res. Commun. 219:521-25

Ont Samson M, LaRosa G, Libert F, Paindavoine P, Detheux M, et al. 1997. The

second extracellular loop of CCRS is the major determinant of ligand specificity. J. Biol. Chem. 272:24934-41 96. Hoffmann C, Moro S, Nicholas RA,

621

. Cotte N, Balestre MN, Phalipou S, Hib-

ert M, Manning M, et al. 1998. Identification of residues responsible for the selective binding of peptide antagonists and agonists in the V2 vasopressin receptor. J. Biol. Chem. 273:29462-68

98 . Bouvier M, Hausdorff WP, De Blasi A, O’ Dowd

BF, Kobilka BF et al. 1988. Re-

moval of phosphorylation sites from the beta 2-adrenergic receptor delays onset of agonist-promoted desensitization. Nature 333:370-73 99 . Seibold A, Bridgette G, Friedman J, Hipkin W, Clark RB. 1998. Desensitization of the beta 2 adrenergic receptors with mutations of the proposed G protein-coupled receptor kinase phosphorylation sites. J. Biol. Chem. 273:7637-42 100. Lezcano N, Mrzljak L, Eubanks S, Levenson R, Goldman-Rakic

P, Bergson C.

2000. Dual signaling regulated by calcyon, a D1 dopamine receptor interacting protein. Science 287:1660—64

101. Koranau H, Seeburg P, Kennedy M. 1997. Interaction of ion channels and receptors with PDZ domain proteins. Curr. Opin. Neurobiol. 7:368—73 102. Brakeman PR, Lanahan AA, O’Brien R, Roche

K, Barnes

CA,

et al. 1997.

Homer: a protein that selectively binds metabotropic glutamate receptors. Nature 386:284-88 103. Janssens R, Paindavoine P, Parmentier M, Boeynaems

J-M.

1999.

Human

P2Y> receptor polymorphism: identification and pharmacological characterization of two allelic variants. Br. J. Pharmacol. 126:709-16

104. Shibata N,

K,

Kawabe

Hirasawa A, Moriyama Ogawa S, Tsujimoto

K,

of amino acids in extracellular loops of the

G. 1996. AlA-adrenoreceptor polymorphism: pharmacological characterization and association with benign prostatic hypertrophy. Br J. Pharm. 118:1403-8

human P2Y, receptor in surface expres-

105. Tanaka H, Moroi K, Iwai J, Takahashi

sion and activation process. J. Biol. Chem. 274:14639-47

tions of the endothelin B receptor gene

Harden TK, Jacobson KA. 1999. The role

H, Ohnuma N, et al. 1998. Novel muta-

622

RANA

#® SHIINA

® INSEL

in patients with Hirschsprung’s disease and their characterization. J. Biol. Chem. 273(18):11378-83

106. Horstmeyer Miiller-Esterl

A, W,

Cramer

H,

Schroeder

Sauer C.

T,

1996.

Palmitoylation of endothelin receptor A. Differential modulation of signal transduction activity by post-translational modification. J. Biol. Chem. 271:20811—

19 107. Eason MG, Liggett SB. 1996. Chimeric mutagenesis of putative G protein coupling domains of the alpha2A-adrenergic receptor. Localization of two redundant and fully competent Gi coupling domains. J. Biol. Chem. 271:12826-32 108. Mason DA, Moore JD, Green SA, Liggett SB. 1999. A gain-of-function polymorphism in a G protein coupling domain of the human f ,-adrenergic receptor. J. Biol. Chem. 274(18):12670-74 109. Arranz MJ, Munro J, Owen MJ, Spurlock G, Sham PC, et al. 1998. Evidence for association between polymorphisms in the promoter and coding regions of the 5-HT2A receptor gene and response to clozapine. Mol. Psychiatry 3:61—66 110. Whaley BS, Yuan N, Birnbaumer L, Clark RB, Barber R. 1994. Differential expression of the beta-adrenergic receptor modifies agonist stimulation of adenylyl cyclase: a quantitative evaluation. Mol. Pharmacol. 45:481—89 111. Chemelli RM, Willie JT, Sinton CM, Elmquist JK, Scammell T, et al. 1999. Narcolepsy in orexin knockout mice: molecular genetics of sleep regulation. Cell 98(4):437-S1 V2, Lin L, Faraco J, Li R, Kadotani H, Rogers W, et al. 1999. The sleep disorder canine narcolepsy is caused by a mutation in the hypocretin (Orexin) receptor 2 gene. Cell 98:365-76 3. Niswender CM, Copeland SC, HerrickDavis K, Emeson RB, Sanders-Bush E. 1999. RNA editing of the human serotonin 5-hydroxytryptamine 2C recep-

tor silences constitutive activity. J. Biol. Chem. 274:9472-78

114. Okuyama Y, Ishiguro H, Toru M, Arinami T. 1999. A genetic polymorphism in the promoter region of DRD4 associated with expression and schizophrenia. Biochem. Biophys. Res. Commun. 258:292-95 bis: Arinami T, Gao M, Hamaguchi H, Toru M. 1997. A functional polymorphism in the promoter region of the dopamine D4 receptor gene is associated with schizophrenia. Hum. Mol. Genet. 6:577— 82 116. Breen G, Brown J, Maude S, Fox H, Collier D, et al. 1999. —141 C del/ins polymorphism of the dopamine receptor 2 gene is associated with schizophrenia in a British population. Am. J. Med. Genet. 88:407-10 BW’ Laruelle M, Abi-Dargham A, Casanova MF, Toti R, Weinberger DR, et al. 1993. Selective abnormalities of prefrontal serotonergic receptors in schizophrenia. A postmortem study. Arch. Gen. Psychiatry 50:810-18 118. Ohara K, Nagai M, Tani K, Tsukamoto T, Ohara K. 1999. Schizophrenia and the serotonin-2A receptor promoter polymorphism. Psychiatr. Res. 85:221-24 119. Collier DA, Arranz MJ, Li T, Mupita D, Brown N, et al. 1997. Association between 5-HT2A gene promoter polymorphism and anorexia nervosa. Lancet 350:412 . Shimron-Abarbanell D, Erdmann J, Vogt

ike

IR, Bryant SP, Spurr NK, et al. 1997. Human 5-HTSA receptor gene: systematic screening for DNA sequence variation and linkage mapping on chromosome 7q34-q36 using a polymorphism in the 5’ untranslated region. Biochem. Biophys. Res. Commun. 233:6-9 Sivagnanasundaram §, Morris AG, Gaitonde EJ, McKenna PJ, Mollon JD, et al. 2000. A cluster of single nucleotide

polymorphisms in the 5’-leader of the human dopamine D3 receptor gene (DRD3)

GPCR POLYMORPHISMS and its relationship to schizophrenia. Neurosci. Lett. 279:13-16

122, Erdmann J, Hegemann N, Weidemann A, Kallisch H, Hummel M, et al. 1998. Screening the human bradykinin B2 receptor gene in patients with cardiovascular diseases: identification of a functional mutation in the promoter and a new coding variant (T21M). Am. J. Med. Genet. 80:521-25 123. McDermott DH, Zimmerman PA, Guignard F, Kleeberger CA, Leitman SF etal.

1998. CCRS promoter polymorphism and HIV-1 disease progression. Multicenter AIDS Cohort Study (MACS). Lancet 352(9131):866—70

124. Bonnardeaux A, Davies E, Jeunemaitre X, Fery I, Charru A, et al. 1994. Angiotensin II type 1 receptor gene polymorphisms in human essential hypertension. Hypertension 24:63-69 125: Wang WY, Zee RY, Morris BJ. 1997. Association of angiotensin [I type 1 receptor gene polymorphism with essential hypertension. Clin. Genet. 51:3134 126. Szombathy T, Szalai C, Katalin B, Palicz T, Romics L, et al. 1998. Association of

angiotensin II type 1 receptor polymorphism with resistant essential hypertension. Clin. Chim. Acta 269:91—100 126a. Miller JA, Thai K, Scholey JW. 1999. Angiotensis II type | receptor gene polymorphism predicts response to losartan and angiotensis II. Kidney Int. 56:217380 PLT. Szombathy T, Janoskuti L, Szalai C, Csaszar A, Miklosi M, et al. 2000. Angiotensin II type 1 receptor gene polymorphism and mitral valve prolapse syndrome. Am. Heart J. 139:101-5 128. Alvarez R, Reguero JR, Batalla A, Iglesias-Cubero G, Cortina A, et al. 1998. Angiotensin-converting enzyme and angiotensin II receptor 1 polymorphisms: association with early coronary disease. Cardiovasc. Res. 40:375-79

623

129) Kozak M. 1997. Recognition of AUG and alternative initiator codons is augmented by G in position +4 but is not generally affected by the nucleotides in positions +5 and +6. EMBO J. 16:2482-92 130. Parola AL, Kobilka BK. 1994. The peptide product of a 5’ leader cistron in the beta 2 adrenergic receptor mRNA inhibits receptor synthesis. J. Biol. Chem. 269:4497-505 1810 McGraw DW, Forbes SL, Kramer LA, Liggett SB. 1998. Polymorphisms of the 5’ leader cistron of the human beta2adrenergic receptor regulate receptor expression. J. Clin. Invest. 102:1927—32 132) Scott MG, Swan C, Wheatley AP, Hall IP. 1999. Identification of novel polymorphisms within the promoter region of the human beta2 adrenergic receptor gene. Br. J. Pharmacol. 126:841—44 1333 Yamada K, Ishiyama-Shigemoto S, Ichikawa F, Yuan X, Koyanagi A. 1999. Polymorphism in the 5/-leader cistron of the beta2-adrenergic receptor gene associated with obesity and type 2 diabetes. J. Clin. Endocrinol. Metab. 84:1754-57 134. Grandy DK, Marchionni MA, Makam H,

Stofko RE, Alfano M, et al. 1989. Cloning of the cDNA and gene for a human D2 dopamine receptor. Proc. Natl. Acad. Sci. USA 86:9762-66 139): Oda Y, Tu CL, Pillai S, Bikle DD. 1998. The calcium sensing receptor and its alternatively spliced form in keratinocyte differentiation. J. Biol. Chem. 273:23344-52 136. Elshourbagy NA, Adamou JE, Gagnon AW, Wu HL, Pullen M, et al. 1996. Molec-

Sie

ular characterization of a novel human endothelin receptor splice variant. J. Biol. Chem. 271:25300-7 Claeysen S, Sebben M, Becamel C, Bockaert J, Dumuis

A.

1999.

Novel

brain-

specific 5-HT4 receptor splice variants show marked constitutive activity: role of the C-terminal intracellular domain. Mol. Pharmacol. 55(5):910—20

624

RANA

® SHIINA

® INSEL

138. Bockaert J, Claeysen S, Sebben M, Dumuis A. 1998. 5-HT4 receptors: gene, transduction and effects on olfactory memory. Ann. NY Acad. Sci. 861:1—15 ISI), Vaisse C, Clement K, Guy-Grand B, Froguel P. 1998. A frameshift mutation in human MCAR is associated with a dominant form of obesity. Nat. Genet. 20:11314 140. Yeo GSH, Faroogi ISF, Aminian S, Halsall DJ, Stanhope RG, O’Rahilly S. 1998. A frameshift mutation in MC4R associated with dominantly inherited human obesity. Nat. Genet. 20:111-12 141. Samson M, Libert F, Doranz BJ, Rucker J, Liesnard C, et al. 1996. Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene. Nature 382:722-25 142. Lu Y, Nerurkar VR, Dashwood WM, Woodward CL, Ablan S, et al. 1999. Genotype and allele frequency of a 32base pair deletion mutation in the CCRS gene in various ethnic groups: absence of mutation among Asians and Pacific Islanders. Int. J. Infect. Dis. 3:186—91

143. Hall IP, Wheatley A, Christie G, McDougall C, Hubbard R, et al. 1999. Association of CCRS delta32 with reduced risk of asthma. Lancet 354:1264—65 144. Chang FM, Kidd JR, Livak KJ, Pakstis AJ, Kidd KK. 1996. The world-wide distribution of allele frequencies at the human dopamine D4 receptor locus. Hum. Genet. 98:91-101 145.

Xie HG, Stein CM, Kim RB, Xiao ZS, He

N, et al. 1999. Frequency of functionally important beta-2 adrenoceptor polymorphisms varies markedly among AfricanAmerican, Caucasian and Chinese individuals. Pharmacogenetics 9:511—16

146. Rana BK, Hewett-Emmett D, Jin L, Chang BH-J, Sambuughin N, et al. 1999. High polymorphism at the human melanocortin | receptor locus. Genetics 151:1547-57 147.

Rutledge

DR.

1991.

Are

there

beta-

adrenergic receptor response differences between racial groups? DICP 25:824-34 148. Mills PJ, Dimsdale JE, Ziegler MG, Nelesen RA. 1995. Racial differences in epinephrine and beta 2-adrenergic receptors. Hypertension 25:88-91

NOTE ADDED IN PROOF

A recent study by Morello et al (89a) introduces a new therapeutic pathway for

disease caused by mutations in the IL domains. The A62-64 mutant of the V, receptor, found in NDI patients, is retained intracellularly in the ER due to improper protein folding and maturation. Treatment with cell permeable, nonpeptide antagonists partially rescued cell surface expression of this mutant receptor, and

these receptors coupled to adenylyl cyclase with the same efficiency as the wild type receptors. The authors reported that they were also able to partially rescue mutants of the TM domains that exhibit defective binding due to improper cell surface expression. This work suggests a novel therapeutic approach in which cell permeable antagonists would act as pharmacological chaperones to stabilize improperly folded mutant receptors and enable their release from the ER, thus allowing the mutant receptors to resume proper function.

Annu. Rey. Pharmacol. Toxicol. 2001. 41:625—59 Copyright © 2001 by Annual Reviews. All rights reserved

DruG TREATMENT EFFECTS ON DISEASE PROGRESSION PLS Chan and NHG Holford Division of Pharmacology and Clinical Pharmacology, School of Medicine, University of Auckland, Private Bag 92019, Auckland 1030, New Zealand; e-mail: p.chan@ auckland.ac.nz, n.holford@ auckland.ac.nz

Key Words disease progress models, symptomatic, protective, Alzheimer’s, Parkinson’s, osteoporosis @ Abstract Degenerative diseases are characterized by a worsening of disease status over time. The rate of deterioration is determined by the natural rate of progression of the disease and by the effect of drug treatments. A goal of drug treatment is to slow disease progression. Drug treatments can be categorized as symptomatic or protective. Symptomatic treatments do not affect the rate of disease progression whereas protective treatments have the ability to slow disease progression down. Many current methods for describing disease progression have two common drawbacks: a linear relationship between time and disease status is assumed, and within- and between-subject variability is ignored. Disease progress models combined with pharmacokineticpharmacodynamic models and hierarchical random effects statistical models provide insights into understanding the time course and management of degenerative disease.

DEFINITION OF DISEASE PROGRESSION Clinical pharmacology can be defined in terms of disease progression and drug action. Disease progression can be defined in terms of changes in disease status as a function of time. Drug action reflects the effect of a drug on disease status. For example, in degenerative disorders such as Parkinson’s disease, natural disease progression is caused by a continuous degeneration of neurons, which is reflected in such disease status measures as the Unified Parkinson’s Disease Rating Scale (UPDRS). In other diseases, such as diabetic neuropathy and nephropathy, natural disease progression is caused by a loss of nerve or kidney function, and status can be defined by nerve conduction velocity or creatinine clearance. Regarding drug effects on disease, there are two main possibilities. Drugs may provide symptomatic benefit without influencing the underlying progression of the disease, or they may influence the underlying time course of progression. The goal of drug treatments in degenerative disorders is not only to relieve clinical symptoms, but also to slow disease progression. 0362-1642/01/0421-0625$14.00

625

626

CHAN

® HOLFORD

The aim of this review is to describe models for disease progression in degenerative diseases and to define the methods and biomarkers that have been used for studying disease progression. We illustrate these models by distinguishing the symptomatic and protective components of drug effects in Alzheimer’s disease, Parkinson’s disease, osteoporosis, diabetic nephropathy, and respiratory disease.

COMPONENTS OF DISEASE PROGRESSION

Natural Disease Progression Cell death and gradual loss of organ function are well-known natural phenomena of aging. Whether the occurrence of degenerative diseases is age related has been questioned (1—5). According to prevalence statistics, the answer is positive, as a higher incidence is found in advanced age groups (6—11). However, aging alone is not sufficient to explain the full story of the occurrence of degenerative diseases. This is, firstly, because the pattern of cell loss in normal aging has been found to be different from the pattern observed in such degenerative diseases as Parkinson’s and Alzheimer’s diseases (12, 13). For example, maximal losses were found in the ventral tier of the substantia nigra in Parkinson’s disease rather than in the dorsal tier in normal aging (12). Secondly, the rate of cell loss has been found to be faster in diseases than in normal aging. For example, the rate of loss of pigmented neurons in the substantia nigra was 4.7% per decade in normal aging compared with a 45% loss in the first decade in parkinsonian patients (12). This implies that natural disease progression in degenerative diseases can only be studied in patients not receiving drug treatment. In other words, the use of healthy subjects as a control group may not be appropriate in studying disease progression in degenerative disorders.

Natural Disease Progress Models Linear Model

Figure | illustrates a linear pattern of natural disease progression. S(E) = S80 + a- 7,

iig

A linear natural history model describes a constant rate of deterioration of disease status. The rate of disease progression solely depends on the slope (@), whereas the baseline disease status is defined by the parameter SO. Many studies assume a linear rate of disease progression because of the convenience of data analysis (14-18).

Asymptotic Model The rate of change of disease status may vary with disease severity and duration of disease. In this case, disease progression is not simply explained by a linear model. For example, using the UPDRS bradykinesia score as a biomarker for disease severity, nonlinear disease progression was found in Parkinson’s disease patients with prior treatment with levodopa/carbidopa and/or

DISEASE PROGRESS AND DRUG ACTION

40 -

: oa

30 J

Sr

Natural History

ve

ecrse Symptomatic

Protective

y

Symptomatic + NO(2)

627

we

Protective

|eS

eo

i,

Hs;

;

e ne

7

Disease Status

104

0

2

4

6

8

10

12

14

Time

Figure 1 Linear disease progress model and drug modifications. Treatment starts at 2 and stops at 10 time units.

bromocriptine (19). Figure 2 illustrates an asymptotic pattern of natural disease progression.

S@) = S0.e0 7 *+Sss-(1—e 7 *),

2

An asymptotic natural history model describes a worsening of disease status with an exponential time course approaching a steady state. The rate of disease progression depends on the progression half-life (TP) whereas the steady state depends on the maximum “burnt-out” disease status (Sss). Both the linear and asymptotic models represent the possible natural history of disease progression without drug modification. However, these natural disease progress models can be modified by drug treatments, and the modification depends on the type of treatment. In general, each parameter in a disease progress model is a target for describing drug action.

Drug Modifications Classification of Treatments

When describing the beneficial effects of drug therapy, treatments may be categorized into two classes, symptomatic and protective. Protective treatments can slow

628

CHAN

40

® HOLFORD

ol.

|.

pend eae

Sse eee

30 __.. Natural History

8 ern Bo

=

Ngee

Symptomatic

ae

Protective TP : .

Protective Sss 5

a p

|

Protective TP +

Sss es

SY INptomatic x

Protective

0+—

=a]

—

:

=

oe

|

Time

Figure 2 Asymptotic disease progress model and drug modifications. Treatment starts at 2 and stops at 10 time units. TP, progression half-life; Sss, maximum burnt-out disease status.

down, halt, or even reverse disease progress. Symptomatic treatments can only reduce symptom severity. A treatment may have both symptomatic and protective benefits, but distinguishing one from the other may be difficult, as the dominant effect is more likely to be expressed and thus mask the subdominant effect. Separation of symptomatic and protective actions may be possible if the time course of onset of these effects is sufficiently different. Symptomatic effects typically come on more rapidly whereas protective effects take a longer time before they are manifest. The categorization of symptomatic and protective is primarily applicable to the beneficial effects of drug treatments. If a drug has an adverse effect, this may be reflected as an offset in the disease status marker or a change in the rate of progression, as with a beneficial effect. If a drug effect modifies the rate of disease progression adversely, it might be described as accelerating disease progression (the opposite of a protective mechanism). When drug effects are described in terms of their effects on the parameters of a disease progression model, it provides a clear and unambiguous definition to

support the claim of different types of drug effect. A change in a disease progress parameter that does not change the rate of progression is a symptomatic effect. An improvement in the rate of progression is a protective effect.

DISEASE PROGRESS AND DRUG ACTION

629

Mechanisms of Action Selegiline and Tocopherol In Parkinson’s disease, two treatments, selegiline and tocopherol, have been suggested as having primarily protective benefits. Selegiline is amonoamine oxidase inhibitor. Its effect is partly due to inhibition of monoamin e oxidase B, with the assumption that this leads to decreased formation of free radicals, such as the hydroxy] radical. Tocopherol is an antioxidant vitamin. Its protective effect is based on the idea of trapping free radicals and thus reducing the degradation of neurons. However, no study has provided definitive support for the protective effects of either selegiline or tocopherol (15, 20-22).

Angiotensin-Converting Enzyme Angiotensin-converting enzyme (ACE) inhibitors have also been shown to have protective effects by slowing the decline of renal function in diabetic nephropathy (17, 23-25). The mechanism of renal protective effect of ACE inhibitors is still not clear. It has been thought that the protective effect of ACE inhibitors is due to the result of antagonizing the effects of a potent vasoconstrictor, angiotensin II, by inhibiting its formation from angiotensin I (23, 26). The disturbance of the renin-angiotensin system by ACE inhibitors results in retaining the balance between the vasoconstrictive and saltand fluid-retentive properties of angiotensin II. The possible mechanism of renal protective effects of ACE inhibitors has been reviewed elsewhere (27).

Time Course of Drug Effects on Disease Progression Symptomatic Effects In Figures 1 and 2, the “symptomatic” effects in both linear and asymptotic disease progress models demonstrate an improvement of disease status while treatment is given. Because there is no change in the underlying process, the drug benefit simply delays the time until the disease reaches the state observed at the start of treatment, e.g. the benefit of tacrine in Alzheimer’s disease is a delay of about 6 months (28). When treatment is stopped, the ben-

eficial symptomatic effect disappears and the same deterioration pattern as the natural disease progression is followed. The disease progress model parameters, such as the slope (a) of the linear model, the disease progression half-life (TP), and the maximum burnt-out disease status (Sss) in the asymptotic model, remain

unchanged. Irrespective of the function used to describe the time course of the disease, symptomatic treatment can be modeled as if it was a function of the baseline disease state parameter, SO.

Protective Effects Protective drug effects describe modifications of the time course of natural disease progression. With the linear model, the protective effect is refiected in a change of the slope of the natural disease progress model. With the asymptotic model, there are three possible variants. The progression state model represents treatments that have an effect on TP. This is reflected in a change of the curvature of the natural disease progress model. The asymptotic state model represents treatments that have an ability to alter Sss.

630

CHAN

® HOLFORD

Treatments that have protective effects on both mechanisms Figure 2.

are illustrated in

METHODS FOR MEASURING DISEASE STATUS Continuous Scale Markers A necessary requirement for studying disease progression is a biomarker (clinical or biochemical) that can relate clinical observations to disease status. Preferably, such a biomarker is easily measured on a repeated basis and is expressed on a continuous scale: for example, creatinine clearance as an index of renal function, velocity of nerve conduction as a marker for diabetic neuropathy, bone mineral density as an index for osteoporosis, and FEV1 (force expiratory volume in | s) as a marker for obstructive lung disease.

Categorical Rating Scales A number of categorical rating scales have been used to describe disease status in such neurodegenerative diseases as Parkinson’s and Alzheimer’s diseases. Each of these rating scales has different components (cognitive, mental, motor, and activity of daily living) to assess the functional condition of patients. The most widely used scales are the Unified Parkinson’s Disease Rating Scale (UPDRS) and Hoehn and Yahr scale (H&Y) in Parkinson’s disease and Mini Mental State (MMSE) and Alzheimer’s Disease Assessment Scale (ADAS) in

Examination

Alzheimer’s disease. Table 1| lists some of the available rating scales for measuring disease severity in neurodegenerative diseases. Because each of the rating scales is constructed differently, the range of scores is different from one to another. This makes it difficult to compare the results of one rating scale with another. In this case, changes expressed in percentage of baseline rather than in absolute scores may be used to compare different rating scales.

Positron Emission Tomography and Single Photon Emission Tomography Positron emission tomography (PET) and single photon emission tomography (SPECT) are quantitative techniques employed to localize and measure physiologic and biochemical processes in the brain. By following the same pharmacological pathway as intrinsic neurotransmitters, radioactive markers can be used to examine the native neural system in different regions of the brain. With different tracers, PET can differentiate between diseased and normal brain, as well as between diseases with similar clinical symptoms (29-31). In Alzheimer’s disease,

a 12%-24% reduction of regional cerebral glucose metabolism (compared with healthy control subjects) has been found (32, 33). SPECT is commonly used for estimating blood flow and receptor binding, as its marker does not depend on

DISEASE PROGRESS AND DRUG ACTION

TABLE 1

631

Common rating scales for assessing disease severity in neurodegenerative disease

Scale

Abbreviation

Component

Range

CURS — MCS H&Y HSD NYUPDS

— — — — — -—

0-128 0-220 0-100 I-V 0-53 0-20

Parkinson’s disease

Columbia University Rating Scale Cornell Weighted Scale Modified Columbia Scale® Hoehn & Yahr Hamilton Scale for Depression New York University Parkinson’s Disease Scale Northwestern University Disability Scale Schwab & England Activities of Daily Living Scale? University of California Los Angeles Scale Unified Parkinson’s Disease Rating Scale

NUDS S&E ADL

— —

0-100 0-100

UCLA UPDRS

Webster Rating Scale

WRS

— Total ADL Mental Bradykinesia Motor —

0-220 0-188 0-52 0-16 0-24 0-108 0-30

BIMC

Total Noncognitive Cognitive Total ADL Cognitive —

0-120 0-50 0-70 0-27 0-16 0-17 0-33

BRSD CDR

Total —

0-164 0-3

CIBIC

—

1-7

Alzheimer’s disease Alzheimer’s Disease Assessment Scale

Blessed Dementia Scale

Blessed Information Memory Concentration Behavior Rating Scale for Dementia Clinical Dementia Rating (in six categories) Clinician’s Interview-Based Impression of Change

ADAS ADASC BDS

Sum of Boxes (Global CDR)

CDR-SB

—

0-18

Dementia Rating Scale Extended Scale for Dementia Global Deterioration Scale

DRS ESD GDS

= — —

0-144 0-250 Qe7/

Mini Mental State Examination‘

MMSE

——

0-30

Progressive Deterioration Scale

PDS

—~

0-100

Severe Impairment Battery

SIB

_

0-100

“Modification of Columbia University Rating Scale. >The scale is in percentage, with no disability 100%. ‘Higher scores indicate less impairment.

632

CHAN

® HOLFORD

dopamine turnover. A 30%-56% reduction in striatal uptake of tracer (Vy) was reported in Parkinson’s disease (34-36). It should be noted that different tracers

might generate different uptake rates because of differences in distribution and elimination processes (36). Recently, PET has been used as a tool for detection of preclinical Parkinson’s disease (37, 38) and determination of rate of disease progression (39-41). The

uptake rate constant (K;) can be taken as a distribution rate constant that describes the rate of tracer storage in neurons. Because radioactive tracer is being taken up by the surviving neurons in the brain, K; can be used as a marker for the number of functioning neurons. It has been shown that K; correlates well to the number of surviving nigral pigmented neurons in Parkinson’s disease (42). Moreover, it has also been shown that K; correlates well with clinical markers such as the UPDRS (41) and the H&Y scale (43). Consequently, K; could be used as a marker for assess-

ing disease progression in neurodegenerative disorders. A correlation between V¥ and UPDRS has also been shown (34). Both PET and SPECT have a high reproduci-

bility (44, 45). With the application of PET, disease progression and the effect of drugs can be measured by determining the change of K;, over time (46, 47). In practical terms, PET and SPECT are time-consuming and expensive screening methods. Because of these reasons, the change of K; or V3 is often computed based on two observations. The assumption of a linear rate of loss of neurons is one of the limitations of using changes in K; as a measure of disease progression. This limitation may be overcome by taking more observations over a longer interval.

METHODS FOR DESCRIBING DISEASE PROGRESSION There have been many reports of the longitudinal change of disease status in degenerative diseases. However, few have attempted to explicitly quantify the rate of disease progression. Generally, there are several methods of dealing with longitudinal data.

ANOVA/ANCOVA Frequently, the treatment effect on an outcome measure is determined by simple statistics (parametric or nonparametric) or through the application of analysis of variance (ANOVA) or analysis of covariance (ANCOVA). The purpose of ANOVA is to test for significant differences between the means of the control and treatment groups. The rate of disease progression in either the control or the treatment group is not taken into account by this method.

Survival Analysis Survival analysis is the use of endpoints, for example death or the need for additional treatment, as an objective to measure the fraction of patients reaching the endpoint over time. Kaplan-Meier analysis is a common approach to interpreting the outcome using survival analysis.

DISEASE PROGRESS AND DRUG ACTION

633

Change from Baseline Change from baseline analysis uses two observations to determine the rate of disease progression. The baseline and the final observations are used, and the rate of progression is determined from the change in the two outcome measures divided by the length difference in the two time points. This is also known as two-point analysis.

Linear and Nonlinear Modeling Modeling is the use of mathematical functions to describe quantitative relationships, e.g. time and disease status, through linear or nonlinear regression. The power of modeling is that it not only describes the data, it also predicts and explains the time course and drug effect beyond the study period. Pharmacokineticpharmacodynamic models relate plasma drug concentrations to clinical responses (48). Parameter estimations can be performed under individual- or populationbased approaches. NONMEM (nonlinear mixed effect model) is a program that allows model building and parameter estimation using a population approach (49). A key feature of population analysis is the ability to account for and describe withinand between-subject variability. Another advantage of modeling is the ability to take into consideration the effects of covariates when estimating parameters. The rate of disease progression depends on the disease status scale used to calculate it. Table 2 lists the natural rate of Alzheimer’s disease progression with different scales and analyses (14, 28, 50-65). In one study, Stern et al (57) has shown that the rate of disease progression varied from 3.9 to 5.2 points/year in patients with Alzheimer’s disease with Blessed test of information, memory, and concentration (BIMC) as a marker for assessing disease severity. Studies of short duration that assume a linear model may overestimate the rate of disease progression if the progression model is actually asymptotic. A high patient drop-out rate is also responsible for the imprecision in estimating the rate of change. None of the studies has taken into account the influence of covariates, such as age and duration of symptoms, in determining the rate of disease progression. More important, these methods lack the ability to determine the within- and between-subject variability.

RATE OF DISEASE PROGRESSION

Changes in Pharmacokinetics and Pharmacodynamics in Parkinson’s Disease Several studies have compared the pharmacokinetics and pharmacodynamics of patients with different stages of Parkinson’s disease (66-69). These studies aimed to find out how the time course of levodopa effects might be modified as Parkinson’s disease progresses (Table 3). Contin et al (70—72) have performed several longitudinal studies to investigate the change of pharmacokinetics and pharmacodynamics

634

CHAN

#® HOLFORD

analysis TABLE 2. Rate of disease progress in Alzheimer’s disease with different methods ae and biomarkers* COCs $e a Rate of progression Reference

Method

Scale

Baseline (points)

(points/year)

(%/year)

50

Two point

MMSE

17.20

2.20

1279

51

Two point

MMSE

16.47

4.18

25.39

Sy

Two point

MMSE

17.40

2.81

16.15

53

Two point

MMSE

10.00

3.50

35.00

54

Two point

MMSE

18.70

3.90

20.86

55

Two point

MMSE

11.10

4.30

38.74

56°

Linear

MMSE

17.90

0.62

3.46

4.10

—-

Si

MIM

BIMC

oo

58

Two point

BIMC

137

4.40

33.41

17.40

4.50

25.86

59

Two point

BIMC

52

Two point

BIMC

16.60

3.24

19.52

—_—

3.90

—

57

Two point

BIMC

60

Two point

BIMC

17.10

2.60

15.20

61

Linear

BIMC

—

4.10

_—

57

Linear

BIMC

_

4.00

_

51

Two point

ADAS

22.40

8.28

36.96

62>

Two point

ADASC

29.60

PAS Wy|

9.36

56°

Linear

ADASC — 28.50

5.88

20.63

28

Linear

ADASC — 28.70

6.17

21.50

63

Linear

ADASC — 28.40

5.00

17.61

64

Linear

ADASC

—

6.29

os

63

Linear

CIBIC

4.00

0.61

1525

64

Linear

CIBIC

0.69

_-

65

MIM

PSMS

12.76

2.44

19/12

65

MIM

IADLS

22-32

2.06

9.23

60

Two point

BDS

17.50

3.50

20.00

14°

Two point

BDS

20.70

7.56

36.52

55

Two point

SIB

79.10

17.10

21.62

2

Two point

DRS

98.30

11.38

11.58

56

Linear

PDS

46.70

13.00

27.84

"MIM, multiple interval method. For other abbreviations, see Table 1. >Rate of progression converted from points/week.

“Rate of progression converted from points/month.

DISEASE PROGRESS AND DRUG ACTION

635

TABLE 3 Pharmacodynamic comparisons in patients with different disease stage of Parkinson’s disease* eee

Reference

66°

Parameter

Levodopa naive

EO (taps/min) Max change from

Stable

Fluctuating

10748

9347

DOI=

49+8

144+ 25 56 + 28 228841499 14+0.8

106 + 23 SEEN 211041420 P3EOI8

166 + 44 AD SSE 18.3.

153 +44 515i: 25

Fluctuating + Peak dose dyskinesia

EO (taps/min)

67

EO (taps/min) Eynax (taps/min) ECs, (ng/ml)° Hill (U)

68

Sopra Ae Emax (taps/min)

11649 44 + 34 250441459 Osseo

ECso (ng/ml)?

2404130

6404260

fe32 Iss A Meas ili

IG sise 7/ else Oss

24+ 10 OEE SI 389+ 138

ail)Se 1 Peas 346 + 203

41421 24+ 13 543 + 245

3 0.8140.49

4 1.284050

5 0.39+0.20

Hill (U) Teq (h)* 69

EO (CURS) Ena (CURS) ECs9 (ng/ml) Hill (U) Teq (h)

er

spies I ithe7 Tee Os 6 0.28+0.22

“For abbreviations, see Table |.

Only simple statistical comparisons were made. No pharmacokinetic-pharmacodynamic modeling has been performed. “30% effective concentration (ECs9) converted from nanomoles per milliliter. 4ECso converted from micrograms per milliliter.

“Equilibration half-life (Teq) converted from minutes.

over time (Table 4). According to these findings, changes in pharmacodynamic parameters appeared after 3-4 years of levodopa treatments. Nutt & Holford (73) used a pharmacokinetic-pharmacodynamic approach to explain the transition from the stable to the fluctuating response state in Parkinson’s disease. They argued that a change in sensitivity (50% effective concentration) could not account for differences in the time course of the acute response to levodopa as the disease progressed. A shortening of the delay between changes in plasma concentration and subsequent changes in response, describable by differences in the equilibration half-life (Teq), was most likely the reason for the altered response in the fluctuating state.

Natural Rate of Disease Progression Parkinson’s Disease The first study looking at disease progression in patients with Parkinson’s disease was conducted in 1967 (74). In this study, the rate of progression was investigated

636

CHAN

TABLE 4

® HOLFORD

Changes in pharmacodynamic parameters over time in Parkinson’s disease“

Reference

Parameter

Baseline

Final

% Change?

Study duration

70

ECsg (ng/ml) Hill (U)

370450 SOS

580 + 80 = 2hOaas. 0

41+5 163°

4 years

Teq (min)

109+ 19

3648

—55+8

71

ECso (ng/ml) Teq (min)

4204260 62 +57

690+ 200 OVENS

64° —66°

3 years

1P

EO (taps/min) Max change from EO (taps/min)

127435 43 +19

Pasa = PAS; 53 + 24

—4° 78%

4 years

*All 50% effective concentrations (ECs) were converted from micrograms per milliliter. ‘Percentage of change was computed by the following equation: (final-baseline)/baseline.

“100% change.

by looking at the time required for deterioration of one stage of H&Y scale. Marttila & Rinne (75) also performed a similar study with 442 levodopa naive parkinsonian patients. Figure 3 shows the plots of H&Y stage against time. The natural rate of disease progression can also be estimated by looking at placebo groups in studies investigating the effects of drug treatments. With the assumption of linear deterioration, the rate of disease progression in Parkinson’s disease was found to be 13.11—14.02 points/year (UPDRS total) and 3.62—13.4 points/year (UPDRS motor) (Table 5) (15, 20, 76-78). Disease Progression Using PET The application of PET to describe the rate of disease progression has been performed for Parkinson’s disease. The published rates of change in Kj range from 0.4% to 7% of the mean baseline K; in healthy control subjects. Table 6 summarizes the rate of K, progression in Parkinson’s disease (39-41, 79, 80). All studies showed that parkinsonian patients have a smaller K; value than do healthy control subjects. For example, putamen (K;) was found to be 0.0054 min”! and 0.0101 min~! in parkinsonian patients and healthy control. subjects, respectively (79). The rate of change is expressed as percentage of normal mean per year. This is the mean annual deterioration in K;, in the patients expressed as a percentage of the mean K; in the control group at baseline scan. The annual rate of progress varied with the method of analysis. The large range of annual rate of progression between studies indicates the difficulties in applying PET techniques. Besides PET, computed tomographic scans and magnetic resonance imaging (MRI) scan have also been employed to monitor disease progression in Alzheimer’s disease. In comparison with normal aging controls, a decrease in brain volume was found in Alzheimer’s disease (81, 82). Based upon this phenomenon, it has been proposed that rates of change in brain volume could be a marker of disease progression in Alzheimer’s disease. Not surprisingly, the annual rate of change

DISEASE PROGRESS AND DRUG ACTION

637

Y H&

Time [Years] Figure 3

Observed rate of disease progression measured by the time required for deteri-

oration of one stage of Hoehn & Yahr (H&Y) scale: closed diamond (74); closed square

(75). varied with the structural measures (Table 7) (83-91). In general, with MRI scan,

a larger decrease in brain volume was shown in Alzheimer’s disease patients in comparison with the control group. For example, the annual decrease in total brain volume was 2.37%-—2.78% in Alzheimer’s disease in comparison with 0.24%0.41% in the normal group.

Aizheimer’s Disease In Alzheimer’s disease, several studies have explored the natural disease progression by using a multiple-interval method (repeatedly computing change over a specified time interval, i.e. every 6 months), two-point analysis, or linear regression (Table 2). The rate of progression has a large range because of the use of different rating scales and analysis methods (2.77-6.29 ADASC, 2.2-4.3 MMSE,

2.6-4.5 BIMC points/year). The absolute scores are different rating scales used; thus, a plot of percentage shown in Figure 4 (28, 51,52, 56,92, 93). The heavy ease progression predicted by using the progression & Peace (28), Knopman

not comparable because of of change from baseline is lines show the rate of disrates reported by Holford

& Gracon (56), and Yesavage et al (51). The Figure

illustrates the variability in rate of disease progress with different rating scales.

638

CHAN

® HOLFORD

TABLE 5_ Natural and treatment-altered rate of disease progress in Parkinson’s disease with different rating scales as clinical markers* Rate of progression Ref.

—__(Points/year)

(%/Year)

Scale

Selegiline

UPDRS UPDRS

25.4+ 11.6 25S E20

13.114 14.30 Spe bee enIsbe

51.61 21.74

UPDRS UPDRS UPDRS UPDRS

254 253: 25.44 PDs is=

14.02 + 12.32 7.00+ 10.76 15.16+ 16.12 7.28+11.11

55.20 27.67 59.69 28.77

Levodopa Levodopa?

UPDRS UPDRS

20.6+ 10.9 23:62 14

3.8+8.5 2 fel

18.45 5.08

Selegiline

UPDRSm UPDRSm

21.4142.18 21.93 = 1:47

13.40+ 1.82 6.75 + 1.05

62.59 30.78

Selegiline

UPDRSm UPDRSm

16.8+8.8 16.8+8.8

58+ 9.88 4.02 + 8.29

51.07 23.93

UPDRSm UPDRSm UPDRSm UPDRSm

16.8+8.8 16.8+8.8 16.8+8.8 16.8+8.8

3.62 + 3.74 2.66 + 3.22 3.92+4.47 2.51 +3.86

21°55 15.83 23:33 14.94

Selegiline + Sinemet Sinemet Selegiline + bromocriptine Bromocriptine

UPDRSm

14.6+1.5

—1.4+1.0

—9.59

UPDRSm UPDRSm

12.8+1.0 14.2+1.0

SS2210 2A ket

25.78 16.90

UPDRSm

114+1.3

aif at

43.86

Levodopa Levodopa>

UPDRSm UPDRSm

14.2+8.6 16.7+8.8

2.6+6.8 Og 6- 1

18.31 4.19

Selegiline

H&Y H&Y

1.46+0.13 L59 — 0.01574AH FEV1 (males) = 758.5 + 634.9H? — 0.128H3(A — 36.3)? FEV1 (females) = 798.2+517.6H? — 0.136H3(A — 36.7) FEV1 (males) = 0.092H — 0.032A — 1.260 FEV1 (females) = 0.089H — 0.025 — 1.932 FEV1 (females) = 67.6H — 23.0A — 918

EVI (males) = 0.036H — 0.027A — 1.65 FEV1 (females) = 0.025H — 0.022A — 0.62 FEV1 (males) = 0.037H — 0.028A — 1.59

25-74

18278 20-75 20-84 [S271 25-74 18-66

“FEV1, force expiratory volume in | s; A, age (years); M, body mass (kilograms); H, height (meters), Height is in centimeters.

“Height is in inches. 4SEX: 0, males; 1, females.

A hazard function has been used to study disease progression and the effect of selegiline in Parkinson’s disease (116). The hazard function defines the probability of patients reaching an end point at a given point in time. We might expect the hazard of patients requiring levodopa to increase with time in patients not receiving drug treatment, whereas drug therapy may decrease the hazard. In this study, selegiline decreased the hazard in the first 300 days compared with the placebo group. After day 300, the hazard of the placebo group unexpectedly decreased and approached the hazard for the selegiline group, at approximately 530 days. Based upon this finding, the authors suggested that the effect of selegiline is symptomatic rather than protective, but no clear explanation has been proposed for the pattern of hazard in the placebo group.

Respiratory Disease Corticosteroids Inhaled corticosteroids such as budesonide and beclomethasone are used in the management of chronic obstructive pulmonary disease. Table 9 shows the rate of disease progression in respiratory disease and the effect of inhaled corticosteroids (117—119). All studies of the rate of disease progression used a linear model and showed that corticosteroid treatments produced a slower decline in FEV1 in respiratory diseases (range 30-46 ml/year; control range 50-64 ml/year)

646

CHAN

® HOLFORD

ADASC of Change [%] Baseline from

+

T

q

0

0.5

| Time

Bes

2

[Years]

Figure 7 Observed effects of treatments in Alzheimer’s disease using Alzheimer’s Disease Assessment Scale—Cognitive (ADASC) as a marker. Solid lines indicate treatment groups. Closed diamonds, idebenone (90 mg/day) (109); open diamonds, idebenone (270 mg/day) (109); closed squares, donepezil (110); closed triangles, eptastigmine (111); closed circles, tacrine (112); open circles, tacrine + oestrogen (112). Dotted line indicates predicted natural disease progression (28). Dashed line is predicted response to treatment with tacrine (28).

(117-119). Nevertheless, these studies claimed that the difference in rate of decline in FEV 1 was not significantly different between the treatment and the control groups.

Bronchodilator The effects of a smoking intervention and the use of an anticholinergic bronchodilator (ipratropium bromide) in patients with chronic ob-structive pulmonary disease has been studied (120). A 27.6 ml increase in FEV1 was shown in the group receiving ipratropium bromide compared with the placebo group (ipratropium bromide, 38.8 ml; placebo, 11.2 ml). The effect of ipratropium bromide is symptomatic, as the rate of decline in FEV1 was similar between the two groups (ipratropium bromide, 52.7 ml/year; placebo, 52.3 ml/year).

Smoking Effect When comparing the rate of decline in FEV1 between the smoking intervention group (without bronchodilator) and the no-intervention group, a similar rate of decline in FEV1 was seen (no intervention, 56.2 ml/year; smoking intervention, 52.3 ml/year). However, the effect of smoking intervention differed when comparing the rate of decline between sustained quitters and continuing

DISEASE PROGRESS AND DRUG ACTION

647

704

50 4

Change of Parkinson's Baseline from status [%]

Time [Years] Figure 8 Effect of selegiline on natural disease progression in Parkinson’s disease. Dotted line predicts the exponential change in bradykinesia score (derived from Unified Parkinson’s Disease Rating Scale) in patients with prior treatment of levodopa/carbidopa and/or bromocriptine (19). Solid lines indicate selegiline-only treatment groups status at time zero. Closed diamonds, Webster Rating Scale (115); closed triangles, Northwestern University Disability Scale

(115); closed squares, Columbia University Rating Scale (115).

smokers (continuing smokers, 63 ml/year; sustained quitters, 34 ml/year) over the

5-year study period. The slowing down in the decline of FEV1 suggested that smoking cessation has a protective effect similar to a protective drug treatment effect or, conversely, that smoking accelerates the natural progression.

Diabetic

Nephropathy

All studies of the rate of disease progression in diabetic nephropathy have assumed a linear model. Table 10 shows the rate of disease progression in diabetic nephropathy and the treatment effect of ACE inhibitors (17, 23-25, 121-123). ACE inhibitors slow the decline of glomerular filtration rate in diabetic nephropathy (range 0.98—9.2 ml/min/year) compared with the placebo control group (range 4,55—-13.4 ml/min/year). Laffel et al (25) reported an increase of 0.9 ml/min/year in glomerular filtration rate after 2 years of treatment with captopril. The ability to alter the rate of disease progression suggests that ACE inhibitors have a protective drug effect rather than a symptomatic drug effect.

648

CHAN

® HOLFORD

TABLE 9 Rate of disease progress in respiratory disease and effect of drug treatment using FEV1 as a biomarker*

Rate of progression Ref.

Treatment

i

(Liters/year)

(%/Year)

Duration (years) 3

2.39

—0.0496

—2.08

Budesonide

2.36

—0.046

—1.95

3

Beclomethasone

2.29 2.38

—0.064 —0.033

—2.79 —1.39

3 3

— Budesonide Budesonide +

1.9 2.16

—0.060 —0.030

—3.16 —1.39

a 2

prednisolone

1.86

—0.040

oS

ve

co 119

Baseline FEV1 (liters)

"FEV1, force expiratory volume in | s.

Osteoporosis The change in bone mineral density with different drug treatments has been described recently (124). The common drug treatments of osteoporosis can be classified into different groups: hormone replacement therapy such as estrogen; selective estrogen receptor modulators such as tamoxifen; bisphosphonates such as alendroate and pamidronate; and calcium supplementation. Figure 9 shows the effects of different drug treatments on bone mineral density in osteoporosis (125-133). A symptomatic treatment effect rather than a protective effect is seen in studies with trial durations longer than 1 year. Pors Nielsen et al (133) compared TABLE 10 Rate of disease progress in diabetic nephropathy and the treatment effect of ACE inhibitors ———— eee

Baseline GFR®

Rate of progression

(ml/min/1.73 m2)

(ml/min/year)

Duration

Ref.

Treatment

23

Enalapril

46+ 14

—2.0

—4,35

121

—

83

—5.7

—6.87

5

24

Captopril

TQS 84+ 46

—13.4 —9.2

—17.00 —11.00

4 4

17

Captopril

O8=E5

—4.4

—4.48

10

2S

— Captopril

81+3 (EES

—4.9 0.9

—6.05 isi

2 2

122

Lisinopril

67+18>

—0.98

—1.5

6

123.

— Lisinopril

Mul Oictaal 113+ 16

—4,55 = 1.33

—4.1 —1.18

3 3

“GFR, glomerular filtration rate. >Values converted from milliliters per second per 1.73 m?.

(%/Year)

(years)

3

DISEASE PROGRESS AND DRUG ACTION

649

Change BMD ef Baseline from [%]

Time [Years]

Figure9 Effects of symptomatic treatments in osteoporosis using bone mineral density in lumbar spine as a marker. Dotted lines indicate placebo groups. Solid lines indicate treatment groups. Closed diamonds, tamoxifen (126); open squares, raloxifene (60 mg/day with calcium) (128); small closed squares, raloxifene (120 mg/day with calcium) (129); large closed squares, raloxifene (150 mg/day with calcium) (128); x, estrogen/progestin (125); open circle, alendronate (2.5 mg/day)

(130); closed circle, alendronate (5 mg/day) (130); closed triangle, pamidronate (150 mg/day with calcium) (131); +, calcium (500 mg/day) (132); open triangle, calctum (1000 mg/day) (127). Heavy dashed lines, predictions of estrogen/progestin effect using exponential model with (heavier) and without (lighter) linear decline of bone mineral density (133). Light dashed line, prediction of natural disease progression by (133). BMD, bone mineral density.

an exponential model with and without a linear component to describe the change in bone mineral density seen in response to estrogen in postmenopausal women (heavy dotted lines). The linear decline in bone mineral density with no drug treatment is illustrated as a dashed line. A similar study was performed by Hart etal (134), with a follow up period of 10 years. Unfortunately, these authors only present graphs of their model without numerical parameter values. In general, our review of a range of diseases and treatments indicates that the

percentage of change from baseline and the rate of progression has a wide range due to different markers, types of treatment, and duration of study. Generally, the shorter the duration of study, the greater the rate is. This seems likely to be due to symptomatic effects rather than protective effects. The current two-point method of computing rate of progression has a critical limitation, which is the assumption

650

CHAN

® HOLFORD

of a linear change over time. This leads to an inability to distinguish protective drug effects from symptomatic ones. A serious limitation of most models described in the literature is the use of the naive pooled approach, which makes it hard to assess the importance of covariates, such as the duration of drug therapy or age at onset. A population-based approach that accounts for individual trajectories is essential for understanding the differences between individual responses.

FACTORS INFLUENCING RATE OF DISEASE PROGRESSION The variability in predicting individual time course of disease progression may be explained in part by covariates such as age of onset, duration of symptoms, gender, initial disease severity, etc. In this section, two common covariates, age of onset and gender, are discussed.

Age of Onset Several factors have been thought to play a role in determining the rate of disease progression in Parkinson’s disease. They are age, duration of drug treatment, gender, age of onset, and levodopa dosage. Among these factors, age of onset seems to be the most notable. A study done by Diamond et al (135) compared 54 parkinsonian patients grouped according to age of onset. They illustrate an increased rate of progression with increased age of onset by using the University of California Los Angeles Scale (UCLA) disability score as a clinical marker, but no specific values were presented. A faster rate of disease progression in patients with older age of onset has been confirmed by others (40, 41, 136-138). A similar finding was also seen in Alzheimer’s disease (14, 60). In addition, age of onset may also have a role in determining the degree of drug improvement. In the study by Diamond et al (135), the degree of drug improvement decreased with increased age of onset. The improvement from baseline in the UCLA disability score after 6 years of levodopa treatment was 39.7, 38, and 7.1 points for groups with age of onset 60 years, respectively.

Gender Gender is another notable cofactor in altering the rate of disease progression in degenerative diseases. It has been suggested that women have a lower risk (0.40) than men of neurodegenerative disorders (74). This is thought to be caused by differences in hormonal state and by the menstrual cycle in premenopausal women (139). In osteoporosis, a higher risk of bone fracture is found in postmenopausal women, but this is much more clearly linked to loss of estrogen. In other studies of gender differences on disease progression rate, there have been inconsistent results (57, 60, 140-142).

DISEASE PROGRESS AND DRUG ACTION

651

CLINICAL TRIAL SIMULATION Disease progression can only be investigated by longitudinal studies. However, longitudinal studies have practical difficulties, such as expense and high patient drop-out rates, for reasons that may be linked to the disease progression itself. Studies with high patient drop-out rates should be analyzed with different approaches when attempting to recover the lost information. Ali & Siddiqui (143) have performed a simulation study to compare different analysis methods in handling missing data results from patients dropping out. A promising technique aimed at helping the design of such clinical trials has been proposed. This is the application of clinical trial simulation (144, 145). The aim of clinical trial simulation is to reduce the cost and shorten the drug development process by helping to design a more informative clinical trial. The power of clinical trial simulation is the ability to test a planned trial and preview the possible outcomes before actually carrying out a trial. This enables an inadequate design to be improved. A few studies have demonstrated the ability to explore designs of clinical trials through the application of clinical trial simulation (146-148).

SUMMARY The current means of studying disease progression in degenerative diseases have several major shortcomings. The methods for describing disease progression are often simplistic and limit the information the data can provide. Failure to identify between-subject variability prevents understanding of individual time course and response to treatment. The use of hierarchical modeling can overcome these shortcomings through its ability to describe the disease time course and through estimating both within- and between-subject variability. The significance of modeling disease progression is in describing not only the time course of disease but also the effects of treatment. Incorporation of pathophysiological understanding with pharmacological concepts holds the promise for developing better drugs and describing their effects more precisely. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED Le Baven DM, White P, Spillane JA, Goodhardt MJ, Curzon G, et al. 1979. Accelerated ageing or selective neuronal loss as an important cause of dementia? Lancet 1:11— 14 . Brayne C, Calloway P. 1988. Normal ageing, impaired cognitive function, and senile

dementia of the Alzheimer’s type: a continuum? Lancet 1:1265—67

. Martin WR, Palmer MR, Patlak CS, Calne

DB. 1989. Nigrostriatal function in humans studied with positron emission tomography. Ann. Neurol. 26:535-42 4. Sawle GV, Colebatch JG, Shah A, Brooks

652

CHAN

® HOLFORD

DJ, Marsden CD, Frackowiak RSJ. 1990.

Alzheimer’s brains. Neuroscience 89:675—

Striatal function in normal aging: implications for Parkinson’s disease. Ann. Neurol. 28:799-804

14. Huff FJ, Growdon JH, Corkin S. Rosen TJ. 1987. Age at onset and rate of progression

. Koller WC, Langston JW, Hubble JP, Irwin

of Alzheimer’s disease. J. Am. Geriatr. Soc.

I, Zack M, et al. 1991. Does a long preclinical period occur in Parkinson’s disease? Neurology 41:8-13

35:27-30 . Tetrud JW, Langston JW. 1989. The effect of depreny] (selegiline) on the natural history of Parkinson’s disease. Science 245:519-22 . Dompeling E, van Schayck CP, Molema J, Folgerimg H, van Grunsven PM, van Weel C. 1992. Inhaled beclomethasone improves the course of asthma and COPD. Eur. Respir. J. 5:945—52 We Parving HH, Rossing P, Hommel E, Smidt UM. 1995. Angiotensin-converting enzyme inhibition in diabetic nephropathy: ten years’ experience. Am. J. Kidney Dis. 26:99-107 . Giugliano D, Acampora R, Marfella R, Di Maro G, De Rosa N, et al. 1995. Tolrestat in the primary prevention of diabetic neuropathy. Diabetes Care 18:536-41 ey Lee CS, Schulzer M, Mak EK, Snow BJ, Tsui JK, et al. 1994. Clinical observations on the rate of progression of idiopathic parkinsonism. Brain 117:501—7 . The Parkinson Study Group. 1993. Effects of tocopherol and deprenyl on the progression of disability in early Parkinson’s disease. N. Engl. J. Med. 328:176-83

. Ohno T, Kato N, Shimizu M, Ishii C, Ito Y, et al. 1993. Effect of age on the de-

velopment or progression of albuminuria in noninsulin-dependent diabetes mellitus (NIDDM) without hypertension. Diabetes Res, 22:115-21 . Ott A, Breteler MM, van Harskamp F, Claus JJ, van der Cammen TJ, et al. 1995.

Prevalence of Alzheimer’s disease and vascular dementia: association with education.

The Rotterdam study. Br. Med. J. WS

310:970—

. Tesfaye S, Stevens LK, Stephenson JM, Fuller JH, Plater M, et al. 1996. Prevalence

of diabetic peripheral neuropathy and its relation to glycaemic control and potential risk factors: the EURODIAB IDDM complications study. Diabetologia 39:1377-84 Pde Rik MC, Tzourio C, Breteler MMB, Dartigues JF, Amaducci L, et al. 1997. Prevalence of parkinsonism and Parkinson’s disease in Europe: the EUROPARKINSON collaborative study. J. Neurol. Neurosurg. Psychiatry 62:10-15 . Burger H, de Laet CE, van Daele PL, Weel AE, Witteman JC, et al. 1998. Risk factors

for increased bone loss in an elderly population: the Rotterdam Study. Am. J. Epidemiol. 147:871-79 . Fratiglioni L, De Ronchi D, Aguero-Torres H. 1999. Worldwide prevalence and incidence of dementia. Drugs Aging 15:365— 75 . Fearnley JM, Lees AJ. 1991. Ageing and Parkinson’s disease: substantia nigra regional selectivity. Brain 114:2283-301 13h, Yew DT, Wong HW, Li WP, Lai HW, Yu WH. 1999. Nitric oxide synthase neurons in different areas of normal aged and

86

. Sano M, Emesto C, Thomas RG, Klauber

MR, Schafer K, et al. 1997. A controlled trial of selegiline, alpha-tocopherol, or both as treatment for Alzheimer’s disease. The Alzheimer’s Disease Cooperative Study. N. Engl. J. Med. 336:1216-22 . The Parkinson Study Group. 1998. Mortality in DATATOP: a multicenter trial in early Parkinson’s disease. Ann. Neurol. 43:318— 25

2% Bjorck S, Mulec H, Johnsen SA, Norden G, Aurell M. 1992. Renal protective effect of enalapril in diabetic nephropathy. Br. Med.1 J. 304:339-43 24. Lewis EJ, Hunsicker LG, Bain RP,

DISEASE PROGRESS AND DRUG ACTION Rohde RD. 1993. The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. N. Engl. J. Med. 329:1456-62 29) Laffel LM, McGill JB, Gans DJ. 1995. The beneficial effect of angiotensin-converting enzyme inhibition with captopril on diabetic nephropathy in normotensive IDDM patients with microalbuminuria. North American Microalbuminuria Study Group. Am. J. Med. 99:497-504 26. Zoja C, Donadelli R, Corna D, Testa D, Facchinetti D, et al. 1997. The renoprotective properties of angiotensin-converting enzyme inhibitors in a chronic model of membranous nephropathy are solely due to the inhibition of angiotensin II: evidence based on comparative studies with a receptor antagonist. Am. J. Kidney Dis. 29:254— 64 Pife Ruggenenti P, Remuzzi G. 1996. The renoprotective action of angiotensin-converting enzyme inhibitors in diabetes. Exp. Nephrol. 4:53-60 28. Holford NHG, Peace KE. 1992. Results and validation of a population pharmacodynamic model for cognitive effects in Alzheimer patients treated with tacrine. Proc. Natl. Acad. Sci. USA 89:11471Us oo: Leenders KL, Salmon EP, Tyrrell P, Perani D, Brooks

DJ, et al. 1990. The nigrostriatal dopaminergic system assessed in vivo by positron emission tomography in healthy volunteer subjects and patients with Parkinson’s disease. Arch. Neurol. 47:1290-98 30. Karbe H, Holthoff V, Huber M, Herholz K, Wienhard K, et al. 1992. Positron emission tomography in degenerative disorders of the dopaminergic system. J. Neural

Transm. 4:121-30 30 Takahashi H, Snow BJ, Nygaard TG, Calne DB. 1993. Clinical heterogeneity of doparesponsive dystonia: PET observations. Ady. Neurol. 60:586—90 30), de Leon MJ, Ferris SH, George AE, Christ-

653

man DR, Fowler JS, et al. 1983. Positron emission tomographic studies of aging and Alzheimer disease. Am. J. Neuroradiol. 4:568-71 333), Alavi A, Newberg AB, Souder E, Berlin JA. 1993. Quantitative analysis of PET and MRI data in normal aging and Alzheimer’s disease: atrophy weighted total brain metabolism and absolute whole brain metabolism as reliable discriminators. J. Nucl. Med. 34:1681-87 34, Seibyl JP, Marek KL, Quinlan D, Sheff K, Zoghbi S, et al. 1995. Decreased singlephoton emission computed tomographic ['*3I}beta-CIT striatal uptake correlates with symptom severity in Parkinson’s disease. Ann. Neurol. 38:589-98 3) Marek KL, Seibyl JP, Zoghbi SS, ZeaPonce

Y,

Baldwin

('*?lbeta-CIT/SPECT strates porters rology 36. Seibyl

RM,

et

al.

1996.

imaging demon-

bilateral loss of dopamine transin hemi-Parkinson’s disease. Neu46:231-37 JP, Marek K, Sheff K, Zoghbi S,

Baldwin

RM,

et al.

1998.

Iodine-123-

beta-CIT and iodine-123-FPCIT SPECT measurement of dopamine transporters in healthy subjects and Parkinson’s patients. J. Nucl. Med. 39:1500-8 Sl. Brooks DJ. 1991. Detection of preclinical Parkinson’s disease with PET. Neurology 41:24-27 38. Sawle GV, Playford ED, Burn DJ, Cunningham VJ, Brooks DJ. 1994. Separating Parkinson’s disease from normality. Discriminant function analysis of fluorodopa F 18 positron emission tomography data. Arch. Neurol. 51:237-43 89) Bhatt MH, Snow BJ, Martin WRW, Pate BD, Ruth TJ, Calne DB.

1991. Positron

emission tomography suggests that the rate of progression of idiopathic parkinsonism is slow. Ann. Neurol. 29:673-77

40. Vingerhoets Schulzer

FJ, Snow BJ, Lee CS, M, Mak E, Calne DB. 1994.

Longitudinal fluorodopa positron emission tomographic studies of the evolution of

654

CHAN

idiopathic 36:759-64

® HOLFORD

parkinsonism.

Ann.

Neurol.

4l. Morrish PK, Sawle GV, Brooks DJ. 1996. An[!8F]dopa-PET and clinical study of the rate of progression in Parkinson’s disease. Brain 119:585-91 42. Snow BJ, Tooyama I, McGeer EG, Yamada

T,

Calne

DB,

et

al.

1993.

Human positron emission tomographic [SF] fluorodopa studies correlate with dopamine cell counts and levels. Ann. Neurol. 34:324-30

43. Eidelberg D, Moeller JR. 1990. The metabolic anatomy of Parkinson’s disease: complementary [!8F]fluorodeoxyglucose and ['8F]fluorodopa positron emission to-

mographic studies. Mov. Disord. 5:203-13 44. Vingerhoets FJ, Snow BJ, Schulzer M, Morrison

S, Ruth

TJ, et al. 1994.

Re-

producibility of fluorine-18-6-fluorodopa positron emission tomography in normal human subjects. J. Nucl. Med. 35:18-24 45. Seibyl JP, Marek K, Sheff K, Baldwin RM, Zoghbi S, et al. 1997. Test/retest reproducibility of iodine-123-betaCIT SPECT brain measurement of dopamine transporters in Parkinson’s patients. J. Nucl. Med. 38:1453-59 46. Bench CJ, Price GW, Lammertsma AA, Cremer JC, Luthra SK, et al. 1991. Measurement of human cerebral monoamine oxidase type B (MAO-B) activity with positron emission tomography (PET): a dose ranging study with the reversible inhibitor Ro 19-6327. Eur. J. Clin. Pharmacol. 40:169-73 47. Sawle GV, Burn DJ, Morrish PK, Lammertsma AA, Snow BJ, et al. 1994. The effect of entacapone (OR-611) on brain ['SF]-6-fluorodopa metabolism: implications for levodopa therapy of Parkinson’s disease. Neurology 44:1292-97 48. Holford NHG, Sheiner LB. 1981. Understanding the dose-effect relationship: clinical application of pharmacokineticpharmacodynamic models. Clin. Pharmacokinet. 6:429-53

49. Boeckmann AJ, Sheiner LB, Beal SL. 1992. NONMEM Users Guides. San Francisco, University of California: NONMEM Project Group 50. Uhlmann RF, Larson EB, Koepsell TD. 1986. Hearing impairment and cognitive decline in senile dementia of the Alzheimer’s type. J. Am. Geriatr. Soc. 34:207-10 ae Yesavage JA, Poulsen SL, Sheikh J, Tanke E. 1988. Rates of change of common measures of impairment in senile dementia of the Alzheimer’s type. Psychopharmacol. Bull. 24:531-34 ee Salmon DP, Thal LJ, Butters N, Heindel WC. 1990. Longitudinal evaluation of dementia of the Alzheimer type: a comparison of 3 standardized mental status examinations. Neurology 40:1225—30 53. Bums A, Jacoby R, Levy R. 1991. Progression of cognitive impairment in Alzheimer’s disease. J. Am. Geriatr. Soc. 39:39-45 . Mornis JC, Edland S, Clark C, Galasko D,

Koss E, et al. 1993. The consortium to establish a registry for Alzheimer’s disease (CERAD). IV. Rates of cognitive change in the longitudinal assessment of probable Alzheimer’s disease. Neurology 43:2457-— 65 . Wild KV, Kaye JA. 1998. The rate of progression of Alzheimer’s disease in the later stages: evidence from the Severe Impairment Battery. J. Int. Neuropsychol. Soc. 4:512-16 56. Knopman D, Gracon S. 1994. Observations on the short-term “natural history” of probable Alzheimer’s disease in a controlled clinical trial. Neurology 44:260-65 Sie Stern RG, Mohs RC, Bierer LM, Silverman JM, Schmeidler J, et al. 1992. Deterioration on the Blessed test in Alzheimer’s disease: longitudinal data and their implications for clinical trials and identification of subtypes. Psychiatry Res. 42:101-10

58. Katzman R, Brown T, Thal LJ, Fuld PA, Aronson M, et al. 1988. Comparison of rate

DISEASE PROGRESS AND DRUG ACTION of annual change of mental status score in four independent studies of patients with Alzheimer’s disease. Ann. Neurol. 24:38489 ay), Thal LJ, Grundman M, Klauber MR. 1988. Dementia: characteristics of a referral population and factors associated with progression. Neurology 38:1083—90 60. Lucca U, Comelli M, Tettamanti M, Tiraboschi P, Spagnoli A. 1993. Rate of progression and prognostic factors in Alzheimer’s disease: a prospective study. J. Am. Geriatr. Soc. 41:45-49 61. OrtofE,Crystal HA. 1989. Rate of progression of Alzheimer’s disease. J. Am. Geriatr. Soc. 37:511-14 62. Antuono PG. 1995. Effectiveness and safety of velnacrine for the treatment of Alzheimer’s

disease.

A double-blind,

placebo-controlled study. Mentane Study Group. Arch. Intern. Med. 155:1766-72 63. Holford NH, Peace K. 1994. The effect of tacrine and lecithin in Alzheimer’s disease. A population pharmacodynamic analysis of five clinical trials. Eur J. Clin. Pharmacol. 47:17—23 64. Holford NHG. 1997. Population models for Alzheimer’s and Parkinson’s disease. In The Population Approach: Measuring and Managing Variability in Response, Concentration and Dose, ed. L Aarons, LP Bal-

ant, pp. 97-104. Brussels: COST B1 Eur. Comm.

65. Green CR, Mohs RC, Schmeidler J, Aryan M, Davis KL. 1993. Functional decline in

Alzheimer’s disease: a longitudinal study. J. Am. Geriatr. Soc. 41:654-61

66. Contin M, Riva R, Martinelli P, Procaccianti G, Cortelli P, et al. 1990. Response to a standard oral levodopa test in parkinsonian patients with and without motor fluctuations. Clin. Neuropharmacol. 13:19-28

67. Nutt JG, Woodward WR, Carter JH, Gancher ST. 1992. Effect of long-term therapy on the pharmacodynamics of levodopa. Relation to on-off phenomenon. Arch. Neurol. 49:1123-30

655

68. Contin M, Riva R, Martinelli P, Baruzzi A. 1993. Pharmacodynamic modeling of oral levodopa: clinical application in Parkinson’s disease. Neurology 43:367-71

69. Harder S, Baas H. 1998. Concentrationresponse relationship of levodopa in patients with different stages of Parkinson’s disease. Clin. Pharmacol. Ther. 64:18391 70. Contin M, Riva R, Martinelli P, Cortelli P, Albani F, Baruzzi A. 1994. Longitudinal

monitoring of the levodopa concentrationeffect relationship in Parkinson’s disease. Neurology 44:1287-92

7A Contin M, Riva R, Martinelli P, Albani F, Baruzzi A. 1997. Relationship between

levodopa concentration, dyskinesias, and motor effect in parkinsonian patients—a 3-year follow-up study. Clin. Neuropharmacol. 20:409-18

IZ Contin M, Riva R, Martinelli R, Cortelli P, Albani A, Baruzzi A. 1998. A levodopa

kinetic-dynamic study of the rate of progression in Parkinson’s disease. Neurology 51:1075-80

Wey, Nutt JG, Holford NHG. 1996. The response to levodopa in Parkinson’s disease: imposing pharmacological law and order. Ann. Neurol. 39:561—73 74. Hoehn MM, Yahr MD. 1967. Parkinsonism: onset, progression, and mortality. Neurology \7:427-42 IP. Marttila RJ, Rinne UK. 1977. Disability and progression in Parkinson’s disease. Acta Neurol. Scand. 56:159-69 76. Shoulson J, Parkinson Study Group. 1992. An interim report of the effect of selegiline (L-depreny]) on the progression of disability in early Parkinson’s disease. Eur. Neurol. 32:46-S3 TAL. Palhagen S, Heinonen EH, Hagglund J, Kaugesaar T, Kontants H, et al. 1998. Selegiline delays the onset of disability in de novo parkinsonian patients. Neurology $1:520-25 78. Olanow CW, Hauser RA, Gauger L, Malapira T, Koller W, et al. 1995. The effect

656

CHAN

® HOLFORD

of deprenyl-and levodopa on the progression of Parkinson’s disease. Ann. Neurol. 38:771-77

WS Morrish PK, Rakshi JS, Bailey D, Sawle GV, Brooks DJ. 1998. Measuring the rate of progression and estimating the preclinical period of Parkinson’s disease with ('!8F]dopa PET. J. Neurol. Neurosurg. Psychiatry 64:314-19

88.

89.

80. Morrish PK, Sawle GV, Brooks DJ. 1996. The rate of progression of Parkinson’s

disease. A longitudinal ['!*F]DOPA PET study. Adv. Neurol. 69:427-31 81. DeCarli C, Grady CL, Clark CM, Katz DA, Brady DR, et al. 1996. Comparison of positron emission tomography, cognition, and brain volume in Alzheimer’s disease with and without severe abnormalities of white matter. J. Neurol. Neurosurg. Psychiatry 60:158-67 82. Mori E, Hirono N, Yamashita H, Imamura

90.

.

T, Ikejiri Y, et al. 1997. Premorbid brain

.

84.

.

86.

87.

size as a determinant of reserve capacity against intellectual decline in Alzheimer’s disease. Am. J. Psychiatry 154:18-24 Luxenberg JS, Haxby JV, Creasey H, Sundaram M, Rapoport SI. 1987. Rate of ventricular enlargement in dementia of the Alzheimer type correlates with rate of neuropsychological deterioration. Neurology 37:1135—40 Burns A, Jacoby R, Levy R. 1991. Computed tomography in Alzheimer’s disease: a longitudinal study. Biol. Psychiatry 29:383-90 DeCarli C, Haxby JV, Gillette JA, Teichberg D, Rapoport SI, Schapiro MB. 1992. Longitudinal changes in lateral ventricular volume in patients with dementia of the Alzheimer type. Neurology 42:2029-36 Jobst KA, Smith AD, Szatmari M, Esiri MM, Jaskowski A, et al. 1994, Rapidly progressing atrophy of medial temporal lobe in Alzheimer’s disease. Lancet 343:829-30 Shear PK, Sullivan EV, Mathalon DH, Lim KO, Davis LF, et al. 1995. Longitudinal volumetric computed tomographic analy-

2.

93°

sis of regional brain changes in normal aging and Alzheimer’s disease. Arch. Neurol. 52:392-402 Kaye JA, Swihart T, Howieson D, Dame A, Moore MM, et al. 1997. Volume loss of the hippocampus and temporal lobe in healthy elderly persons destined to develop dementia. Neurology 48:1297—304 Fox NC, Freeborough PA. 1997. Brain atrophy progression measured from registered serial MRI: validation and application to Alzheimer’s disease. J. Magn. Reson. Imaging 7:1069-75 Jack CR Jr, Petersen RC, Xu Y, O’Brien PC, Smith GE, et al. 1998. Rate of medial temporal lobe atrophy in typical aging and Alzheimer’s disease. Neurology 51:99399 Fox NC, Cousens S, Scahill R, Harvey RJ, Rossor MN. 2000. Using serial registered brain magnetic resonance imaging to measure disease progression in Alzheimer disease: power calculations and estimates of sample size to detect treatment effects. Arch, Neurol. 57:339-44 Galasko D, Edland SD, Morris JC, Clark C, Mohs R, Koss E. 1995. The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). XI. Clinical milestones in patients with Alzheimer’s disease followed over 3 years. Neurology 45:1451—55 Berg L, Miller JP, Storandt M, Duchek J, Morris JC, et al. 1988. Mild senile dementia of the Alzheimer type. 2. Longitudinal assessment. Ann. Neurol. 23:477-84

94. Brooks JOD, Kraemer HC, Tanke ED; Yesavage JA. 1993. The methodology of studying decline in Alzheimer’s disease. J. Am. Geriatr. Soc. 41:623-28 953 Riggs BL, Wahner HW, Seeman E, Offord KP, Dunn WL, et al. 1982. Changes in bone mineral density of the proximal femur and spine with aging. Differences between the postmenopausal and senile osteoporosis syndromes. J. Clin. Invest. 70:71628 96. Glastre C, Braillon P, David L, Cochat P,

DISEASE PROGRESS AND DRUG ACTION

Meunier

PJ, Delmas PD. 1990. Measurement of bone mineral content of the lumbar spine by dual energy x-ray absorptiometry in normal children: correlations with growth parameters. J. Clin. Endocrinol. Metab. 70:1330-33 Sie Dockery DW, Ware JH, Ferris BG Jr, Glicksberg DS, Fay ME, et al. 1985. Dis-

tribution of forced expiratory volume in one second and forced vital capacity in healthy, white, adult never-smokers in six

U.S. cities. Am. Rev. Respir. Dis. 131:51120

98. Burrows B, Lebowitz MD, Camilli AE, Knudson RJ. 1986. Longitudinal changes in forced expiratory volume in one second in adults. Methodologic considerations and findings in healthy nonsmokers. Am. Rev. Respir. Dis. 133:974-80 99. Gore CJ, Crockett AJ, Pederson DG, Booth ML, Bauman

A, Owen

N. 1995.

Spirometric standards for healthy adult lifetime nonsmokers in Australia. Eur. Respir. J. 8:773-82

100. Franceschi S, Schinella D, Bidoli E, Dal Maso L, La Vecchia C, et al. 1996. The in-

fluence of body size, smoking, and diet on bone density in pre and postmenopausal women. Epidemiology 7:411-14 101. Nguyen TV, Sambrook PN, Eisman JA. 1998. Bone loss, physical activity, and weight change in elderly women: the Dubbo Osteoporosis Epidemiology Study. J. Bone Miner. Res. 13:145867 102. Ravn P, Cizza G, Bjarnason NH, Thompson D, Daley M, et al. 1999. Low body mass index is an important risk factor for low bone mass and increased bone loss in early postmenopausal women. Early Postmenopausal Intervention Cohort (EPIC) study group. J. Bone Miner. Res. 14:1622-27 103. Kristufek P, Brezina M, Ciutti P, Strmen J, Mayer M. 1987. Reference values and modelling of lung function development as a transcendent function of age, body

657

height and mass. Bull. Eur. Physiopathol. Respir. 23:139-47

104. Lebowitz MD, Holberg CJ, Knudson RJ, Burrows B. 1987. Longitudinal study of pulmonary function development in childhood, adolescence, and early adult-

hood. Development of pulmonary function. Am. Rev. Respir. Dis. 136:69-75

105. Morris JF, Koski A, Johnson LC. 1971. Spirometric standards for healthy nonsmoking adults. Am. Rev. Respir. Dis. 103:57-67 106. Lindall A, Medina A, Grismer JT. 1967. A re-evaluation of normal pulmonary function measurements in the adult female. Am. Rev. Respir. Dis. 95:1061—64 107. Ferris BG, Anderson DO, Zickmantel R. 1965. Prediction values for screening tests of pulmonary function. Am. Rev. Respir. Dis. 91:252-61 108. Kory RC, Callahan R, Boren HG, Syner MJC. 1961. The veterans administrationarmy cooperative study of pulmonary function. Am. J. Med. 30:243-58 109. Weyer G, Babej-Dolle RM, Hadler D, Hofmann S, Herrmann WM. 1997. A con-

trolled study of 2 doses of idebenone in the treatment of Alzheimer’s disease. Neuropsychobiology 36:73—82 110. Rogers SL, Friedhoff LT. 1998. Longterm efficacy and safety of donepezil in the treatment of Alzheimer’s disease: an interim anatysis ofthe results of aUS multicentre open label extension study. Eur. Neuropsychopharmacol. 8:67—75 111. Imbimbo BP, Verdelli G, Martelli P, Marchesini D. 1999. Two-year treatment of Alzheimer’s disease with eptastigmine. The Eptastigmine Study Group. Dement. Geriatr. Cogn. Disord. 10:139-47 iL. Schneider LS, Farlow MR, Henderson VW, Pogoda JM. 1996. Effects of estrogen replacement therapy on response to tacrine in patients with Alzheimer’s disease. Neurology 46:1580-84 INS Vingerhoets FJG, Schulzer M, Calne DB, Snow BJ. 1997. Which clinical sign of

658

CHAN

® HOLFORD

Parkinson’s disease best reflects the nigrostriatal lesion? Ann. Neurol. 41:58—64 114. Schulzer M, Lee CS, Mak EK, Vingerhoets FJG, Calne DB. 1994. A mathemat-

123:

ical model of pathogenesis in idiopathic parkinsonism. Brain 117:509-16

HLS). Myllyla VV, Sotaniemi KA, Vuorinen JA, Heinonen EH. 1992. Selegiline as initial treatment in de novo parkinsonian patients. Neurology 42:339-43 116. Schulzer M, Mak E, Calne DB. 1992. The antiparkinson efficacy of deprenyl derives from transient improvement that is likely to be symptomatic. Ann. Neurol. 32:795— 98 LONG, Vestbo J, Sorensen T, Lange P, Brix A,

124.

125%

Torre P, Viskum K. 1999, Long-term ef-

fect of inhaled budesonide in mild and moderate chronic obstructive pulmonary disease: a randomised controlled trial. Lancet 353:1819-23 118. Kerstjens HA, Brand PL, Hughes MD, Robinson

NJ, Postma

.

DS, et al. 1992.

A comparison of bronchodilator therapy with or without inhaled corticosteroid therapy for obstructive airways disease. Dutch Chronic Non-Specific Lung Disease Study Group. N. Engl. J. Med. 327:1413-19

119. Renkema TE, Schouten JP, Koeter GH. Postma DS. 1996. Effects of long-term treatment with corticosteroids in COPD. Chest 109:1156-62 120. Anthonisen NR, Connett JE, Kiley JP, Altose MD, Bailey WC, et al. 1994. Effects of smoking intervention and the use of an inhaled anticholinergic bronchodilator on the rate of decline of FEV1. The Lung Health Study. JAMA 272:1497-505 12 Gall MA, Nielsen FS, Smidt UM, Parving HH. 1993. The course of kidney function in type 2 (noninsulin-dependent) diabetic patients with diabetic nephropathy. Diabetologia 36:1071-78 . Bakris GL, Copley JB, Vicknair N, Sadler R, Leurgans S. 1996. Calcium channel blockers versus other antihypertensive

1

.

129:

therapies on progression of NIDDM associated nephropathy. Kidney Int. 50:164150 Crepaldi G, Carta Q, Deferrari G, Mangili R, Navalesi R, et al. 1998. Effects of lisinopril and nifedipine on the progressionto overt albuminuria in IDDM patients with incipient nephropathy and normal blood pressure. The Italian Microalbuminunia Study Group in IDDM. Diabetes Care 21:104-10 Reid IR. 2000. Drug treatment in postmenopausal osteoporosis. New Ethicals J. 3:43-5S1 Hillard TC, Whitecroft SJ, Marsh MS, Ellerington MC, Lees B, et al. 1994. Long-term effects of transdermal and oral hormone replacement therapy on postmenopausal bone loss. Osteoporosis Int. 4:341-48 Grey AB, Stapleton JP, Evans MC, Tatnell MA, Ames RW, Reid IR. 1995. The effect of the antiestrogen tamoxifen on bone mineral density in normal late postmenopausal women. Am. J. Med. 99:6364] Reid IR, Ames RW, Evans MC, Gamble GD, Sharpe SJ. 1995. Long-term effects of calcium supplementation on bone loss on fractures in postmenopausal women: a randomized controlled trial. Am. J. Med. 98:331-35 Meunier PJ, Vignot E, Garnero P, Confavreux E, Paris E, et al. 1999. Treatment of postmenopausal women with osteoporosis or low bone density with raloxifene. Raloxifene Study Group. Osteoporosis Int. 10:330-36 Ettinger B, Black DM, Mitlak BH, Knickerbocker RK, Nickelsen T, et al. 1999. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: results from a 3-year randomized clinical trial. Multiple Outcomes of Raloxifene Evaluation (MORE) Investigators. JAMA 282:637-45

DISEASE PROGRESS AND DRUG ACTION 130. Hosking D, Chilvers CED, Christiansen C, Ravn P, Wasnich R, et ai. 1998. Pre-

vention of bone loss with alendronate in postmenopausal women under 60 years of age. NV. Engl. J. Med. 338:485—92

Se

Reid IR, Wattie DJ, Evans MC, Gamble

GD, Stapleton JP, Cornish J. 1994. Con-

1S2

disease: age-related side effects. Neurology 33:1518-22

139: Mizuta E, Yamasaki S, Nakatake M, Kuno S. 1993. Neuroleptic malignant syndrome in a parkinsonian woman during the premenstrual period. Neurology 43:1048-49 140. Kramer-Ginsberg E, Mohs RC, Aryan M,

tinuous therapy with pamidronate, a potent bisphosphonate, in postmenopausal osteoporosis. J. Clin. Endocrinol. Metab. 79:1595-99 Delmas PD, Bjarnason NH, Mitlak BH,

ical predictors of course for Alzheimer patients in a longitudinal study: a preliminary report. Psychopharmacol. Bull. 24:458-62

Ravoux A-C, Shah AS, etal. 1997. Effects

141. Diamond SG, Markham CH, Hoehn MM,

Lobel D, Silverman J, et al. 1988. Clin-

of raloxifene on bone mineral density,

McDowell

serum

examination of male-female differences in progression and mortality of Parkinson’s disease. Neurology 40:763-66

cholesterol

concentrations,

and

uterine endometrium in postmenopausal women. NV. Engl. J. Med. 337:1641-47

33}. Pors Nielsen S, Barenholdt O, Hermansen F, Munk-Jensen N. 1994. Magnitude and pattern of skeletal response to long term continuous and cyclic sequential oestrogen/progestin treatment. Br. J. Obstet. Gynaecol. 101:319-24 134. Hart DM, Farish E, Fletcher CD, Barnes JF, Hart H, et al. 1998. Long-term effects

of continuous combined HRT on bone turnover and lipid metabolism in postmenopausal women. Osteoporosis Int. 8:326-32

Si)

659

Diamond SG, Markham CH, Hoehn MM, McDowell FH, Muenter MD. 1989. Ef-

fect of age at onset on progression and mortality in Parkinson’s disease. Neurology 39:1187-90 136. Hely MA, Morris JGL, Reid WGI, O’Sullivan

DJ, Williamson PM, et al. 1995. Age at onset: the major determinant of outcome in Parkinson’s disease. Acta Neurol. Scand. 92:455—63

WaMfhe Ransmayr G, Kunig G, Neubauer M, Wagner M, Falk M. 1995. Effect of age and disease duration on parkinsonian motor scores under levodopa therapy. J. Neural Transm. 9:177-88 . 138. Pederzoli M, Girotti F, Scigliano G, Aiello G, Carella F, Caraceni T. 1983. L-

dopa long-term treatment in Parkinson’s

FH, Muenter MD.

1990. An

142. Lyons KE, Hubble JP, Troster AI, Pahwa R, Koller WC.

1998. Gender differences

in Parkinson’s disease. Clin. Neuropharmacol. 21:118-21 143. Ali WA, Siddiqui O. 2000. Multiple imputation compared with some informative dropout procedures in the estimation and comparison of rates of change in longitudinal clinical trials with dropouts. J. Biopharm. Stat. 10:165-81 144. Hale M, Gillespie WR, Gupta SK, Tuk B,

Holford

NH.

1996.

Clinical

trial

simulation—streamlining your drug development process. Appl. Clin. Trials 5:35—40

145. Holford N, Kimko H, Monteleone J, Peck C. 2000. Simulation of clinical trials. Annu. Rev. Pharmacol. Toxicol. 40:20934 146. Jones DR. 1979. Computer simulation as a tool for clinical trial design. Int. J. Biomed. Comput. 10:145-50 147. Sheiner LB, Hashimoto Y, Beal S. 1991. A simulation study comparing designs for dose ranging. Stat. Med. 10:303-21 148. Gieschke R, Reigner BG, Steimer J-L. 1997. Exploring clinical study design by computer simulation based on pharmacokinetic/pharmacodynamic modelling. Int. J. Clin. Pharmacol. Ther. 35:469-74

_

ai?

hy

ly

ae

\a@

i: im nly

brings:

Annu. Rey. Pharmacol. Toxicol. 2001. 41:661—90

PROSTANOID RECEPTORS: Subtypes and Signaling* Richard M Breyer,' Carey K Bagdassarian,” Scott A Myers,’ and Matthew D Breyer? ' Division of Nephrology and Departments of Medicine and Pharmacology, Vanderbilt University, Nashville, Tennessee 37232; e-mail: [email protected], scott.myers @mcmail.vanderbilt.edu ?Department of Chemistry, College of William and Mary, Williamsburg, Virginia 23187; e-mail: [email protected] 3Division of Nephrology and Departments of Medicine and Molecular Physiology and Biophysics, Vanderbilt Univesity, and Department of Veterans Affairs Medical Center, Nashville, Tennessee 37232; e-mail: matthew.breyer @ memail.vanderbilt.edu

Key Words

prostaglandin, GPCR, G protein, mutagenesis, alternative splicing

@ Abstract Cyclooxygenases metabolize arachidonate to five primary prostanoids: PGE>, PGF>,,, PGI,, TxA>, and PGD>. These autacrine lipid mediators interact with specific members of a family of distinct G-protein-coupled prostanoid receptors, designated EP, FP, IP, TP, and DP, respectively. Each of these receptors has been cloned, expressed, and characterized. This family of eight prostanoid receptor complementary DNAs encodes seven transmembrane proteins which are typical of G-protein-coupled receptors and these receptors are distinguished by their ligand-binding profiles and the signal transduction pathways activated on ligand binding. Ligand-binding selectivity of these receptors is determined by both the transmembrane sequences and amino acid residues in the putative extracellular-loop regions. The selectivity of interaction between the receptors and G proteins appears to be mediated at least in part by the C-terminal tail region. Each of the EP,, EP3, FP, and TP receptors has alternative splice variants described that alter the coding sequence in the C-terminal intracellular tail region. The C-terminal variants modulate signal transduction, phosphorylation, and desensitization of these receptors, as well as altering agonist-independent constitutive activity.

INTRODUCTION Prostaglandins (PGs) comprise a diverse family of autacoids, whose synthesis is initiated by cyclooxygenase-mediated metabolism of the unsaturated 20-carbon fatty acid arachidonic acid to PGG/H),, generating five primary bioactive prostanoids: *The US Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper.

661

662

BREYER ET AL

PGE;, PGF;,,, PGD>, PGI,, and TXA, (1, 2). The importance of this pathway in

a broad array of diseases including cancer, inflammation, and hypertension is underscored by the classic and novel uses of cyclooxygenase-inhibiting nonsteroidal anti-inflammatory drugs that nonselectively inhibit the synthesis of all of these compounds. Each prostanoid is synthesized in specific compartments within the body via the action of specific synthases. These autacoids then act within the tissue where they are synthesized via specific G-protein-coupled receptors (GPCRs), designated EP for PGE, receptors and FP, DP, IP, and TP for PGF,,, PGD,, PGI,, and TXA, receptors, respectively (3,4). The chemical structure of each of the five major PGs is shown in Figure 1A. The energy-minimized geome- _ tries of the endogenous ligands are very similar (Figure 1B), and, although each PG binds with the highest affinity to its cognate receptor, considerable ligand-binding cross-reactivity can be observed between a given prostanoid and other receptors within the family (Table 1). The PG receptors have the characteristic seven-hydrophobic-transmembranesegment architecture typical of GPCRs, and several of the prostanoid receptors display alternatively spliced variants in the C-terminal sequence that can alter receptor function (Figure 2). The prostanoid receptors belong to the family A GPCRs (5). Phylogenetic studies have broken family A down into five evolutionarily conserved groups, with the PG receptors in group V (5). In addition to the prostanoid receptors, this group includes a number of receptors for autacrine, paracrine, and endocrine factors such as small tripeptides, pituitary hormones, glycoprotein hormones, opioids, and platelet-activating factor. Interestingly, the prostanoid family is most closely related to the vasopressin receptor family of peptide-binding hormone receptors, and as described below, the ligand-binding motif of the prostanoid receptors shares some similarities with this class of peptide-binding receptors rather than with other receptors that bind small-molecule ligands, for example the adrenergic receptor family (6). Most of the receptors in group V signal via stimulation of

phospholipase C to produce IP, and di-acyl-glycerol or via inhibition of adenylyl cyclase through inhibitory guanine nucleotide-binding regulatory protein (G,),

>

Figure 1 (A) The structure of the five principal prostaglandin metabolites. (B) Energyminimized prostanoid molecular geometries. These structures result, with only small perturbations, from minimization in vacuo with either the AM1 or PM3 semi-empirical method and also from several ab initio schemes (STO-3G, 6-31G*, 6-31G**). For each prostanoid molecule, bond lengths, bond angles, and dihedral angles are systematically varied from a starting molecular geometry until a minimum energy structure is located. Ab initio schemes are the more exact since semi-empirical calculations introduce further approximations to the quantum mechanical calculations (4a). Calculations were performed with the Gaussian 94 and Molecular Simulations Inc. computational packages. Note similar three-dimensional geometries for the structurally different prostanoids. The orientation of the prostanoid structures is similar to that shown for the chemical structures in panel A, with the prostanoid ring on the left, the carboxyl-containing alpha side chain to the upper right, and the omega chain to the lower right. Carbon atoms are medium lines, oxygen atoms are bold lines, and hydrogen atoms are gray (light) lines.

PROSTANOID RECEPTORS

red

OH

PGD?2

663

664

BREYER ET AL

B

PGE?

PGD?2

PGF2q

TxA Figure 1

(Continued)

PGI?

ate

ls

@)TOOOHOOQOOOQOOOTOOOOPAOHMPHAGQOOOQOOHGH®

NHo

OD©

() A) (R) © ©) (0)

: P Alternatively spliced C-termini

ermini

72A

=.

= 74,

©

e1

Ce mS

E)

©COOH

site

, of point

divergence

B0A

Or

)

&

@

w

®

ee bee ® TOBGOOWOOOGHEDOOV

© ©

tld

COOH

clone terminus

Splice

77A

se)

COOK OMAOSDHAHNOMOOGONCOOODOOD OM a

‘

oF ay)

© (n)

® LOOQOOEQHPHODOOOHOHDOGHOHOGOOQWEL TOMOORONGHORS cS) ® $ 2

“ORPCXYXGXG)C OOH

Figure 2 Identification of conserved residues in rabbit EP3 receptor splice variants differing only Sequence alignments of the predicted amino acid were carried out utilizing the TMAP program (145). and mouse EP; receptors; the human EP, receptor;

the EP; receptor and sequence of five in their intracellular carboxy] termini. sequences for the prostanoid receptors The sequences aligned were the human EP; receptors from rabbit, rat, mouse,

cow, and human tissues; the EP, receptor from rabbit, mouse, rat, and human tissues; the TXA, receptor from mouse, rat, and human tissues; the FP receptor for mouse, rat, cow,

and human tissues; the IP receptor from mouse and human tissues; and the DP receptor

from mouse and human tissues. Conserved residues are indicated by gray circles, and invariant residues are indicated by black circles. Residues with “bulls-eye” symbols are conserved across the entire superfamily of GPCRs (146), those without this inset are unique to the prostanoid receptors. The predicted amino acid sequences of each splice variant are represented by the one-letter amino acid code. The carboxyl-variable tails range from 56-amino-acid residues for clone 77A to none for the NT (no-tail) clone.

BREYER ET AL

666

00rT 0091

0091 Orel< 9LI ¥ OL81

£Tl ¥F CKO

(Or€—-061) 0ST

(€Z-€D LI *

*

*

(00P-0S1)OFT (OrT-€8) O11

+9 F $96

* + *

*

* ‘

OOe!

* (9T-LI) 17

*

* *

+

*

* Ol F €@ Orel < a an0 pOEL F OvS8 COt ¥F OSLI 6vl F SIB CIv6 F CEE

pOl F Lys 70 F 81

(€6-8P) LO (L6'0-L¥'0) 89°0 (tL'0-'0) 95'0

(I-96)El r

CFSE

vO FIP *

(0091-095) 0001 *

+

(OvI-€L) 001 + '-

Ovel< 6t ¥ L8v9 + + ZOL F £719

(TP-BTO) VE

(3'P-1'€) BE

Fe

* *

*

* +

*

OOLI

0L9 ¥ OOL8 10@ # 1711

+

(99-%¢£) LY +

(O8P-0S7) OSE 000°01< (pCI-ETD VI (0€6-09€) 08S

* Il F €ve P

*

* TET F STEP * Ly ¥ LOS x ueuiny dL

asnouwl gi

(LL-L¥) 09 (Ov I-€L) 001

*

*

+ k

Ips # 7099

*

+

*

*

1081 + O78S

*

*

OOET

asnow dL

* *

*

*

*

* +

*

* + BE ¥F 6TE CL ¥ 6I¥ cI # LOI *

Cl + 16

90 + 09 61 F 61c OOI # tLoc

(OS 1-46) 071 x * (OSI-OLD OZI (SZ-LI) IZ *

* 6L (I-11) PI (97-S1) 0%

*

SO F 6

uewny 'qq

86 F LITI L8I #F TEEI LO ¥ Sel

(S1-@'6) TI

asnow ! qq

8+

IZ

61997 dejue Cy1-S dO

8rjue/gpS6cOS urpoAonque)

ysoides9g

* 067 ‘LUV

joua}soidny-q

ysosdvot> ysosdoyy (gy) sosdouryey (vx) sosdouryeT joua)sosdo[D

£7@

OStT MA

MAD Iv8011AZ

TOF

Orel< SOI # 619 86 F SSS 70 F 87 S00 = LY0

*

SOcETHV

yy FIN [o1sosds yd [oysosdosty] ZADAdIAMIOUIP 91°91 66LE9AD LOL8TAN auojsosding

1d9d-(HO)I

7d0Dd

COFTE #

SOFLI

* OIS9S F T8E6

I€l + Iv7l vl ¥ OIS Ol + 361

ya 1soideing

AN 1so1deing

pda

¢di

ssejo puesry]

x

3ejue/6089HV 3eue/6801SOS BeyUue/TTEISOS 7ADd (Aueyd-L1 TdADd

TaTa VL

qpuesry

*

x *

Tl ¥ 61!

ueuny di

(QU) JURIsUOD IO}IqrYU] psiojdaoal prourysoid paydnoo-uiajoid-E pouoyo 0} Surpulq prourjsoid jo

uewiny “qa

dL

dl

1dd

* a, 2 2

asnowl “qa

4

*

*

* *

* *

077

( panuijuo)) (O1E-O9D)

‘x “TOP +09 8E F 9

7GDd

MA OSPT TMZ IP8OI mADd

di

Foe

dL *

JUasardar CG%

sUePIUOSs[RAroyUI SP Uae) Wo

ZIMOWIQY 39 [2 °(g) ..x,,

* *

*

* *

LO0

uewny

681

€10€F

*

TSE¥

x

“

*

x

dd

* *

SIP¥

k +

(Ga-eiorLOE +

asnow

J0U,,

[AtHOU“ays

x

*

ILI Orel


ve).

wow) e.

AD>Ties4O4

a)

4 z

sg

ro

RTS

a]

Im,

a

be

sq ,

fed 2

BOs rin

Le

i

i

:

«@

g

6

é «"

Copyright

Annu, Rey. Pharmacol. Toxicol. 2001. 41:775—87 © 2001 by Annual Reviews. All rights reserved

MOLECULAR APPROACH TO ADENOSINE

RECEPTORS: Receptor-Mediated Mechanisms of Tissue Protection J Linden Departments of Cardiovascular Medicine and Molecular Physiology and Biological Physics, University of Virginia, Charlottesville, Virginia 22908; e-mail: jlinden@ virginia.edu

mast cells, inflammation, ischemia/reperfusion injury, asthma, Key Words preconditioning

H Abstract Adenosine accumulation during ischemia and inflammation protects tissues from injury. In ischemic tissues adenosine accumulates due to inhibition of adenosine kinase, and in inflamed tissues adenosine is formed from adenine nucleotides that . are released from many cells including platelets, mast cells, nerves, and endothelium intidases ecto-nucleo of family a by adenosine to converted Nucleotides are rapidly cluding CD39 and CD73. Activation of A; and possibly A3 adenosine receptors (ARs) protects heart and other tissues by preconditioning through a pathway including protein kinase C and mitochondrial K arp channels. Activation of Aj, receptors limits reperfus sion injury by inhibiting inflammatory processes in neutrophils, platelets, macrophage that receptors by mediated responses atory proinflamm produces Adenosine cells. T and human vary among species; A; and App receptors mediate degranulation of rodent and ligective subtype-sel receptor adenosine or canine mast cells, respectively. Novel MRS1220 blocker), (App MRS1754 include These developed. ands have recently been (A; blocker), MRE

3008F20 (human A; blocker), MRS1523

(rat A; blocker), and

investigators ATL146e (Aza agonist). These new pharmacologic tools will help therapeutic new identify to and injury from to sort out how adenosine protects tissues diseases. ischemic and ry inflammato of treatment the agents that hold promise for

BACKGROUND modulate physAdenosine is a primordial signaling molecule that has evolved to of its effects, it breadth the iological responses in all mammalian tissues. Due to anding of underst our in is not possible to summarize all of the new developments confined is review This adenosine receptor physiology even within the past year. and the g signalin and on to recent insights in the understanding of receptor regulati tion informa new on description of significant new pharmacological tools. I focus or ischemic limit may about how activation or inhibition of adenosine receptors 0362-1642/01/0421-0775$14.00

Tle

776

LINDEN

inflammatory tissue injury, an area of particularly interesting mechanistic and therapeutic importance.

ADENOSINE RECEPTORS: Physical and G Protein Coupling Characteristics Receptor-mediated effects of adenosine are mediated by four G protein-coupled receptors designated A,, Aj,, Aog, and A; (Table 1). All four receptors are N-linked glycoproteins, and all but A,, have sites for palmitoylation near the carboxyl terminus (1-4). Glycosylation has no effect on the affinity of ligands for receptors and may be involved in targeting newly formed receptors to the cell surface. In view of emerging evidence that certain G protein-coupled receptors may form homo- or heterodimers, it is possible that hydrophilic glycosylation of adenosine receptors, and G protein-coupled receptors in general, may inhibit dimerization reactions driven by hydrophobic interactions. All of the adenosine receptors can be readily deglycosylated upon incubation with N-glycosidase F. One practical application of deglycosylation is its use to distinguish between specific and nonspecific antibody binding to putative receptors detected by western blotting. Figure 1 illustrates the use for western blotting of two different anti-A,, receptor antibodies. Both antibodies specifically detect overexpressed recombinant A>, receptors, but only one of the antibodies is able to detect low levels of receptors found on tissues. Furthermore, nonreceptor immunoreactivity is found in proteins that have nearly the same molecular mass as the receptor. This illustrates two points about the use of antireceptor antibodies: (a) Detection of overexpressed recombinant receptors is not necessarily predictive of the ability of an antibody to detect endogenous receptors that are expressed at a much lower levels and (b) deglycosylation is an important control to check for specificity when doing western TABLE 1

Properties of adenosine receptor subtypes

Adenosine receptor subtype

Ay Avda Aop A3

Genbank accession number

Amino acids

G Protein coupling

Chromosomal location

References

Human

$45235

326

i,0

1q32.1

(48)

Mouse

U05671

326

Mouse

Human

S$46950

412

U05672

— 410

Human

X68487

332

Mouse

U05673

332

Human Mouse

122607 120331

318 319

(49)

s,olf

22q11-2

(50)

(49) S.q

17p11.2-12

(51) (49)

i,0

SS

1p13.3

(52) (53)

ADENOSINE RECEPTORS

Monoclonal Anti-A2A Receptor Antibody Recombinant

FRAME

A2A

athe

Polyclonal Antipeptide Antibody Recombinant eae 2.

Striatum

Odi.

BENE

Striatum

CEERI

135 88—' Aber mi i ee

3 ‘a

1

pA

kKDanan

a

: a8

Fe

N-Glycanase

tc

‘

f

ig— B kDa

~

ay

=

N-Glycanase

a

Figure 1 Western blots showing specific and nonspecific binding of anti-A», receptor antibodies. in their Treatment of adenosine receptors with N-glycosidase F produces a characteristic decrease A>» striatal and molecular mass. (A) Specific immunoreactivity of overexpressed recombinant coma of ivity immunoreact nonspecific and receptors with a monoclonal antibody. (B) Specific by glycosidase mercially available polyclonal antibody. The absence of a shift in molecular mass 43.) Reference from (Adapted ivity. immunoreact nonspecific of indicative is F probably

blotting. The use of poorly characterized antireceptor antisera may have resulted n, in several instances of erroneous conclusions about adenosine receptor expressio detect to used distribution, and regulation. Anti-A>, receptor antibodies have been (5), receptors selectively in areas where they are highly expressed, such as striatum lower. is but not convincingly in endogenous tissues in which the receptor density from mice The best control for immunohistochemical experiments is tissue derived indicate tions in which specific receptor genes have been deleted. These considera still binding that for the purpose of quantifying adenosine receptors, radioligand is generally preferable to immunohistochemistry or western blotting.

ADENOSINE METABOLISM all stimulate It has long been known that hypoxia, ischemia, or inflammation adenosine, to barrier a is lium endothe the Because local adenosine production. large part in derived be may vessels blood of lumen adenosine formed within the intersticontrast, By cells. ial endothel or platelets from nucleotides released from recepAy, on acting by nantly predomi tion tial adenosine may produce vasodila ial interstit to le accessib arly particul are that tors on vascular smooth muscle cells or cells ymal parench ischemic likely is ne nucleoside. The source of this adenosi

778

LINDEN

nucleotides derived from nerves or interstitial mast cells. Ithas recently been shown in the heart that hypoxia-induced inhibition of adenosine kinase amplifies small changes in free myocardial AMP into a major rise in adenosine. This mechanism plays an important role in causing high sensitivity of the myocardium and other tissues to impaired oxygenation (6). The concentration of endogenous adenosine acting at the receptor level during an ischemic episode was estimated to be 30 4M in rat hippocampal slices, based on the ability of the selective A, receptor antagonist, 8-cyclopentyl-1,3-dipropylxanthine (CPX), to reverse the effects of ischemia (7). Adenosine also can be derived from adenine nucleotides released from many cell types by mechanisms that are not yet clearly understood (8). Substantial amounts of adenosine may be formed from the breakdown of adenine nucleotides that are present in the granules of autonomic nerves, platelets, and mast cells. Adenine nucleotides are rapidly converted to adenosine by a family of ecto-ATP/ADPases including CD39 (NTPDase 1) and ecto-S’ nucleotidases including CD73 (Figure 2). The expression of CD39 on the endothelial cell surface may be regulated because palmitoylation targets the enzyme to caveolae (9). This in turn may regulate the rate of ADP conversion to adenosine. Inosine, formed by adenosine deamination, accumulates to even higher levels (>100 uM) than adenosine in ischemic tissues. Inosine has been found to activate

rat and guinea pig A; receptors with Ki values in the range of 15-25 uM (10). In contrast to its effects to activate rodent A; receptors, inosine is a weak partial agonist of the human A, receptor (J Linden, unpublished).

A, RECEPTORS A, adenosine receptors signal through Gi/o pathways and inhibit adenylyl cyclase,

activate K* channels, or inhibit Ca** channels in various cells. A, receptors also can stimulate Ca**+ mobilization via a pertussis toxin-sensitive pathway through activation of PLCB with G protein By subunits (11). This signaling pathway appears to be synergistic with receptors that activate PLC via Gq, possibly including P2Y receptors and A>, receptors. A; receptors couple preferentially to G proteins containing y2 or y3 over subunits containing y 1 (12). This preference has been shown to be mediated by the prenylation state of the y-subunit. G proteins containing geranyl geranylated y subunits (including y2 and y 3) interact more effectively: with A, receptors then do G proteins containing farnesylated y subunits (13). It has recently been shown that PKC can phosphorylate the y 12 subunit of heterotrimeric G proteins, resulting in increased G protein affinity for A, receptors (14). Much recent work has focused on A, receptor-PKC signaling cascades. The A, agonist, N°-cyclopentyladenosine (CPA), has been recently recognized as a facilitator of insulin-stimulated leptin release through a pathway involving protein kinase C (15). Activation of A, adenosine receptors increases nucleoside efflux from DDT 1 MF-2 cells by a PKC-dependent inhibition of adenosine kinase activity (16). A, receptors and the heterotrimeric G protein, Go, are abundant in the brain. Recently, GRIN1, a probable regulator of neurite growth, has been identified as an effector of Go (17). This suggests a possible newly identified function of Ay

ADENOSINE RECEPTORS

Ta

Platelet

AaB

'

NO=Ca2+

'

cAMP ———-————

adhesion

endothelium

~*~ —

AoB_

: Inosine

or A3

Macrophage

Figure 2 Purinergic regulation of inflammation. ADP derived from activated platelets exerts a pro-aggregatory effect on platelets through cell surface P2Y receptors, including a newly identified P2Y12 receptor (9a), that is countered by ecto-nucleotidases that degrade ADP and produce adenosine (ado) that activates anti-aggregatory A>, receptors. Activation of Aj, receptors also reduces histamine and cytokine release from certain mast cells and macrophages and inhibits the expression of adhesion molecules on endothelium. A3 receptors regulate rodent mast cells and App receptors regulate human and canine mast cells. Inosine activates rodent, but not human, A3 receptors. A>p receptors are dually coupled to adenylyl cyclase via Gs and to Ca*+ via Ga.

receptors to regulate neurite growth in the central nervous system through a Go/GRIN1 pathway. Palmitoylation of A, receptors has no effect on receptor-effector coupling, receptor downregulation, or receptor interactions with G proteins. However, A, palmitoylation may divert new synthesized receptors from a pathway leading to rapid receptor degradation (1).

A,, RECEPTORS A>, receptors are most highly expressed in intermediate spiney neurons of the striatum. A>, receptors are known to activate Gs, but receptors in striatum may interact predominantly with G,; (first identified in the olfactory epithelium) because

780

LINDEN

G,ir is much more highly expressed in the striatum than Gs (18). Functionally, activation of Aj, receptors opposes the action of D2 dopamine receptors in spiney neurons, and antagonists of Aj, receptors are being investigated for possible use in Parkinson’s disease (19)

Activation of Aj, receptors causes vasodilatation of coronary arteries and, to a variable extent, other blood vessels. A>, receptors also are found on bone marrow— derived cells including neutrophils, monocytes, macrophages, platelets, and mast cells. Activation of Aj, receptors produces a constellation of responses that in general can be classified as anti-inflammatory (20, 21).

A,3 RECEPTORS A» receptors have long been known to couple to Gs. Recent studies indicate that Aj, receptors also couple to Gq and produce Ca** mobilization and mitogenactivated protein kinase (MAPK) activation (3, 4). Ca** mobilization is not limited

to cells that overexpress Ajp receptors because the endogenous A>, receptors of HEK-293 cells produce a robust A5,-mediated Ca** mobilization (22). In HEK-293 cells the agonist, N-ethylcarboxamido-adenosine (NECA) is equipotent in elevating cyclic AMP and stimulating MAPK activation. The protein kinase A inhibitor, H89, blocks forskolin but not NECA-stimulated

MAPK

activation

in HEK cells, suggesting that the Gq pathway contributes to MAPK activation through a pathway including MEK and Ras (22). A»g receptors on vascular endothelial cells may contribute to an NO-dependent component of vasodilation mediated by Ca?*-dependent NO synthase activation (Figure 2).

A, RECEPTORS A; receptors appear to signal through Gi in much the same way as A, receptors. Unlike A, receptors, depalmitoylation of A; receptors renders them susceptible to phosphorylation by G protein-coupled receptor kinases (GIRKs) (23). This in turn leads to rapid phosphorylation and desensitization of A, receptors that does not occur in the case of A, receptors (2, 24).

ADENOSINE AND TISSUE PROTECTION Adenosine protects tissues from hypoxic or ischemic damage through multiple receptor subtypes. The activation of A, and possibly A; (25) receptors produces preconditioning to protect the heart and other tissues from subsequent ischemic injury. Adenosine has been postulated to trigger preconditioning by increasing mitochondrial K-ATP channel activity through a pathway including PKC (26). A late phase of preconditioning in response to A, receptor activation in the rabbit

ADENOSINE RECEPTORS

781

heart appears to be mediated in part by the induction of manganese superoxide dismutase (27). In contrast to preconditioning, agonists of A,, receptors can protect tissues from ischemia/reperfusion damage when added during the reperfusion period. The agonist CGS21680, which is highly selective for A,, over A, and A5p receptors, was found to attenuate reperfusion injury in the dog heart (28). This effect is correlated with an inhibition of neutrophil accumulation, superoxide generation, and coronary endothelial adherence, suggesting that reduced inflammation may be responsible for protecting the heart during reperfusion injury. A new agonist, ATL 146e, which is >50 times more potent at human A,, receptors than CGS21680, has recently been found to produce profound protection of the rabbit lung (29) and rat kidney (30) from reperfusion injury at concentrations that are well below the threshold to produce hemodynamic responses. Protection from reperfusion injury is accompanied not only by reduced neutrophil accumulation in ischemic renal microvessels, but also by reduced expression of the adhesion molecules,

P-selectin, and [CAM-1 on the reperfused vascular endothelium (31). Aj, agonists also protect isolated crystalloid-perfused rabbit heart from ischemia/reperfusion injury. The effect is manifest as a significant reduction in contracture development during reperfusion. These data imply a role for Aj, receptors on cardiomyocytes or tissue-resident inflammatory cells in A,, receptor—mediated cardioprotection (32). Activation of Aj, ARs on human monocytes inhibits, by a cyclic AMPdependent mechanism, the secretion of IL-12, a proinflammatory cytokine and a major inducer of Th1 responses (33). Through this mechanism, adenosine re-

leased in excess during inflammatory and ischemic conditions, or tissue injury, may contribute to selective suppression of Th1 responses and cellular immunity. It is thought that activation of T lymphocytes is required for neutrophil recruitment during ischemia reperfusion injury in the liver. The subacute phase of ischemia/reperfusion injury in the liver is absent in athymic nude mice but can be restored by adoptive transfer of CD4* T-cells (34). Inhibition of T-cell activa-

tion may contribute to the protection of tissues from ischemia/reperfusion injury because inhibitory A, receptors are found on CD4* T cells (35).

ADENOSINE RECEPTORS ON MAST CELLS Aerosolized adenosine has the effect of causing mast-cell-dependent bronchoconstriction in asthmatic subjects but bronchodilation in nonasthmatics (36, 37). More-

over, the nonselective adenosine receptor antagonist, theophylline, is widely used as an antiasthmatic drug, although its mechanism of action is uncertain. A related xanthine, enprofylline, is also therapeutically efficacious in the treatment of asthma and was thought to act through a non—adenosine receptor—-mediated mechanism owing to its low affinity at A, and A>, receptors (37). The A; adenosine receptor was initially implicated as the receptor subtype that facilitates the degranulation of

782

LINDEN

rat RBL 2H3 mast-like cells (38) and perivascular mast cells of the hamster cheek pouch (10). There is also evidence of mast cell degranulation when agonists of A; ARs are administered to rats or mice (39, 40). In contrast, the A>z, AR has been

implicated as the receptor subtype that facilitates the release of allergic mediators from canine BR and human HMC-1 mastocytoma cells (41,42). A role for A>, ARs in human asthma is also suggested by the efficacy of enprofylline, which at therapeutic concentrations of 20-50 uM only blocks the Az AR subtype (4). In sum, the literature indicates that the release of allergic mediators in some mast

cells is mediated by A; ARs and in other cells is mediated by Az ARs. This may result from species differences, with rodent (rat, guinea pig, and mouse) and canine or human mast cells responding primarily to A; or A>p adenosine receptor stimulation, respectively.

ADENOSINE RECEPTOR PHARMACOLOGY Potent and selective agonists and antagonists of adenosine receptor subtypes developed recently (Figure 3) have been valuable for further defining the physiological effects of the various adenosine receptor subtypes. However, it is not always appreciated that the selectivity of these compounds is limited. For example, at concentrations above 1 uM, CGS21680 is not a selective A,, agonist, IB-MECA is not a selective A; agonist, and CPX is not a selective A, antagonist (Table 2).

TABLE 2

Antagonist binding to human adenosine receptor subtypes, (Ki, nM)

Ay

Ada

AoR

A3

References

A, selective CPX WRC0571

2 3

156 157

40 19,000

509 6,500

(53) (53)

Apa Selective ZM241385 SCH58261

536 287

1.4 0.6

31 5,011

269 >10,000

(4) (54)

A»p-Selective MRS1754 Enprofylline*

403 44,000

503 32,000

ey) 6,300

570 158,000

(46) (4)

A; selective MRS1220 MRS1523 (rat) MRE300F20

305 15,600 1,100

52 2,050 140

— — 2,100

0.65 519° (113) 0.29

(55) (29) (56)

_ sO

“It is notable that plasma concentration of enprofylline achieved in the treatment of asthma is 20-50 EM, sufficient to block human Aj, receptors but not other adenosine receptor subtypes.

PA K, of 519 + 86 nM (N = 6) of MRS1523 for the rat Aj receptors determined in the author’s laboratory is somewhat higher than the K, reported previously.

ADENOSINE RECEPTORS

bal Ne

HN NS al s) N

783

[ Mt ae

ot Ne

NZ

j

—

CPX

(A, Antagonist)

HO

(A; Agonist)

NH> NON - is O

4 OH

l SN

0 Soot

(Aza Agonist)

O

.

~

(0) -CHsC- NH- \

hoy OH

< i) S

OH

(nonselective agonist,

;

most potent A2B)

ia

NvNi~ j

NH ¢

= aera

Figure 3

N

e OHaOHe.

i SN

i

O

~N7

—

_~OCHs

07 NH

oh

|

ue \

N

Led

Je }

J ;

Nia

ZA

ia N~

0

(Azg Antagonist)

N

NECA

a

pon

MRS1754

r

N

fe)

ha

0

ay

Oth N

yt

ah oh wg) FO ONT ON 7241385 H

Ni

on

NH»

alee

en

(Aza Antagonist)

=

ATL146e OH

wk

SS

N

aes

Z

OH OH

f

oO

HO.

(

O

NA

c|

WN N

J

Ay O

MRS1220 ;

(Ag Antagonist)

MRE 3008F20 (Az antagonist)

N*phe~Cl Cl-IB-MECA agonist lle

(Az selective

Selective agonists and antagonists of adenosine receptor subtypes.

The use of agonists to determine the receptor subtype that mediates a particular response is particularly problematic because agonists bind with two different affinity states of G protein-coupled adenosine receptors, resulting in wide differences in reported binding affinities, depending on the source of receptors (recombinant receptors tend to be uncoupled from G proteins) and the use of agonist or antagonist radioligands (antagonists detect more uncoupled receptors). Moreover, the EDs, of agonists in functional responses is highly variable, depending on the extent of receptor reserve (43). The judicious use of new selective antagonists is simplifying the process of identifying which adenosine receptor subtypes mediate various physiological responses. It is important to note, however, that there are substantial species differences in the affinity of these compounds. For example, the widely used A, selective antagonist, CPX binds with >10 times lower affinity

784

LINDEN

to canine than to human or rat A, receptors (44) and is fairly potent as an antagonist of human A,, receptors (Table 2). Compounds that block human A; ARs generally are weak antagonists of rodent A, receptors, including the new selective antagonist of human A, receptors, MRE 3008-F20 (45). MRS1523 is the only compound reported to be moderately selective as an antagonist of the rat A; receptor (Table 2). Recently, the first potent and selective antagonists of Aj, adenosine receptors have been described. These compounds, exemplified by MRS1754 are anilide derivatives of 8-phenylxanthines (46). It has recently been suggested that selective inhibitors of Ap receptors may be useful for the treatment of asthma and other allergic diseases (4, 42, 46).

SUMMARY AND FUTURE DIRECTIONS Because adenosine is a metabolic breakdown product of ATP, adenosine receptors may have evolved in part to protect tissues from various injurious stimuli. As summarized in this review, activation of A, and A; receptors ellicit protective responses in various tissues by several processes collectively referred to as preconditioning. Activation of A, receptors produces a constellation of effects on various inflammatory cells types that can attenuate injury due to ischemia/reperfusion or inflammation. The roles played by A,z and A; receptors in inflammation are at the present time somewhat confusing and controversial and apparently vary among species. Recent developments using pure reconstituted components have begun to refine our understanding of specific G protein subunits that participate in various adenosine receptor signaling cascades. Advances in medicinal chemistry have led to the generation of new potent and selective receptor-subtype-selective agonists and antagonists, particularly for the A,p and A; subtypes. These and additional selective agonists and antagonists that are currently under development will provide investigators in the adenosine field with important new tools to understand the receptors and cell-types that contribute to the various effects of adenosine. It is also becoming increasingly apparent that highly selective agonists or antaognists of adenosine receptor subtypes have great potential as therapeutic agents for the treatment of various inflammatory or ischemic diseases. Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Gao Z, Ni Y, Szabo G, Linden J. 1999, Palmitoylation of the recombinant human Al adenosine receptor: enhanced proteolysis of palmitoylation-deficient mutant receptors. Biochem. J. 342:387-95 . Gao Z, Robeva AS, Linden J. 1999, Purification of Al adenosine receptor-G-

protein complexes: effects of receptor downregulation and phosphorylation on coupling. Biochem. J. 338:729-36 . Linden J, Auchampach JA, Jin X, Figler RA. 1998. The structure and function of Al and

A2B adenosine receptors. Life Sci. 62:151924

ADENOSINE RECEPTORS

. Linden J, Thai T, Figler H, Jin X, Robeva

AS.

1999.

Characterization

of human

A(2B)

adenosine receptors: radioligand binding, western blotting, and coupling to G(q) in human embryonic kidney 293 cells and HMC-1 mast cells. Mol. Pharmacol. 56:705-13 . Rosin DL, Robeva A, Woodard RL, Guyenet PG, Linden J. 1998. Immunohis-

tochemical localization of adenosine A2A

receptors in the rat central nervous system. J. Comp. Neurol. 401:163-86 . Decking UKM, Schlieper G, Kroll K, Schrader J. 1997. Hypoxia-induced inhibition of adenosine kinase potentiates cardiac adenosine release. Circ. Res. 81:154-64 . Latini S, Bordoni F, Pedata F, Corradetti

R. 1999. Extracellular adenosine concentrations during in vitro ischaemia in rat hippocampal slices. Br J. Pharmacol. 127:729-39 . Harden TK, Lazarowski ER. 1999. Release

of ATP and UTP from astrocytoma cells. Prog. Brain Res. 120:135-43 . Koziak K, Kaczmarek E, Kittel A, Sevigny J, Blusztajn JK, et al. 2000. Palmitoylation targets CD39/endothelial ATP diphosphohydrolase to caveolae. J. Biol. Chem. 275:2057-62 9a. Hollopeter G, Jantzen H-M, Vincent D, Li G, England L, et al. 2001. Identification of

the platelet ADP receptor targeted by antithrombotic drugs. Nature 409:202-7 10. Jin X, Shepherd RK, Duling BR, Linden J. 1997. Inosine binds to A3 adenosine receptors and stimulates mast cell degranulation. J. Clin. Invest. 100:2849-57

File Dickenson JM, Hill SJ. 1998. Involvement of g-protein beta-gamma subunits in coupling the adenosine a(1) receptor to phospholipase c in transfected cho cells. Eur. J. Pharmacol. 355:85—93 12. Fi gler RA, Lindorfer MA, Graber SG, Garrison JC, Linden J. 1997. Reconstitution of

bovine A, adenosine receptors and G proteins in phospholipid vesicles: betagammasubunit composition influences guanine

785

nucleotide exchange and agonist binding. Biochemistry 36:16288—99

13% Yasuda H, Lindorfer MA, Woodfork KA, Fletcher JE, Garrison JC. 1996. Role of

the prenyl group on the G protein gamma subunit in coupling trimeric G proteins to Al adenosine receptors. J. Biol. Chem. 271:18588—95 14. Lindorfer MA, Myung CS, Savino Y, Yasuda H, Khazan R, Garrison JC. 1998. Differential activity of the G protein beta5 gamma? subunit at receptors and effectors. J. Biol. Chem. 273:34429-36 15 Cheng JT, Liu IM, Chi TC, Shinozuka K, Lu FH, et al. 2000. Role of adenosine in

insulin-stimulated release of leptin from isolated white adipocytes of Wistar rats. Diabetes 49:20-24 16. Sinclair CJ, Shepel PN, Geiger JD, Parkinson FE. 2000. Stimulation of nucleoside efflux and inhibition of adenosine kinase by Al adenosine receptor activation. Biochem. Pharmacol. 59:477-83 If Chen LT, Gilman AG, Kozasa T. 1999. A candidate target for G protein action in brain. J. Biol. Chem. 274:26931-38 18. Herve D, Levi-Strauss M, Marey-Semper I, Verney C, Tassin JP, et al. 1993. Golf) and Gs in rat basal ganglia: possible involvement of G(olf) in the coupling of dopamine D1 receptor with adenylyl cyclase. J. Neurosci. 13:2237-48 19. Aoyama S, Kase H, Borrelli E. 2000. Rescue of locomotor impairment in dopamine D2 receptor-deficient mice by an adenosine A2A receptor antagonist. J. Neurosci. 20:5848-52 20. Cronstein BN. 1994. Adenosine, an endogenous anti-inflammatory agent. J. Appl. Physiol. 76:5—13 Pail Sullivan GW, Linden J. 1998. Role of Ay, adenosine receptors in inflammation. Drug Dev. Res. 45:103-12 Dp Gao Z, Chen T, Weber MJ, Linden J.

1999. Anz adenosine and P2Y> receptors stimulate mitogen-activated protein kinase in human embryonic kidney-293 cells:

786

LINDEN . Okusa MD, Linden J, Huang L, Rieger JM,

cross-talk between cyclic AMP and protein kinase C pathways. J. Biol. Chem. 274:5972-80

MacDonald

He Palmer TM, Stiles GL. 2000. Identification of threonine residues controlling the

agonist-dependent

phosphorylation

and

desensitization of the rat A(3) adenosine

receptor. Mol. Pharmacol. 57:539-45 24. Palmer TM, Benovic JL, Stiles GL. 1996. Molecular basis for subtype-specific desensitization of inhibitory molecular adenosine receptors—analysis of a chime-

a2;

ric A(1)-A(3) adenosine receptor. J. Biol.

33.

Chem. 271:15272-78 Me Liang BT, Jacobson KA. 1998. A physiological role of the adenosine A3 receptor: sustained cardioprotection. Proc. Natl. Acad. Sci. USA 95:6995-99 26. Sato T, Sasaki N, O’ Rourke B, Marban E. 2000. Adenosine primes the opening of mitochondrial ATP-sensitive potassium channels: a key step in ischemic preconditioning? Circulation 102:800-5 Pile Dana A, Jonassen AK, Yamashita N, Yellon DM. 2000. Adenosine A(1) receptor activation induces delayed preconditioning in rats mediated by manganese superoxide dismutase. Circulation 101:2841-48 28. Jordan JE, Zhao ZQ, Sato H, Taft S, VintenJohansen J. 1997. Adenosine A2 receptor activation attenuates reperfusion injury by inhibiting neutrophil accumulation, superoxide generation and coronary endothelial adherence. J. Pharmacol. Exp. Ther. 280:301-9 DIS). Ross SC, Tribble CG, Linden J, Gangemi JJ, Lanpher BC, et al. 1999. Selective adenosine-A(2A) activation reduces lung reperfusion injury following transplantation. J. Heart Lung Transplant. 18:994— 1002 30. Okusa MD, Linden J, Macdonald T, Huang LP. 1999. Selective A(2A) adenosine receptor activation reduces ischemia-

reperfusion injury in rat kidney. Am. J. Physiol. Renal Fluid Electrolyte Physiol. 277:F404-12

.

35:

36.

Sif

.

TL, Huynh LP. 2000. A2A-

adenosine receptor mediated inhibition of ischemia reperfusion injury and neutrophil adhesion in rat and mouse kidneys. Am. J. Physiol. Renal Fluid Electrolyte Physiol. 279:F809-18 Cargnoni A, Ceconi C, Boraso A, Bernocchi P, Monopoli A, etal. 1999. Role of A2A receptor in the modulation of myocardial reperfusion damage. J. Cardiovasc. Pharmacol. 33:883—93 Link AA, Kino T, Worth JA, McGuire JL, Crane ML, et al. 2000. Ligand-activation of the adenosine A2a receptors inhibits IL12 production by human monocytes. J. Jmmunol. 164:43642 Fan C, Zwacka RM, Engelhardt JF. 1999. Therapeutic approaches for ischemia/reperfusion injury in the liver. J. Mol. Med. 77:577-92 Koshiba M, Rosin DL, Hayashi N, Linden J, Sitkovsky MV. 1999. Patterns of A2A extracellular adenosine receptor expression in different functional subsets of human peripheral T cells. Flow cytometry studies with anti-A2A receptor monoclonal antibodies. Mol. Pharmacol. 55:614-24 Cushley MJ, Tattersfield AE, Holgate ST. 1983. Inhaled adenosine and guanosine on airway resistance in normal and asthmatic subjects. Br: J. Clin. Pharmaco. 15:161—65 Vilsvik JS, Persson CG, Amundsen T, Brenna E, Naustdal T, et al. 1990. Comparison between theophylline and an adenosine non-blocking xanthine in acute asthma. Eur. Respir. J. 3:27—32 Ramkumar V, Stiles GL, Beaven MA, Ali H. 1993. The A3 adenosine receptor is the unique adenosine receptor which facilitates release of allergic mediators in mast cells.

J. Biol. Chem. 268:16887—90 So: Fozard JR, Pfannkuche HJ, Schuurman HJ. 1996. Mast cell degranulation following adenosine A(3) receptor activation in rats.

Eur. J. Pharmacol. 298:293-97 40. Tilley SL, Wagoner VA, Salvatore CA,

ADENOSINE RECEPTORS

Jacobson MA, Koller BH. 2000. Adenosine and inosine increase cutaneous vasopermeability by activating A(3) receptors on mast cells. J. Clin. Invest. 105:361-67 41. Auchampach JA, Jin J, Wan TC, Caughey GH, Linden J. 1997. Canine mast cell adenosine receptors: cloning and expression of the A3 receptors and evidence that degranulation is mediated by the A2B receptor. Mol. Pharmacol. 52:846-60 42. Feoktistov I, Biaggioni I. 1995. Adenosine A>, receptors evoke interleukin-8 secretion in human mast cells: an enprofylline-sensitive mechanism with implications for asthma. J. Clin. Invest. 96:1979-86 43. Shryock JC, Snowdy S, Baraldi PG, Cacciari B, Spalluto G, et al. 1998. A(2A)Adenosine receptor reserve for coro-

nary 18

vasodilation.

Circulation

787

1994. Cloning of two adenosine receptor subiypes from mouse bone marrowderived mast cells. J. Immunol. 152:4508— 15 49. Furlong TJ, Pierce KD, Selbie LA, Shine J. 1992. Molecular characterization of a human brain adenosine A> receptor. Mol. Brain Res. 15:62-66 50. Jacobson MA, Johnson RG, Luneau CJ, Salvatore CA. 1995. Cloning and chromosomal localization of the human A>, adenosine receptor gene (ADORA2B) and its pseudogene. Genomics 27:374—-76

SIE Salvatore CA, Jacobson MA, Taylor HE, Linden J, Johnson RG.

3

1993. Molecular

cloning and characterization of the human A3 adenosine receptor. Proc. Natl. Acad. Sci. USA 90:10365—69 Wilkie TM, Chen Y, Gilbert DJ, Moore KJ,

98:711-

Yu L, et al. 1993. Identification, chromosomal location, and genome organization of

44, Tucker AL, Robeva AS, Taylor HE, Hole-

mammalian G-protein-coupled receptors. Genomics 18:175-84 ae) Robeva AS, Woodard RL, Jin XW, Gao

ton D, Bockner M, et al. 1994. A, adeno-

sine receptors: two amino acids are responsible for species differences in ligand recognition. J. Biol. Chem. 269:279006

45. Baraldi

PG,

Cacciari

B, Romagnoli

R,

Varani K, Merighi S, et al. 2000. Synthe-

sis and preliminary biological evaluation of [H-3]-MRE 3008-F20: the first high affinity radioligand antagonist for the human A(3) adenosine receptors. Bioorg. Med. Chem. Lett. 10:209-11

46. Kim Y-C, Ji X-D, Melman N, Linden J, Jacobson KA. 2000. Anilide derivatives of an 8-phenylxanthiine carboxylic congener are highly potent and selective antagonists at human A2B adenosine receptors. J. Med. Chem. 43:1165-72 47. Libert F, Van Sande J, Lefort A, Czernilofsky A, Dumont JE, et al. 1992. Cloning and functional characterization of a human Al adenosine receptor. Biochem. Biophys. Res. Commun. 187:919-26 48. Marquardt DL, Walker LL, Heinemann S.

ZH, Bhattacharya S, et al. 1996. Molec-

ular characterization of recombinant human adenosine receptors. Drug Dev. Res. 39:243-52

54. Ongini E, Dionisoiti S, Gessi S, Irenius E, Fredholm BB. 1999. Comparison of CGS 15943, ZM241385, and SCH58261 as antagonists at human adenosine receptors. Naunyl-Schmiedegergs Arch. Pharmacol. 359:7-10 yD) Li AH, Moro S, Melman N, Ji XD, Jacobson KA. 1998. Structure-activity relationships and molecular modeling of 3,5-diacyl-2,4-dialkylpyridine derivatives as selective a(3) adenosine receptor antagonists. J. Med. Chem. 41:3186-201 56. Varani K, Merighi S, Gessi S, Klotz KN, Leung E, et al. 2000. [H-3 MRE 3008F20: anovel antagonist radioligand for the pharmacological and biochemical characterization of human A(3) adenosine receptors.

Mol. Pharmacol. 57:968—75

(hg = hiomsmneal yy

a

=—2

ei

A

ah

OM eee af,

-

Annu. Rey. Pharmacol. Toxicol. 2001. 41:789-813 Copyright © 2001 by Annual Reviews. All rights reserved

MOLECULAR TARGETS OF LITHIUM ACTION Christopher J Phiel and Peter S Klein Department of Medicine and Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104; e-mail: [email protected], phiel@ hhmi.upenn.edu

Key Words _ inositol monophosphatase, phosphomonoesterase, glycogen synthase kinase-3 (GSK-38), bipolar disorder, Wnt signaling @ Abstract Lithium is highly effective in the treatment of bipolar disorder and also has multiple effects on embryonic development, glycogen synthesis, hematopoiesis, and other processes. However, the mechanism of lithium action is still unclear. A number of enzymes have been proposed as potential targets of lithium action, including inositol monophosphatase, a family of structurally related phosphomonoesterases, and the protein kinase glycogen synthase kinase-3. These potential targets are widely expressed, require metal ions for catalysis, and are generally inhibited by lithium in an uncompetitive manner, most likely by displacing a divalent cation. Thus, the challenge is to determine which target, if any, is responsible for a given response to lithium in cells. Comparison of lithium effects with genetic disruption of putative target molecules has helped to validate these targets, and the use of alternative inhibitors of a given target can also lend strong support for or against a proposed mechanism of lithium action. In this review, lithium sensitive enzymes are discussed, and a number of criteria are

proposed to evaluate which of these enzymes are involved in the response to lithium in a given setting.

INTRODUCTION Lithium has been used for more than fifty years as the primary therapy for bipolar disorder (BPD) (1, 2), but its mechanism(s) of action is still unknown (3-10). Other

putatively therapeutic uses prior to the work of Cade had been described for lithium salts, and an effect of lithium on embryonic development has been recognized for at least 100 years (11). Lithium also affects metabolism, neuronal communication, and cell proliferation in a diverse array of organisms, from cellular slime molds

to humans. Some attempts have been made to explain the mechanism of lithium action in these diverse settings through a single unifying hypothesis. These efforts have guided valuable research on the pharmacology of lithium action, but it remains unclear whether these or other possible mechanisms are sufficient to explain any or all of the effects of lithium. In this review, we focus on potential targets that are known to be directly inhibited by lithium in vitro. The vast and expanding 0362-1642/01/0421-0789$14.00

789

790

PHIEL

® KLEIN

literature on downstream consequences of lithium action cannot be adequately reviewed here; for this and related topics, readers are referred to several excellent reviews (3-7, 12; see also 9, 10).

The clinical importance of defining the direct target(s) of lithium action is twofold. First, while lithium is highly effective in treating BPD in many patients, the therapeutic window between effective dosing and toxicity is narrow, side effects are common even within the therapeutic dose range, and a significant number of patients do not respond (3, 8). While anticonvulsants such as valproic acid offer an alternative mode of therapy, an understanding of the targets of lithium (and of valproic acid) will make it possible to identify additional therapies for this common disorder. Second, little is known about the pathogenesis of BPD or other mood disorders, and therefore identification of the molecular target of lithium should

shed light on the etiology of this disorder. Finally, it is of great scientific interest to understand how a simple, small cation like lithium can demonstrate such relative specificity in its range of actions when other monovalent cations have no apparent effect. It is indeed remarkable that millions of people can actually tolerate this simple drug with a minimum of prohibitive side effects. Lithium is clearly able to inhibit multiple enzymes, and it is improbable that all of the actions of lithium can be explained by interaction with a single target. Since these diverse potential targets are often present in the same cell, it remains a challenge to distinguish which, if any, of the known targets of lithium is involved in the in vivo response to lithium. Thus, after a brief discussion of some of the better characterized effects of lithium, we also discuss criteria that can be applied to validate a putative target in a given setting.

EFFECTS OF LITHIUM Lithium has numerous effects in humans and in model organisms; it would be difficult to describe all of them in detail here. A few of the more common effects of lithium are described to provide a physiological context for the discussion of the potential molecular targets of lithium action that follows.

Neuropsychiatric Bipolar disorder is a common psychiatric ailment characterized by cycling periods of extreme elation (mania) and severe depression (9); features of psychosis,

including delusions and hallucinations, can be associated with either extreme. Although lithium salts were used in the nineteenth century for various,and sometimes dubious purposes, for example as a soporific and a gout treatment (2, 13— 15), the use of lithium to treat the manic phase of BPD was first described by

Cade in 1949, after noticing the sedative effect of lithium on guinea pigs (1). Cade boldly extended his observations to humans with BPD, providing a potent

MOLECULAR TARGETS OF LITHIUM ACTION

791

therapy for BPD. The occurrence of BPD has been estimated to be in 0.3% to 1.5% of the population, based on epidemiological studies sampling diverse regions of the world (16). Although a genetic link is suspected given the clustering of BPD in genetically homogenous populations (17—21), specific loci have not been identified (22, 23). In addition to controlling mania, lithium has been used

as a mood stabilizer in the control of bipolar depression and may also have a therapeutic value as adjunctive therapy in the treatment of unipolar depression (8,24) and in cluster headache. Lithium in cell culture models confers protection from excitotoxic neurotransmitters in cortical and cerebellar cells and leads to synaptic remodeling in cerebellar cells (25-27). As described below, lithium can cause an increase in neurotransmitter release in Drosophila neuromuscular junctions.

Developmental Effects Lithium alters the development of phylogenetically diverse organisms (Figure 1) (11, 28-32). For example, in Dictyostelium, a simple eukaryotic organism, exposure to lithium during early development blocks spore cell fate and promotes the formation of stalk cells (28,29). In sea urchins, lithium causes vegetalization of animal blastomeres (32-34). In vertebrates such as Xenopus and zebrafish, lithium causes expansion of dorsal mesoderm, leading to duplication of the dorsal axis in Xenopus or, in extreme cases, entirely dorsalized embryos lacking ventral tissues (30, 31). Inmammals, isolated nephrogenic mesenchyme undergoes mesenchymal to epithelial differentiation when exposed to lithium (35), and mouse mammary tumor cells have an increased proliferative index in the presence of lithium (36). In humans teratogenic effects of lithium have been reported at surprisingly low frequency when one considers the dramatic effects on the development of lower vertebrates. Some studies have indicated an increased frequency of congenital heart defects, particularly Ebstein’s anomaly, characterized by downward displacement of the tricuspid valve in the right ventricle of the heart, but the reported frequency of this defect has varied widely (9).

Metabolic Effects Lithium can stimulate glycogen synthesis in mammals through activation of glycogen synthase, mimicking insulin action (37-41). In addition, lithium therapy in humans is associated with subclinical hypothyroidism and nontoxic goiter, nephrogenic diabetes insipidus (decreased renal concentrating ability), weight gain, hyperparathyroidism, and a large number of other less common side effects (8, 9).

Hematopoiesis One of the most prevalent benign side effects of lithium therapy in humans is an increase in the number of circulating granulocytes (up to 1.5-fold), predominantly

792

PHIEL

A.

® KLEIN

Xenopus

embryos

control

B.

Dictyostelium

spores stalk control

Figure 1 The effects of lithium on developing systems. (a) Exposure of Xenopus embryos to lithium during an early stage of development causes expansion and duplication of dorsal and anterior structures (30). On the left is a control tadpole. On the right, a lithium-treated embryo with duplicated dorsal and anterior structures.

(b) Dictyostelium discoideum is a simple eukary-

ote that, upon starvation, develops into two general cell types, spores (within a fruiting body) and stalk cells (supporting the fruiting body). Exposure to lithium during development diverts cells away from the spore cell fate with expansion of the stalk cell population (adapted from Reference 28).

neutrophils, although effects on other lineages have also been reported (42-44). Lithium appears to increase the level of pluripotent hematopoietic stem cells, either indirectly by stimulating the release of cytokines or directly by acting on stem cells (or both) (44-48). Thus, lithium treatment reduces chemotherapy-induced neutropenia and febrile complications of marrow suppressive therapies in a number of clinical trials (49-52) (reviewed in 44). In spite of these early encouraging studies, lithium has not been used extensively in neutropenic patients, perhaps because of its narrow therapeutic window.

MOLECULAR TARGETS OF LITHIUM ACTION

793

MOLECULAR TARGETS OF LITHIUM Considerable information has been gained in the past 50 years concerning some of the indirect physiological consequences of lithium therapy (3,4,6,53—56). Lithium can affect neurotransmitter release, metabolism of biogenic monoamines, and neuronal signal transmission through perturbation of the distribution of sodium, magnesium, and calcium (9, 10). Lithium can inhibit depolarization-induced and

calcium-dependent release of norepinephrine and dopamine, (57), and conversely, may stimulate the release of serotonin (S58). A number of phosphoproteins have been identified whose phosphorylation or expression level is sensitive to lithium treatment, including neurofilament proteins (59), microtubule-associated proteins (60), and protein kinase C (PKC) substrates such as the MARCKS

protein, which

is downregulated in response to lithium therapy (4, 6,53). Lithium also inhibits ADP-ribosy] transferase in extracts derived from rat frontal cortex (61). Because of the characteristic delay in clinical response to lithium therapy, an effect of lithium on neuronal gene expression has also been proposed (56): In support of this, lithium has been shown to activate AP-1-dependent transcription (62-67)

and Tcf/Lef-dependent transcription (see below). A number of studies have indicated that lithium interferes with signal transduction through G protein-coupled pathways, inhibiting the G proteins themselves or downstream effectors, including adenyly! cyclase, phospholipase C, and protein kinase C (4, 68). However, in each case, these effects of lithium have not been shown to be direct, largely

because of the difficulty in reconstituting these signaling systems from purified components. These observations are not necessarily in conflict with each other, and each could lie in a pathway regulated by a common signaling molecule that is the direct target of lithium action. Our focus ison targets known to be inhibited directly by lithium (Table 1). These include inositol monophosphatase (69) and the large family of related phosphomonoesterases (70), as well as the recently identified target glycogen synthase kinase-38 (GSK-3) (71). Since each of these potential targets is expressed widely, and generally in the same cell types, it remains a

TABLE 1

Lithium-sensitive enzymes

Target

Ki(mM)

Type of inhibition

Reference(s)

IMPase

0.8

Uncompetitive

1

IPPase

0.3

Uncompetitive

91

FBPase

_0.3-0.8

BPntase

0.16-0.3

Hal2p GSK-3B

0.1 © 1-2

Uncompetitive

.

193 85, 90

Un-/noncompetitive

86

Uncompetitive

71, 101

i

794

PHIEL

® KLEIN

challenge to distinguish which, if any, are involved in a given response to lithium. To validate a molecule as a relevant direct target of lithium action, one would expect the following: 1. The target should show direct sensitivity to lithium (and not to other monovalent cations) in vitro.

2. The target should be sensitive to lithium in vivo at physiological concentrations of lithium. 3. Loss-of-function mutations in the putative target should be phenocopied by exposure to lithium. 4. Alternative inhibitors of the target molecule should mimic the effect of

lithium. 5. The effect of lithium should be reversed by appropriate modulation of components downstream of the putative target. 6. A lithium-resistant form of the molecule should block the effect of lithium in Vivo.

Inositol Depletion Hypothesis One of the most compelling hypotheses to explain lithium action is the inositol depletion hypothesis (Figure 2), which is based on the observation that lithium inhibits inositol monophosphatase (IMPase) in vitro at a Ki (0.8 mM) within the therapeutic range (0.5 to 1.5 mM) for lithium treatment of patients with bipolar disease (69, 72). Since IMPase regenerates inositol from inositol monophosphate (IMP), inhibition of this step could deplete inositol if the cells do not have an alternative source. In principle, this should lead to depletion of phosphatidylinositol bisphosphate (PIP), a necessary precursor for the generation of the second messenger inositol-1,4,5 trisphosphate (IP) in response to extracellular signals. For example, numerous neurotransmitters bind to G protein—coupled receptors that activate phospholipase C, which hydrolyzes PIP, to diacylglycerol (DAG) and IP,; DAG activates PKC, while IP; causes release of calcium from intracellular stores into the cytoplasm. The net effect of lithium, then, would be to block liganddependent signaling through PKC and IP;/calcium. Lithium is an uncompetitive inhibitor of IMPase, and it has been argued that this may explain why lithium therapy is effective in BPD but has no apparent effect on the psychiatric state of normal subjects (73). For an uncompetitive inhibitor, the fractional inhibition is proportional to the level of enzyme-substrate complex, which means that at higher concentrations of substrate (e.g. IMP), lithium would have a greater effect. Thus, if mania or other disturbances of mood arise because of excess signaling that leads to increased IMP levels, lithium would inhibit IMPase to a greater extent than if given to “normal” patients with lower levels of IMP. It follows, then, that an uncompetitive inhibitor cannot be overcome by increased concentration of substrate, in contrast to a competitive inhibitor.

MOLECULAR TARGETS OF LITHIUM ACTION

795

au “—~—— INOSITOL iIMPase

PKC

[Ca?*]j

Figure 2 Inositol depletion hypothesis. Ligand binding to a surface receptor activates phospholipase C (PLC), which hydrolyzes the phospholipid PIP to yield two second messengers: diacylglycerol (DAG) and inositol-1,4,5 trisphosphate (IP3). As shown in this simplified model, lithium inhibits IMPase, which regenerates inositol from inositol monophosphate (IP). If this inhibition is sufficient to deplete inositol, then it should also deplete PIP and prevent the formation of IP; and DAG, thus indirectly inhibiting transmembrane signaling. The bisphosphonate compound L-690,330 is an IMPase inhibitor that is 1000-fold more potent than lithium.

The inositol depletion hypothesis is supported by observations that lithium can inhibit IMPase in vivo (74), leading to accumulation of IMP. However, reduction

of inositol in mammals is seen only at toxic doses of lithium, therapeutic doses in mammals do not deplete inositol, PIP;, or IP; in vivo, either acutely or after chronic administration

(3,4). Numerous

studies have shown

inositol depletion

in brain slices treated with lithium, but in these in vitro assays, a marked drop in inositol reserves is incurred even before addition of lithium (3, 4). Still, therapeutic

concentrations of lithium could lower inositol levels within small, restricted regions of the brain or within specific subcellular pools that are particularly sensitive to inhibition of IMPase, and this might not be detected with the available assays for inositol and inositol phosphates. The inositol depletion hypothesis has also been invoked to explain the profound effects of lithium on developing organisms (29,75). Indeed, teratogenic doses of lithium inhibit IMPase in vivo in Xenopus embryos (71) and cause a 30% reduction in inositol levels (76), while in Dictyostelium, lithium lowers intracellular

inositol by 20% (77). In spite of these small changes in inositol levels, coinjection

796

PHIEL

® KLEIN

of inositol can prevent this teratogenic effect of lithium in Xenopus, providing additional support for the inositol depletion hypothesis (75). Although the inositol depletion hypothesis is an attractive model for explaining the effects of lithium in some settings, a closer examination of this hypothesis casts doubt on whether inhibition of IMPase is sufficient to explain the developmental effects of lithium (71, 78). Alternative inhibitors of IMPase (bisphosphonates) that

are 1000-fold more potent than lithium (79) have no effect on the development of Xenopus embryos despite complete inhibition of IMPase in vivo (71). This observation suggests that inhibition of IMPase is not sufficient for the developmental effects of lithium in Xenopus. Why then does coinjection of inositol with lithium reverse the developmental effects of lithium in Xenopus (as above)? Elevated levels of inositol or inositol phosphates may have unexpected and indirect effects on other lithium-sensitive pathways; in support of this, raising the level of inositol in Xenopus blocks dorsal axis induction by agents other than lithium that are unlikely to act through depletion of inositol, as discussed below (78). In addition, agonist-induced increases in IP; can be blocked in slime molds by disruption of the gene encoding phospholipase C (although IP; is still detectable at basal levels), yet this mutation has no effect on Dictyostelium development (80), despite the dramatic effect of lithium treatment. Furthermore, Dictyostelium has at least three forms of IMPase, two of which are insensitive to lithium, which may explain the modest effect of lithium on inositol levels (20% reduction) in this organism (77). Taken together, these observations suggest that the developmental effects of lithium are mediated through a target distinct from IMPase, although the possibility remains that inhibition of IMPase is necessary but not sufficient for lithium action in these settings.

Phosphomonoesterases Other Than IMPase In addition to IMPase, a number of structurally related phosphomonoesterases have been described that require metal ions (especially magnesium) and are inhibited by lithium (70). Biochemical and crystallographic analyses of three of these, IMPase, inositol polyphosphate |-phosphatase (IPPase), and fructose 1,6-bisphosphatase (FBPase), have revealed similar tertiary structures and a consensus sequence (D-X,-EE-X,-DP(i/1)D(s/g/a)T-X,-WD-X, ,-GG) that is involved in metal ion binding, which plays a role in catalysis, and that most likely interacts with lithium: ions (70, 81-84). A number of other phosphomonoesterases contain this consensus sequence, including Hal2p and Toll (from budding yeast and fission yeast, respectively), SALI (from Arabidopsis), the bacterial protein cysQ, and related bisphosphate 3’ nucleotidases (BPntase) from mammals (70, 85-90), defining a large family of lithium-sensitive enzymes. IPPase catalyzes the removal of the 1-phosphate from inositol 1,4-bisphosphate or 1,3,4-trisphosphate to form inositol 4-phosphate or inositol 3,4-bisphosphate (83,91). An in vivo correlation between lithium action and inhibition of IPPase is found in Drosophila neuromuscular junctions, where lithium exposure pheno-

copies mutations in IPPase, both of which lead to a similar defect in synaptic

MOLECULAR TARGETS OF LITHIUM ACTION

797

transmission due to increased vesicle release (92). The fact that lithium inhibits

IPPase in vitro offers a likely explanation for the effects of lithium in this setting. It would be interesting to test whether overexpression of [PPase or expression of a lithium-resistant form of IPPase prevents the effect of lithium on the Drosophila neuromuscular junction and in other settings. HAL2 was identified in Saccharomyces cerevisiae as a gene that, when over-

expressed, confers tolerance to high-salt conditions (93). (HAL2 is identical to MET22, which is required for methionine synthesis (94).) Hal2p removes the 3’ phosphate from adenosine 3’, 5’ bisphosphate (pAp) and is inhibited by lithium at submillimolar concentrations (IC50 = 0.1 mM (86)). Although lithium toxicity in yeast requires 100-200 mM LiCl in the growth medium (93, 95), it is not

clear what the intracellular concentration is under these conditions. Inhibition of Hal2p by salt (lithium or sodium) or loss-of-function mutations in HAL2 leads to

the accumulation of pAp, which in turn causes inhibition of sulphotransferases and RNA processing enzymes such as exoribonuclease Xrn1p (95, 96). Hal2p was recently crystallized, and this structure, combined with genetic analysis, led to the identification of residues important in binding metal ions and in mediating inhibition by monovalent cations (96). In fact, mutations in Hal2p were identified that reduced sensitivity to monovalent cations; corresponding mutations were generated in human IMPase and were found to reduce sensitivity to lithium almost ninefold (96). It will be fascinating to test whether transgenic expression of lithiumresistant IMPase (or other phosphomonoesterases, if possible) confers resistance to lithium in model organisms or cell lines. BPntase (also identified as RnPIP) is a bisphosphate 3’-nucleosidase with activity toward pAp and other bisphosphate nucleotides, as well as toward 3’ phosphoadenosine 5’ phosphosulfonate (85,89, 90). BPntase/RnPIP is dependent on magnesium and is uncompetitively inhibited by lithium with a low Ki (157— 300 4M). Since BPntase/RnPIP is structurally related to Hal2p and SAL1, which

are involved in salt tolerance in yeast and plants, Speigelberg et al have proposed that inhibition of BPntase might account for lithium-induced nephrogenic diabetes insipidus observed in mammals, a conditioned characterized by poor concentration of urine with consequent polyuria. BPntase/RnPIP also remove the 1-phosphate from inositol 1,4 bisphosphate (85, 89, 90), similar to SALI (88).

In summary, a large family of lithium-sensitive phosphomonoesterases has been described in a broad range of organisms; these enzymes offer attractive targets to

consider as potential targets of lithium action, and in a few cases are associated

with lithium effects in vivo, particularly with IPPase in Drosophila, where loss of function is phenocopied by lithium, and with Hal2p in S. cerevisiae, in which overexpression prevents the effects of high lithium or sodium concentrations.

Glycogen Synthase Kinase-3 A novel hypothesis to explain the effects of lithium action on embryonic development, glycogen synthesis, and hematopoiesis has been proposed based on the

798

PHIEL

® KLEIN

observation that lithium is a direct inhibitor of glycogen synthase kinase-3 (GSK-3) (71). GSK-3 was first identified as a negative regulator of glycogen synthesis that phosphorylates and inhibits glycogen synthase (97, 98). Two isozymes of GSK-3, GSK-3a and GSK-38, have been identified (99). Although published work has focused on GSK-36, GSK-3a has similar but not identical biochemical properties (98 but see also 100). GSK-3£ is expressed broadly in eukaryotes, including protozoans such as yeast and Dictyostelium, higher plants such as Arabidopsis, rice, and corn, invertebrates such as Caenorhabditis elegans, Drosophila, and sea urchins, and vertebrates including Xenopus, mice and humans. In mammals, GSK-3f is expressed in early embryos and in most adult tissue types, including brain (98, 99). GSK-3£ is inhibited by lithium with a Ki (1-2 mM) within the effective range for lithium action (71). Lithium does not inhibit other protein kinases tested, in-

cluding casein kinase II, protein kinase A, p34°*?, MAP kinase, and protein kinase C. Lithium is an uncompetitive inhibitor of GSK-3£ with respect to substrate (71), as seen with IMPase and other phosphomonoesterases but appears to be competitive with respect to magnesium, which may explain its mechanism of inhibition (LJ Conrad & PS Klein, unpublished data). Lithium inhibition of GSK-3£, as well as GSK-3a, has subsequently been confirmed in vitro (101).

Lithium also

inhibits GSK-3£ in vivo, as demonstrated in a number of systems by the reduced phosphorylation of known GSK-3 substrates such as tau protein, MAP-1B, and others (62, 101-106). Furthermore, lithium inhibits GSK-3 derived from diverse species, from Dictyostelium to mammals (33, 62, 101). How then does this inhibition of GSK-3 explain the mechanism of lithium action in developmental settings? Strong support for this hypothesis comes from genetic disruption of GSK-38 in Dictyostelium (GSKA). Loss of GSKA in Dictyostelium alters cell fate so that presumptive spore cells are diverted to a stalk cell fate (107). The phenotype described by Harwood et al (107) is remarkably similar to the effect of lithium described by Maeda 25 years earlier (28), and thus, lithium phenocopies loss of GSKA in this setting. This was confirmed by demonstrating that lithium

inhibits GSKA (62, 108). In higher eukaryotes, GSK-3 is a negative regulator of the Wnt signaling pathway (Figures 3, 4), which plays a central role in the patterning of early embryos (109-113), in the regulation of cell proliferation (114-116), and in neuronal signal transduction in adults (27). In the absence of Wnt signaling, GSK-38 is found in a complex with several other proteins, including the adenomatous polyposis coli protein (APC), axin, protein phosphatase 2A (PP2A), dishevelled (dsh), and 6-catenin (117). This multiprotein complex promotes GSK-36-mediated phosphorylation of 6-catenin, targeting 6-catenin for ubiquitination and degradation via the proteosome pathway. Upon activation of Wnt signaling, GSK-3£ is inhibited (118-120), allowing B-catenin protein to accumulate (121). Accumulate d

6-catenin translocates to the nucleus where it forms a transcriptionally competent complex with members of the Tcf/Lef family of DNA binding factors, resulting in the transcription of Wnt target genes (109, 110). Perturbations that interfere with GSK-3B—mediated phosphorylation of B-catenin, such as null mutations in the GSK-3£ gene, known as zeste-white-3 or shaggy

MOLECULAR TARGETS OF LITHIUM ACTION

799

nucleus

Figure 3 The Wnt signaling pathway. Wnts are a family of secreted glycoproteins involved in embryonic patterning, axonal remodeling, regulation of cellular proliferation, stem cell development, and numerous other processes. In the absence of Wnt, GSK-36 phosphorylates -catenin, resulting in its rapid degradation. Binding of Wnt to the frizzled receptor leads to inhibition of GSK-38 and stabilization of B-catenin, which accumulates in the nucleus where it interacts with Tcf/Lef DNA binding factors to activate downstream target genes. Lithium activates Wnt signaling by inhibiting GSK-36 directly. (Arrows indicate positive effect and horizontal or inverted T indicates inhibitory effect.)

800

PHIEL

® KLEIN

glycogen’ synthase-P (reduced activity)

B-catenin-P

(unstable)

GSK-3B

, terminal half-life; F, bioavailability; PK, pharmacokinetics (parameters); C,nin.gs, trough concentration at steady state.

“Normalized for body weight.

Interethnic differences in CYP3A4-mediated drug metabolism have been studied in vitro and in different populations. Caucasian liver microsomes had higher nifedipine oxidase activity and significantly higher testosterone 66-hydroxylase activity than Japanese samples (n = 30 each). Hepatic P-450 3A (principally P-450 3A4) content correlated well with nifedipine oxidation (r = 0.79) and testosterone 68-hydroxylation (r = 0.81) activities. Also, CYP3A4-mediated metabolic activation of aflatoxin B1 and sterigmatocystin correlated well with microsomal P-450 3A content (r = 0.78 and 0.83, respectively) and was significantly higher in Caucasian than in Japanese samples (15). As summarized in Table 4, the apparent oral clearance of alprazolam was found to be similar in native and American-born Asians, but to be significantly higher in Caucasians than in Asians

(150). Similarly, the area under the plasma concentration—time curve of nifedipine was significantly higher in Asians than Caucasians (155, 157-159). The ability to metabolize oral nifedipine was similar in Asian Indians and Malaysians who

resided in the same geographic area (160) and codeine N-demethylation, mediated by CYP3A4 (164), was more extensive in Caucasian than Chinese subjects (106). These data suggest that CYP3A4 activity may be higher in Caucasians than other

populations. However, such an ethnic variation may be substrate-dependent, since erythromycin N-demethylation did not show a good correlation with the content of

hepatic P-450 3A (r = 0.28) (15), and no difference was present between Asians and Caucasians when erythromycin (155), triazolam (163) or cerivastatin (151)

were used as metabolic markers of CYP3A4 activity.

ETHNIC VARIATION IN DRUG RESPONSE

825

Conflicting results have been found when CYP3A4-catalyzed oxidation of nifedipine (161, 162) was compared in black and white subjects. The metabolism of cerivastatin (151) and midazolam

(156) does not differ between blacks and

whites. Based on published data (see Table 4), ethnic comparisons with different CYP3A4 substrates have yielded inconsistent results that are difficult to interpret and may reflect an interplay of genetic and environmental factors. An early in vivo study of nifedipine oxidation showed an apparent bimodality of the area under the plasma concentration—time curve of nifedipine after a 20 mg oral dose in 53 healthy Dutch individuals (165). Subsequent studies, however, in a larger number of subjects, did not confirm this finding (137, 147, 148, 158, 166),

suggesting that nifedipine’s metabolism is unlikely to be governed by genetic variations that result in the presence or absence of enzyme activity, such as occurs with CYP2C19 and CYP2D6.

The CYP3A4 gene was mapped on chromosome 7q7*! (167-171) and found to be ~27 kb long, with 13 exons and 12 introns (172). The sequence of CYP3A4

cDNA obtained from hepatic libraries has been extensively examined (GenBank accession numbers: M18907 for the CYP3A4 cDNA, and D11131 for its 5/-flanking region) (168, 172-176). To date, at least four

CYP3A4 allelic variants

have been identified (for more information on the nomenclature, go to http://www.

imm.ki.se/CYPalleles/cyp3a4.htm). The first common CYP3A4 allelic variant is an A to G transition in the 5’ promotor region at position —290 (from the transcription initiation site) (177), altering the 10-bp nifedipine-specific response element localized at —287 to —296 of the 5’ regulatory region (172). This allele (previously termed CYP3A4-V and now designated as CYP3A4*/B) was found to be more common in patients with prostate cancer of a more invasive clinical stage than patients with a low-level clinical stage (177), and to be over-represented in patients with leukemia (178). Recent studies indicate that this allelic variant results

in a modest reduction in hepatic CYP3A4 activity (156) but does not significantly alter the metabolism of CYP3A4 substrate drugs (179-181), although there are marked differences in frequency among various ethnic populations (181a). The frequency of CYP3A4*1B was low in white and Hispanic subjects (3.6-11.0%) (177, 179, 181a, 182), absent in Chinese and Japanese subjects (179, 182, 183),

and much higher in black subjects (53.0-69.0%) (179, 181a, 182, 183).

A second allelic variant of the

CYP3A4 gene was found in exon 7 (a Ser***Pro

CYP3A4*2 (183). This allele was uncommon in 55 substitution) and designated not observed in black or Chinese groups of similar was and (2.7%), subjects white cted cDNA expression system, the intrinsic baculovirus-dire a Using size (183). was decreased approximately six- to oxidation nifedipine for in) (Vnax/K clearance with the wild-type enzyme, but was compared enzyme variant the ninefold with n (183). The C YP3A4*2 66-hydroxylatio testosterone for different not significantly ent altered kinetics, but substrate-depend with enzyme variant a allele may encode substantially to contribute to likely not is variant this because it is uncommon, in the differences ethnic together, Taken activity. ethnic variations in CYP3A4 inconsistent, and characterized poorly are drugs disposition of CYP3A4 substrate and currently recognized molecular variations in the CYP3A4 gene do not appear to

826

XIE ET AL

contribute substantially to interindividual variability in the disposition of CYP3A4 substrate drugs.

DRUG TRANSPORTER

P-Glycoprotein There is increasing evidence that drug metabolism alone does not account for the observed interindividual variability in drug disposition or response (184), but that other processes, including drug transport, are important determinants of drug disposition. Although a number of drug transporters have been shown to play a | key role in drug disposition (184-186; RB Kim & GR Wilkinson, submitted for publication), P-glycoprotein (P-gp), the MDR/ gene product, is one of the best studied and characterized. Although initial interest in P-gp focused on its role as a mediator of multidrug resistance (MDR) in tumor cells, recent studies have demonstrated its more general role in drug disposition (184-188). P-gp is expressed in many tissues other than tumor cells, particularly those associated with excretory function (189), such as the canalicular domain of hepatocytes and the (luminal) brush border membrane of both intestinal epithelial cells and renal proximal tubule cells. P-gp is an ATP-dependent drug efflux pump, actively transporting many structurally diverse compounds from the inside to the outside of cells against a concentration gradient. Its apical distribution in cells results in decreased drug absorption from the gut lumen and enhanced drug excretion into bile and urine. Moreover, expression of P-gp in the capillary endothelium of the blood-brain barrier prevents penetration of substrate drugs into the central nervous system. Accordingly, P-gp plays an important role in drug absorption, distribution, and excretion. Of importance for drug disposition is that P-gp and CYP3A4 are frequently co-expressed in the same cells and share a large number of substrates and modulators (139).

The disposition of such drugs is thus affected by both transport and metabolism (184). Variability in P-gp-mediated drug transport in the gastrointestinal tract alters the oral bioavailability (F’) of P-gp substrates (7, 186, 190). Such effects are seen most easily when either hepatic or extrahepatic drug metabolism is negligible (188). Ethnic variation in P-gp activity has not been widely studied. Lindholm et al investigated the effects of demographic factors on the pharmacokinetics of cyclosporine, a drug that is a substrate for both P-gp and CYP3A4, in 187 kidney transplant recipients, and found that the oral bioavailability of cyclosporine was significantly lower in blacks (n = 58, mean = 30.9%) than whites Gay = 86, mean = 39.6%) or Hispanics (n = 40, mean = 42.1%), with no ethnic variation in clearance and volume of distribution at steady state (V,.) (153). Be-

cause a 10-fold variation in the levels of intestinal CYP3A4 had no clear effect on oral cyclosporine pharmacokinetics, Lown and colleagues postulated that intestinal P-gp transport activity was the major determinant of bioavailability and

ETHNIC VARIATION IN DRUG RESPONSE

827

Cyax Of cyclosporine (190), so that patients with lower levels of intestinal P-gp had higher bioavailability and higher C,,,,, and vice versa. Furthermore, the concentration/dose ratio of tacrolimus (also a substrate of CYP3A4 and P-gp) was correlated with the mRNA expression of MDRI1 but not CYP3A4 (191). Like

cyclosporine (153, 154, 192, 193), higher doses of tacrolimus were required in blacks than whites to attain similar plasma levels (194), suggesting a lower oral bioavailability of tacrolimus. Although there is no supporting evidence, one explanation for the lower bioavailability of cyclosporine and tacrolimus would be greater P-gp-mediated drug transport in blacks. We found no difference in the disposition of cyclosporine in healthy black and white men studied on a controlled diet (CM Stein, AJ Sadeque, JJ Murray,

C Wandel, RB Kim, et al, manuscript submit-

ted). Recently, we found that the oral clearance of fexofenadine (a P-gp substrate that is not metabolized) exhibited ~10-fold interindividual variation and tended to be slightly higher in white women than in black women (195). However, addi-

tional studies with larger sample sizes will be required to define the relationship between ethnicity and P-gp activity. The interrelationship between CYP3A4 and P-gp, and the effects of environmental factors such as diet, make defining ethnic differences in P-gp-mediated drug disposition difficult. The MDRI gene encoding P-gp is located on chromosome 7q”! (196), with 28 exons encoding a protein of 1280 amino acids. Significant information about the structure-function analysis of P-gp has recently been summarized (197). Some naturally occurring polymorphisms of the MDR/ gene have been found to correlate with potential clinical effects (198), or with the levels of intestinal MDR/ expression and uptake of orally administered digoxin (a substrate of P-gp) (188). We are currently examining the hypothesis that allelic variants of MDR1 might be associated with interindividual or interethnic variations in the disposition of P-gp drug substrates. Using PCR-based single-stranded conformational polymorphism (SSCP) methods, we found that a number of single nucleotide polymorphisms exist in multiple exons and the 5’-flanking promotor region of the MDR/ gene in Japanese, black, and white American populations, and that these point mutations are distributed differently among the ethnic groups (198a). Indeed, genetic variability in MDR/ appears to be more frequent than previously thought. Future genotype-phenotype relationship studies may provide additional insights into the role of P-gp as a determinant of interindividual or interethnic variability in drug

response.

DRUG RECEPTORS

a-Adrenergic Receptor The a-adrenergic receptor (a-AR) family comprises two subfamilies (a )-AR and a5-AR). Three subtypes of each have been identified pharmacologically and through molecular cloning: o;, (formerly ac), @)p, jp (formerly @&;ayp), @2a; 5p, and ac (199, 200). Evidence suggests that the human a; ,-AR predominates

828

XIE ET AL

in arteries (201), whereas all three ~;-AR subtypes (in particular a}, and @),) are expressed in veins (201) and peripheral blood lymphocytes (202). The major a,-AR subtypes mediating vasoconstriction and regulating peripheral vascular resistance are a), and @j,p. Several studies have demonstrated that blacks have greater vascular reactivity in response to a-adrenergic stimuli than whites (203—208). The vascular responses to intrabrachial artery infusion of phenylephrine (an a ,-AR agonist) and to cold stress (a stimulator of endogenous norepinephrine release) were compared in AfricanAmerican and Caucasian normotensive men (207—208) and @ ;-AR-mediated vasoconstrictor reactivity was significantly increased in blacks. The response in the superficial dorsal hand vein to local infusion of phenylephrine was reported to be blunted in normotensive blacks (209). The different results may be due to the site of drug action (artery versus vein), because the responsiveness of different types of blood vessel differs quantitatively (210). Known polymorphisms of the human @ ;p-AR are rare and appear to not be associated with the interindividual variations in response to phenylephrine (211, 212). Although African Americans had a significantly lower frequency of the amino

acid variant Cys*”” of the w;,-AR than Caucasian Americans, this polymorphism was not associated with hypertension and its effects on sensitivity to phenylephrine are not known (213). At present, the mechanisms underlying the increased a-adrenergic vascular sensitivity in African Americans are unknown.

B-Adrenergic Receptor Three different B-adrenergic receptor (B-AR) subtypes have been cloned and phar-

macologically characterized: B,, B>, and B; (200, 214). The presence of a putative fourth B-AR subtype (64) has been proposed based on recent pharmacological studies in human and rat cardiac tissue (200). Although both 6,-AR and £-AR subtypes co-exist in the human cardiovascular system (200, 215), 6;-ARs predominate. Many studies have revealed that 6 ,-AR-mediated effects include exerciseinduced increase in heart rate and systolic blood pressure, as well as renin release (for reviews, see 215-217), whereas 6,-AR-mediated responses include a decrease in total peripheral resistance and diastolic blood pressure (215). Bl-Adrenergic Receptor (B1-AR) Ethnic differences in B-AR-mediated responses to drugs have been extensively investigated among Caucasians, East Asians, and blacks of African descent (4). Compared with Caucasian men, Chinese men had a greater sensitivity to the effects of propranolol, a nonselective B-AR antagonist, which produced a greater reduction in mean arterial blood pressure and exercise-induced tachycardia (218) and greater suppression of exercise-induced

plasma renin release (a 6 ;-AR-mediated effect) (219).

By contrast, normoten-

sive blacks had decreased sensitivity to isoproterenol (a nonselective B-AR agonist) compared with whites, before and after 6-blockade with propranolol (220). Furthermore, clinical observations from many investigations have indicated that

ETHNIC VARIATION IN DRUG RESPONSE

829

black patients with hypertension respond less well to monotherapy with several B-AR blockers (221), including the nonselective 6-blockers propranolol (222) and nadolol (223), and the 6 ,-selective blocker atenolol (222—226). Decreased sensi-

tivity to B-AR antagonists may be associated with the lower levels of plasma renin activity and higher proportion of low-renin hypertension found in blacks (4). The human £,-AR is a protein of 477 amino acid residues encoded by an in-

tronless gene (227, 228) localized on chromosome 10q***° (227). Recently, 18 single nucleotide polymorphisms have been identified in the human 6 ;-AR gene, 17 of which are located in the N-terminal and C-terminal region of the coding exon, resulting in 7 amino acid substitutions (229). Two common allelic vari-

ants of the human 6,-AR gene, A'?G (or Ser*Gly) and C!!®G (or Arg**°Gly), were identified (228-234). The first variant Gly”? was associated with a decreased mortality risk in patients with congestive heart failure (230) and was observed significantly more frequently in a group of patients with idiopathic dilated cardiomyopathy (229). However, no ethnic differences in the frequency of this allelic variant existed among African Americans,

Caucasians,

and Chinese (233).

A

second common variant Gly**’ receptor was found to have a decreased receptorG,-protein interaction and reduced cyclic AMP production following exposure to agonist (232), suggesting that this variant receptor exhibits diminished response to a 6,-AR agonist in vitro. In vivo studies to determine the functional sig-

nificance of the Arg*®’Gly 6,-AR polymorphism in humans are underway, and a preliminary population-based case-control study shows no association of this polymorphism with essential hypertension in African Americans or Caucasian Americans (234a). The frequency of the variant Gly**” receptor is significantly higher in African Americans (42%) than in Caucasian Americans (25%), Chinese

(27%), or Hispanics (33%) (234b). The physiological significance of ethnic dif-

ferences in the frequency of the Gly**? variant f ;-AR, which is characterized as a loss-of-function polymorphism in vitro (232), is uncertain.

f2-Adrenergic Receptor (B2-AR)

The human f,-AR is a protein of 413 amino

acids that is encoded by an intronless gene mapped to chromosome 5q*!~? (235) and is distributed in the vascular smooth muscle cells of atria, ventricles, some arte-

rioles (e.g. coronary and skeletal muscle vessels), and systemic veins (215). Selec-

tive B,-AR agonists produce vasodilation in humans. Thus, enhanced blood pressure responses to stress in blacks might be the result of blunted B5-AR-mediated vasodilation. In the human dorsal hand vein attenuated £,-AR-mediated vasodilation was observed in Asian Indians who resided in the United States compared with white Americans (236). We and others have compared forearm blood flow re-

sponses to isoproterenol in young black and white American normotensive men and found that responses were markedly blunted in blacks (220, 237-240). Endothelial release of NO has recently been found to contribute to the vasodilator effect of B5AR stimulation (240-244).

However, the vasodilator effect of isoproterenol was

attenuated in normotensive black subjects both before and after V “_monomethyl-

L-arginine (an eNOS inhibitor) (240). Thus, the decreased vasodilator response

830

XIE ET AL

to isoproterenol in blacks is independent of the NO component of isoproterenolinduced vasorelaxation. The ethnic differences in 6,-AR-mediated vasodilation raise the possibility that 6,-AR polymorphisms may play a role. Recently, two common naturally occurring allelic variants of the B»-AR, A*°G

(or Arg!®Gly) and C’”’G (or Gln?’Glu), have been identified and their functional significance characterized (245). In contrast to the findings in vitro (246, 247), clinical studies showed that the Gly!® variant is resistant to isoproterenol-induced

desensitization (248, 248a). The Gly!® variant, however, was associated with attenuated systemic vasodilation in response to intravenous infusion of a selective 6,-AR agonist in normotensive Australian (249) and normotensive American

white subjects (250). The Glu*’ £5-AR variant is resistant to agonist-stimulated B>-AR desensitization in vitro (246-248), and in vivo is associated with greater vasodilator responses to isoproterenol (248a, 251). There are ethnic differences in the distribution of the two common $,-AR polymorphisms (252,253), with

a significantly lower frequency of the variant Glu’ allele, approximately 18%, in normotensive and hypertensive African Americans compared with Caucasian Americans (~35%). Whether the decreased frequency of the Glu’’ allele in blacks contributes significantly to their attenuated responses to B-AR agonists such as isoproterenol is currently under investigation.

OTHER FUNCTIONALLY IMPORTANT PROTEINS

Endothelial Nitric Oxide Synthase Since the identification in 1987 of NO as a biological mediator, there has been an explosion of information about the physiological, pathophysiological, pharmacological, and therapeutic roles played by this molecule. NO is synthesized from L-arginine and molecular oxygen by the constitutive enzyme endothelial nitric oxide synthase (eNOS) that is expressed in the endothelium. NO diffuses from the endothelium into vascular smooth muscle cells, where it increases the levels of cGMP by stimulating soluble guanylate cyclase (GC), resulting in vascular relaxation (254). The release of NO by the endothelium contributes to basal vascular tone (255) and regulates blood flow and blood pressure (256). Re-

cently, interethnic variations in NO-mediated responses to vasodilators such as acetylcholine, methacholine, bradykinin, and sodium nitroprusside, have been extensively assessed (239, 240, 257-259). Blacks were found to have markedly decreased NO-dependent vasodilator responses to acetylcholine (240, 258), metha-

choline (239), bradykinin (259), and sodium nitroprusside (an exogenous NO donor) (239, 240, 257), suggesting decreased cGMP-mediated vasorelaxation in blacks. In addition, blacks were found to have a reduced NO-dependent vasodilation during mental stress (257), and N°-monomethyl-L-arginine significantly inhibited the stress-induced increase in the forearm blood flow in whites but not in blacks (257). Together, these data indicate that blacks have less endotheliumdependent and endothelium-independent NO-mediated vasodilation than whites.

ETHNIC VARIATION IN DRUG RESPONSE

831

The enzyme eNOS is encoded by the NOS3 gene of 26 exons that is located on chromosome 7q*>~*° (260). A common missense mutation was identified as a G**T single base exchange at the genomic position 1917 in exon 7 (or at position

894 in its cDNA sequence), producing an amino acid substitution (Glu?’*Asp). Preliminary in vivo observations suggest that acetylcholine-mediated, endotheliumdependent vasodilator responses are attenuated in healthy white Americans ho-

mozygous for the Asp*”® variant compared with the wild-type homozygotes (261). The vasoconstrictor response to phenylephrine was also significantly higher in the variant 894T carriers (TT and GT) of French origin than in the GG carriers (262).

These data suggest that the Asp*”* (or 894T) variant of eNOS might be functionally important, resulting in decreased vasodilation.

African Americans have a lower

frequency of the variant Asp*”® allele than Caucasian Americans (14.3% versus 35.3%) (263). Thus, this polymorphism would not explain the reduced vascular response to NO in African Americans.

Pertussis Toxin—Sensitive G;-Type Protein GTP binding proteins (G proteins) comprise a superfamily of ubiquitous signaltransducing proteins that participate in many intracellular signaling cascades and mediate the functional responses to numerous agonists. G proteins are heterotrimers

with w, B, and y subunits. A frequent genetic polymorphism (C**°T) has recently been identified by Siffert et al (264) in exon 10 of the GNB3 gene (chromosome

12p!3) (265) encoding the f; subunit of pertussis toxin-sensitive G;-type protein.

This single nucleotide polymorphism is related to alternative splicing of exon 9, resulting in the loss of 41 amino acids, which results in increased sensitivity to agonists that stimulate intracellular signaling through the pertussis toxin—sensitive G protein (264). Because of the large number of receptors, including adrenoceptors that function through G protein interactions, functional polymorphisms might have important pathophysiological consequences. Case-control studies suggest that the allelic 825T variant is associated with increased blood pressure in German (264, 266, 267) and Australian Caucasians (268) and black Caribbeans of West African descent (269), but not in Japanese (270-272), aboriginal Oji-Cree Canadians (273), French individuals (274), and African Americans (275), suggesting potential ethnic differences in the nature

of genetic susceptibility loci. Furthermore, this variant was associated with left-ventricular hypertrophy in a Spanish (276) but not a German (277) population with hypertension. The 825T polymorphism was also associated with lower renin levels (266) and obesity in some studies (278—282a) but not in others

(268). a>-ARs are coupled to pertussis toxin-sensitive G; protein and mediate vasoconstriction (283, 284). a>-AR-mediated coronary blood flow was significantly decreased in subjects with the GNB3 825T allele (285), but dorsal hand vein con-

strictor responses were not different (286). Individuals of African descent have a higher frequency of the 825T allele than Caucasians (79% versus 33%) (269, 278),

832

XIE ET AL

and the 825T allele was associated with hypertension in black Caribbeans of West-African descent (269) but not in African Americans (275).

SUMMARY AND FUTURE DIRECTIONS Human P-450 enzymes associated with drug metabolism belong mainly to the CYP families CYP1, CYP2, and CYP3. The major forms (percent of total P-450

content) include CYP 3A (~30%), 2C (~20%), 1A2 (~13%), 2E1 (~7%), 2A6

(~4%), and 2D6 (~2%) (15). Approximately 40% of human P-450-mediated drug metabolism is catalyzed by polymorphic enzymes (134), and ~50% of commonly used drugs are metabolized by CYP3A4 (140). The greatest variability in the ~ levels of enzyme activity is found with CYP 2D6 and 2C enzymes because of frequently occurring functionally significant polymorphisms (136). In this review, the relationship between such polymorphisms and ethnicity has been discussed. Other phase I and phase II drug metabolizing enzymes have been extensively reviewed elsewhere (11,67, 287—290).

In addition, drug transporters (e.g. P-gp) function as drug efflux pumps in intestine, liver, and kidney and play an important role in drug absorption, distribution, and excretion. P-gp and CYP3A4 are commonly co-expressed in the same tissues and share substrate specificity. Thus, for many drugs both metabolism and transport are important determinants of disposition. Conclusive evidence for ethnic variations in the P-gp transporter activity is not available but is an intriguing possibility. Ethnic differences in B-AR-mediated responses exist. Sensitivity to propranolol is greater in Chinese, and £,-AR-mediated vasodilation is attenuated in African Americans compared with Caucasians. There are ethnic differences in the distribution of functionally significant polymorphisms, but the molecular basis for ethnic differences in vascular response is undefined. NO plays an important role in the regulation of basal vascular tone and vasodilation. Evidence suggests that African Americans have attenuated NO-mediated (both endothelium-dependent and -independent) vasodilation compared with white Americans. However, the genetic or environmental explanations for this ethnic variation remain unclear. A common genetic polymorphism of the G; protein 6;-subunit gene (GNB3)

C*T is associated with hypertension, low renin levels, and obesity in some but

not all populations of different ethnic backgrounds. In addition, the existence of an 825T allele predicted selective a,-AR-mediated coronary vasoconstriction. Although blacks have a higher allele frequency of 825T, the clinical significance is unknown. In summary, proteins that determine drug disposition and response, such as drug-metabolizing enzymes, drug transporters, and drug receptors, are the products of the genes encoding them. The study of their molecular genetics may provide a clearer understanding of ethnic differences in drug response. The principal focus so far has been on drug metabolism and drug receptors. Future directions include

ETHNIC VARIATION IN DRUG RESPONSE

833

(a) characterizing the in vivo functional significance of known polymorphisms, (b) comparing phenotypic and genotypic characterization among multiple ethnic groups, (c) identifying new polymorphisms in genes that regulate drug disposition or response, and (d) defining the relative contribution of genetic and environmental factors to ethnic variations in drug response. ACKNOWLEDGMENTS Dr. Xie was a Merck International Fellow in Clinical Pharmacology. This work was supported by grants HL 56251, HL 04012, and GM 31304 from the US Public Health Service.

Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Senior PA, Bhopal R. 1994. Ethnicity as a variable in epidemiology research. Br. Med. J. 309:327-30 . Editorial. 2000. Census, race and science.

Nature Genet. 24:97-98 . Lewis P, Rack PH, Vaddadi

KS, Allen

JJ. 1980. Ethnic differences in drug response. Postgrad. Med. J. 56(suppl.1): 4649 . Wood AJJ, Zhou HH. 1991. Ethnic differ-

ences in drug disposition and responsiveness. Clin. Pharmacokinet. 20:350-73 . Smith MW, Mendoza RP. 1996. Ethnicity and pharmacogenetics. Mt. Sinai J. Med. 63:285--90 . Wood

AJJ.

1998.

Ethnic

differences

in

drug disposition and response. Ther. Drug Monit. 20:525-26 _ Johnson JA. 2000. Predictability of the effects of race or ethnicity on pharmacokinetics of drugs. Intl. J. Clin. Pharmacol. Ther. 38:53-60 . Gaedigk A. 2000. Interethnic differences of drug-metabolizing enzymes. Intl. J. Clin. Pharmacol. Ther. 38:61-68 8a. Stein CM, Lang CC, Xie HG, Wood AJJ. 2001. Hypertension in blacks: Study of specific genotypes and phenotypes will provide a greater understanding of interindividual and interethnic variability in blood pressure regulation than studies

based on race. Pharmacogenetics In press Kalow W. 1982. Ethnic differences in drug metabolism. Clin. Pharmacokinet. 7:373400 Kalow W. 1991. Interethnic variation of drug metabolism. Trends Pharmacol. Sci. 12:102-7 . Kalow W, Bertilsson L. 1994. Interethnic

factors affecting drug response. Adv. Drug Res. 25:1-53 iV. Bertilsson L. 1995. Geographical/interracial differences in polymorphic drug oxidation: current state of knowledge of cytochromes

P450 (CYP) 2D6

and 2C19.

Clin. Pharmacokinet. 29:192—209 13) Meyer UA, Zanger UM. 1997. Molecular mechanisms of genetic polymorphisms of drug metabolism. Annu. Rev. Pharmacol. Toxicol. 37:269-96 14. Kalow W. 1984. Pharmacoanthropology: outline, problems, and the nature of case histories. Fed. Proc. 43:2134-38 15; Shimada T, Yamazaki H, Mimura M, Inui Y, Guengerich FP. 1994. Interindividual variations in human liver cytochrome P450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J. Pharmacol. Exp. Ther. 270:414-23 16. Goldstein JA, de Morais SMF. 1994.

834

XIE ETAL

Biochemistry and molecular biology of the human CYP2C subfamily. Pharmacogenetics 4:285—-99 . Nelson

DR,

Stegeman

Koymans

JJ, Feyereisen

L, Kamataki

S, Sachse

T,

R, et al. 1996.

P-450 superfamily: update on the new sequences, gene mapping, accession numbers, and nomenclature.

26 . Aynacioglu

Zi .

Pharmacogenet-

ics 6:1-42 18. Inoue K, Yamazaki H, Imiya K, Akasaka S, Guengerich

FP, et al. 1997. Relation-

ship between CYP2C9 and 2C/9 genotypes and tolbutamide methy] hydroxylation and

28.

S-mephenytoin 4’-hydroxylation activities in livers of Japanese and Caucasian populations. Pharmacogenetics 7:103—13 19. Miners JO, Birkett DJ. 1998. Cytochrome P4502C9: an enzyme of major importance in human drug metabolism. Br. J. Clin. Pharmacol. 45:525-38 20. Scott J, Poffenbarger PL. 1979. Pharmacogenetics of tolbutamide metabolism in humans. Diabetes 28:41-5S1 Ze Veronese ME, Miners JO, Rees DLP, Birkett DJ. 1993. Tolbutamide hydroxylation in humans: lack of bimodality in 106 healthy subjects. Pharmacogenetics 3:86— 93 22. Page MA, Boutagy JS, Shenfield GM. 1991. A screening test for slow metabolizers of tolbutamide. Br. J. Clin. Pharmacol. 31:649-54 23% Spielberg S, McCrea J, Cribb A, Rushmone T, Waldman S, et al. 1996. A mutation in CYP2C9 is responsible for decreased metabolism of losartan. Clin. Pharmacol. Ther. 59:215 (Abstr.) 24. Sullivan-Klose TH, Ghanayem BI, Bell DA, Zhang ZY, Kaminsky LS, et al. 1996. The role of the CYP2C9-Leu359 allelic variant in the tolbutamide polymorphism. Pharmacogenetics 6:341—49 23% Steward DJ, Haining RL, Henne KR, Davis G, Rushmore TH, et al. 1997. Genetic association between sensitivity to warfarin

and expression of CYP2C9*3. Pharmacogenetics 7:361-67

28a .

Zo:

.

oule

.

.

AS, Brockmdller C, Giizelbey

J, Bauer

P, et al. 1999.

Frequency of cytochrome P450 CYP2C9 variants in a Turkish population and functional relevance for phenytoin. Br. J. Clin. Pharmacol. 48:409-15 McCrea JB, Cribb A, Rushmore T, Osborne B, Gillen L, et al. 1999. Phenotypic and genotypic investigations of a healthy volunteer deficient in the conversion of losartan to its active metabolite E-3174. Clin. Pharmacol. Ther. 65:348—S2 Aithal GP, Day CP, Kesteven PJL, Daly AK. 1999. Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications. Lancet 353:717— 19 Ogg MS, Brennan P, Meade T, Humphries SE. 1999. CYP2C*3 allelic variants and bleeding complications. Lancet 354:1124 Aithal GP, Day CP, Kesteven PJL, Daly AK. 1999. Warfarin dose requirement and CYP2C9 polymorphisms. Lancet 353:1972-73 Meehan RR, Gosden JR, Rout D, Hastle ND, Friedberg T. 1988. Human cytochrome P450 PB-1: a multigene family involved in mephenytoin and steroid oxidations that maps to chromosome 10. Am. J. Hum. Genet. 42:26-37 de Morais SMF, Schweikl H, Blaisdell J, Goldstein JA. 1993. Gene structure and upstream regulatory regions of human CYP2C9 and CYP2C18. Biochem. Biophys. Res. Commun. 194:194-201 Rettie AE, Wienkers LC, Gonzalez FJ; Trager WF, Korzekwa KR. 1994. Impaired (S)-warfarin metabolism catalyzed by the R/44C allelic variant of CYP2C9. Pharmacogenetics 4:39-42 Haining RL, Hunter AP, Veronese ME, Trager WF, Rettie AE. 1996. Allelic variants of human cytochrome P450 2C9: baculovirus-mediated expression,

purification,

structural characterization,

substrate stereoselectivity, and prochiral

ETHNIC VARIATION IN DRUG RESPONSE

tic polymorphisms of CYP2C9 and CYP2C19 on phenytoin metabolism in Japanese adult patients with epilepsy: studies in stereoselective hydroxylation and population pharmacokinetics. Epilepsia 39:1317—23

selectivity of the wild-type and [359L mutant forms. Arch. Biochem. Biophys. 333:447-5S8

34. Miners JO, Coulter S, Birkett DJ, Goldstein JA. 2000. Torsemide metabolism by CYP2C9 variants and other human CYP2C subfamily enzymes. Pharmacogenetics 10:267-70 So! Takanashi K, Tainaka H, Kobayashi K, Yasumori T, Hosakawa M, et al. 2000. CYP2C

Ile*>? and Leu*»? variants: enzyme kinetic study with seven substrates. Pharmacogenetics 10:95—104

835

43. Odani A, Hashimoto Y, Otsuki Y, Uwai Y, Hattori H, et al. 1997. Genetic polymorphism of the CYP2C subfamily and its effect on the pharmacokinetics of phenytoin in Japanese patients with epilepsy. Clin. Pharmacol. Ther. 62:287-92 44. Takahashi H, Kashima T, Nomizo Y,

36. Yamazaki H, Inoue K, Shimada T. 1998.

Muramoto

Roles of two allelic variants (Arg144Cys and [le359Leu) of cytochrome P4502C9 in the oxidation of tolbutamide and warfarin by human liver microsomes. Xenobiotica 28:103-15 She Kaminsky LS, de Morais SMF, Faletto MB, Dunbar D, Goldstein JA. 1993. Correlation of human cytochrome P4502C substrate specificities with primary structure: warfarin as a probe. Mol. Pharmacol. 43:234— Bg 38. Furuya H, Fernandez-Salguero P, Gregory

Metabolism of warfarin enantiomers in Japanese patients with heart disease having different CYP2C9 and CYP2C19 genotypes. Clin. Pharmacol. Ther. 63:519—28

W, Taber H, Steward A, et al. 1995. Genetic

polymorphism of CYP2C9 and its effect on warfarin maintenance dose requirement in patients undergoing anticoagulation therapy. Pharmacogenetics 5:389—-92 39. Bhasker CR, Miners JO, Coulter S, Birkett DJ. 1997. Allelic and functional variability of cytochrome P4502C9. Pharmacogenetics 7:51-58 40. Kidd RS, Straughn AB, Meyer MC, Blaisdell J, Goldstein JA, et al. 1999. Pharmacokinetics of chlorpheniramine, phenytoin, glipizide and nifedipine in an individual homozygous for the CYP2C9*3 allele. Pharmacogenetics 9:71—80 41. Miners J, Wing LMH, Birkett DJ. 1985.

Normal metabolism of debrisoquine and theophylline in a slow tolbutamide metabolizer. Aust. N.Z. J. Med, 15:348-49

42. Mamiya K, Ieiri 1, Shimamoto J, Yukawa E, Imai J, et al. 1998. The effects of gene-

45. Takahashi

N, Shimizu

H, Kashima

T, et al. 1998.

T, Nomoto

S,

Iwade K, Tainaka H, et al. 1998. Com-

parisons

between

in-vitro

and

in-vivo

metabolism of ($)-warfarin: catalytic ac-

tivities of CDNA-expressed CYP2C9, its Leu359 variant and their mixture versus unbound clearance in patients with the corresponding CYP2C9 genotypes. Pharmacogenetics 8:365—73 45a. Thijssen HHW, Verkooijen IWC, Frank HLL. 2000. The possession of the CYP2C9*3 allele is associated with low dose requirement of acenocoumarol. Pharmacogenetics 10:757—60 46. Arnold K, Gerber N. 1970. The rate of decline of diphenylhydantoin in human plasma. Clin. Pharmacol. Ther. 11:121— 34 47. Edeki TI, Brase DA. 1995. Phenytoin disposition and toxicity: role of pharmacokinetic and interethnic factors. Drug Metab. Rey. 27:449-69 47a.

Yoon YR, Shon JH, Kim MK, Lim YC,

Lee HR, et al. 2001. Frequency of cytochrome P450 2C9 mutant alleles in a Korean population. Br. J. Clin. Pharmacol. In press

48. Wang S-L, Huang J-D, Lai M-D, Tsai J-J.

836

XIE ET AL

1995.

Detection

of CYP2C9

polymor-

phism based on the polymerase chain reaction in Chinese. Pharmacogenetics 5:37—

49.

Nasu K, Kubota T, Ishizaki T. 1997. Ge-

netic analysis of CYP2C9 polymorphism in a Japanese population. Pharmacogenetics 7:405—9 50. Kimura M, Ieiri I, Mamiya K, Urae A, Higuchi S. 1998. Genetic polymorphism of cytochrome P450s, CYP2C19, and CYP2C9 in a Japanese population. Ther. Drug Monit. 20:243-47 ytIE London SJ, Daly AK, Leathart JB, Navidi WC, Idle JR. 1996. Lung cancer risk in relation to the CYP2C9* 1/CYP2C9*2 genetic polymorphism among AfricanAmericans and Caucasians in Los Angeles county, California. Pharmacogenetics 6:527-33 2: Stubbins MJ, Harries LW, Smith G, Tarbit MH, Wolf CR. 1996. Genetic analysis of the human cytochrome P450 CYP2C9 locus. Pharmacogenetics 6:429-39 53, Brochmoller J, Rost KL, Gross D, Schenkel A, Roots I. 1995. Phenotyping of CYP2C19 with enantiospecific HPLCquantification of R- and S-mephenytoin and comparison with the intron 4/exon 5 G— A-splice site mutation. Pharmacogenetics 5:80-88

54. Ackermann E, Cascorbi I, Sachse C, BrockmOller J, Mrozikiewicz PM, et al. 1997. Frequencies and the allelic linkage of CYP2C9 mutations in a German population, and the detection of a C/T mutation inintron 2. Eur. J. Clin. Pharmacol. 52:A71 (Abstr. )

3)

ile

42

Yasar U, Eliasson E, Dahl ML, Johansson I, Ingelman-Sundberg M, et al. 1999. Validation of methods for CYP2C9 genotyping: frequencies of mutant alleles in a Swedish population. Biochem. Biophys. Res. Commun, 254:628-31 56. Blann A, Hewitt J, Siddiqui F, Bareford D. 1999, Racial background is a determinant of average warfarin dose required to main-

58.

59:

60.

6l.

.

tain the INR between 2.0 and 3.0. Br. J. Haematol. 107:207-9 Yu HCM, Chan TYK, Critchley JAJH, Woo KS. 1996. Factors determining the maintenance dose of warfarin in Chinese patients. QO. J. Med. 89:127-35 Schwarz UI, Choo EF, Dresser GK, Stein CM, Wood AJJ, et al. 2000. Identification of a new CYP2C9 variant in African-Americans. Clin. Pharmacol. Ther. 67:169 (Abstr.) Imai J, Ieiri I, Mamiya K, Miyahara S, Furuumi H, et al. 2000. Polymorphism of the cytochrome P450 (CYP) 2C9 gene in Japanese epileptic patients: genetic analysis of the CYP2C9 locus. Pharmacogenetics 10:85—89 Gotoh O. 1992. Substrate recognition sites in cytochrome P-450 family 2 (CYP2) proteins inferred from comparative analysis of amino acid and coding nucleotide sequences. J. Biol. Chem. 267:83-90 Chang M, Tybring G, Dahl M-L, Gétharson E, Sagar M, et al. 1995. Interphenotype differences in disposition and effect on gastrin levels of omeprazole - suitability of omeprazole as a probe for CYP2C19. Br J. Clin. Pharmacol. 39:511-18 Caraco Y, Lagerstrom PO, Wood AJJ. 1996. Ethnic and genetic determinants of omeprazole disposition and effect. Clin. Pharmacol. Ther. 60:157-67

. Furuta T, Ohashi K, Kamata T, Takashima

M, Kosuge K, Kawasaki T, et al. 1998. Effect of genetic differences in omeprazole metabolism on cure rates for Helicobacter pylori infection and peptic ulcer. Ann. Intern. Med. 129:1027-30 . Ohashi K, Furuta T, Kosuge K, Kimura M, Nishimoto M, et al. 1998. Disposition and effects of omeprazole and related to CYP2C19 genotype. Clin. Pharmacol. Ther. 63:152 (Abstr.) 64. Furuta T, Ohashi K, Kosuge K, Zhao X-J, Takashima M, et al. 1999. CYP2C19

837

ETHNIC VARIATION IN DRUG RESPONSE

genotype status and effect of omeprazole on intragastric pH in humans. Clin. Pharmacol. Ther. 65:552-61

65. Furuta T, Takashima M, Shirai N, Xiao F, Hanai

H, et al. 2000. Cure of refrac-

tory duodenal ulcer and infection caused by Helicobacter pylori by high doses of omeprazole and amoxicillin in a homozygous CYP2C19 extensive metabolizer patient. Clin. Pharmacol. Ther. 67:684— 89

66. Tanigawara Y, Aoyama N, Kita T, Shirakawa K, Komada F, Kasuga M, et al., 1999. CYP2C19 genotype-related efficacy of omeprazole for the treatment of infection caused by Helicobacter pylori. Clin. Pharmacol. Ther. 66:528-34 67. Daly AK. 1995. Molecular basis of polymorphic drug metabolism. J. Mol. Med. 73:539-53 68. Romkes M, Faletto MB, Blaisdell JA, Raucy JL, Goldstein JA. 1991. Cloning and expression of complementary DNA for multiple members of the human cytochrome P450 TiC subfamily. Biochemistry 30:3247-55 69. Zaphiropoulos PG. 1999. RNA molecules containing exons originating from different members of the cytochrome P450 2C gene subfamily (CYP2C) in human epidermis and liver. Nucleic Acids Res. 27:2585— 90 . de Morais SMF, Wilkinson GR, Blaisdell J, Nakamura K, Meyer UA, et al.

1994. The major genetic defect responsible for the polymorphism of S-mephenytoin metabolism in humans. J. Biol. Chem. 269:15419-22

ile de Morais SMF, Wilkinson GR, Blaisdell J, Meyer UA, Nakamura

K, et al. 1994.

Identification of a new genetic defect responsible for the polymorphism of (S)mephenytoin metabolism in Japanese. Mol. Pharmacol. 46:594—98

Wh de Morais SMF, Goldstein JA, Xie HG, Huang SL, Lu YQ, et al. 1995. Genetic analysis of the S-mephenytoin in a Chi-

nese population. Clin. Pharmacol. 58:404-11 . Xie

HG,

Stein

CM,

son GR, Flockhart

lelic, genotypic,

Kim

RB,

Ther.

Wilkin-

DA, et al. 1999. Al-

and phenotypic

distri-

bution of S-mephenytoin 4’-hydroxylase (CYP2C19) in healthy Caucasian populations of European descent throughout the world. Pharmacogenetics 9:539-49

74. Xie HG, son

GR,

Kim

RB,

Wood

Stein AJJ.

CM, 1999.

WilkinGenetic

polymorphism of (S)-mephenytoin 4’hydroxylation in populations of African descent. Br. J. Clin. Pharmacol. 48:402-8 TSS Xie HG. 2000. Genetic variations of Smephenytoin 4’-hydroxylase (CYP2C19) in the Chinese population. Life Sci. 66:PL175-81 76. Alvan G, Bechtel P, Iselius L, GundertRemy U. 1990. Hydroxylation polymorphisms of debrisoquine and mephenytoin in European populations. Eur J. Clin. Pharmacol. 39:533-37 ae Wilkinson GR, Guengerich FP, Branch RA. 1989. Genetic polymorphism of Smephenytoin hydroxylation. Pharmacol. Rev. 43:53-76 78. Zhang Y, Reviriego J, Lou YQ, Sjéqvist F, Bertilsson L. 1990. Diazepam metabolism in native Chinese poor and extensive hydroxylators of S-mephenytoin: interethnic differences in comparison with white subjects. Clin. Pharmacol. Ther. 48:496-S02 WS). Bertilsson L, Henthorn TK, Sanz E, Tybring G, Siwe J, et al. 1989. Importance of genetic factors in the regulation of diazepam metabolism: relationship to Smephenytoin, but not debrisoquin, hydroxylation phenotype. Clin. Pharmacol. Ther. 45:348-55 80. Bertilsson L, Kalow W. 1993. Why are diazepam metabolism and polymorphic S-mephenytoin hydroxylation associated with each other in white and Korean populations but not in Chinese population? Clin. Pharmacol. Ther. 53:608—10 81. Xie HG. 1997. Direct evidence for the

838

XIE ET AL

higher frequency of CYP2C/9 allelic heterozygotes in Chinese subjects than in white subjects. Clin. Pharmacol. Ther. 62:691—92

82. Kumana CR, Lauder IJ, Chan M, Ko W, Lin HJ. 1987. Differences in diazepam pharmacokinetics in Chinese and white Caucasians-relation to body lipid stores. Eur. J. Clin. Pharmacol. 32:211—15 83. Andersson T, Regardh CG, Lou YC, Zhang Y, Dahl ML, et al. 1992. Polymorphic hydroxylation of S-mephenytoin and omeprazole metabolism in Caucasian and Chinese subjects. Pharmacogenetics 2:25—31 84. Caraco Y, Lagestrom PO, Wood AJJ. 1995. Omeprazole disposition in Caucasian and Chinese subjects. Clin. Pharmacol. Ther. 57:216 (Abstr.) 85. Caraco Y, Tateishi T, Wood AJJ. 1995. Interethnic difference in omeprazole’s inhibition of diazepam metabolism. Clin. Pharmacol. Ther. 58:62—72 86. Ishizaki T, Sohn DR, Kobayashi K, Chiba K, Lee KH, et al. 1994. Interethnic dif-

87.

88.

89,

90.

ferences in omeprazole metabolism in the two S-mephenytoin hydroxylation phenotype studied in Caucasians and Orientals. Ther. Drug Monit. 16:214-15 Funck-Brentano C, Becquemont L, Leneveu A, Roux A, Jaillon P, et al. 1997. Inhibition by omeprazole of proguanil metabolism: mechanism of the interaction in vitro and prediction of in vivo results from the in vivo experiments. J. Pharmacol. Exp. Ther. 280:730-38 Kaneko A, Bergqvist Y, Takechi M, Kalkoa M, Kaneko O, et al. 1999. Intrinsic efficacy of proguanil against falciparum and vivax malaria independent of the metabolite cycloguanil. J. Infect. Dis. 179:974-79 Idle JR, Corchero J, Gonzalez FJ. 2000. Medical implications of HGP’s sequence of chromosome 22. Lancet 355:319 Xie HG, Xu ZH, Luo X, Huang SL, Zeng FD, et al. 1996. Genetic polymorphisms

of debrisoquine and S-mephenytoin oxidation metabolism in Chinese populations: a

meta- analysis. Pharmacogenetics 6:235— 38 OF Lennard MS, Iyun AO, Jackson PR, Tucker GT, Woods HF. 1992. Evidence for a disassociation in the control of sparteine, debrisoquine and metoprolol metabolism in Nigerians. Pharmacogenetics 2:89-92 pe Sachse C,. Brockmdller J, Bauer S, Roots I. 1997. Cytochrome P450 2D6 variants in a Caucasian population: allele frequencies and phenotypic consequences. Am J. Hum. Genet. 60:284—95 os Gaedigk A, Gotschall RR, Forbes NS, Simon SD, Kearns GL, et al. 1999. Optimization of cytochrome P4502D6 (CYP2D6) phenotype assignment using a genotyping algorithm based on allele frequency data. Pharmacogenetics 9:669—82 94. Wennerholm A, Johansson I, Massele AY, Jande M, Alm C, et al. 1999. Decreased capacity for debrisoquine metabolism among black Tanzanians: analyses of the CYP2D6 genotype and phenotype. Pharmacogenetics 9:707-14 . Fukuda T, Yamamoto

Q,

Ohno

M,

et

al.

I, Nishida Y, Zhou

1999. Effect of the CYP2D6*/0 genotype on venlafaxine pharmacokinetics in healthy adult volunteers. Clin. Pharmacol. Ther. 47:450—53 96. Johansson I, Oscarson M, Yue Q-Y, Bertilsson L, Sjéqvist F, et al. 1994. Genetic analysis of the Chinese cytochrome P4502D locus: characterization of variant CYP2D6 genes present in subjects with diminished capacity for debrisoquine hydroxylation. Mol. Pharmacol. 46:452-59 Oi Hou ZY, Chen CP, Yang WC, Lai MD, Buchert ET, et al. 1996. Determination of dextromethorphan metabolic phenotype by salivary analysis with a reference to genotype in Chinese patients receiving renal hemodialysis. Clin. Pharmacol. Ther. 59:411-17 98. Lai ML, Wang SL, Lai MD, Lin ET, Tse M, et al. 1995. Propranolol disposition in Chinese subjects of different CYP2D6 genotypes. Clin. Pharmacol. Ther. 58:264-68

ETHNIC VARIATION IN DRUG RESPONSE

netics of codeine and its metabolites in healthy Chinese and Caucasian extensive hydroxylators of debrisoquine. Br. J. Clin. Pharmacol. 31:643-47

99 . Oscarson M, Hidestrand M, Johansson I, Ingelman-Sundberg M. 1997. A combination of mutations in the CYP2D6*17 (CYP2D6Z) allele causes alterations in enzyme function. Mol. Pharmacol. 52:1034—40

100. Tateishi T, Chida M, Ariyoshi N, Mizorogi Y, Kamataki T, et al. 1999. Analysis of the CYP2D6 gene in relation to dextromethorphan O-demethylation capacity in a Japanese population. Clin. Pharmacol. Ther. 65:570-75 101. Tseng CY, Wang SL, Lai MD, Lai ML, Huang JD. 1996. Formation of morphine from codeine in Chinese subjects of different CYP2D6 genotypes. Clin. Pharmacol. Ther. 60:177-82 102. Yokota H, Tamura S, Furuya H, Kimura

839

108. Caraco Y, Sheller J, Wood AJJ. 1999. Impact of ethnic origin and quinidine coadministration on codeine’s disposition and pharmacodynamic effects. J. Pharmacol. Exp. Ther. 290:413ape 109. Horai Y, Nakano M, Ishizaki T, Ishikawa K, Zhou

HH, et al. 1989. Metoprolol

and mephenytoin oxidation polymorphism in Far Eastern Oriental subjects: Japanese versus mainland Chinese. Clin. Pharmacol. Ther. 46: 198-207 110. Sohn DR, Shin SG, Park CW, Kusaka M,

Evide-

Chiba K, et al. 1991. Metoprolol oxida-

nce for a new variant CYP2D6 allele CYP2D6J in a Japanese population associated with lower in vivo rates of sparteine metabolism. Pharmacogenetics 3:25663 103. Bertilsson L, Lou Y-Q, Du Y-L, Liu

tion polymorphism in a Korean population: comparison with native Japanese and Chinese populations. Br J. Clin. Pharmacol. 32:504-7

S, Watanabe

M,

et al.

1993.

ul.

105. Rudorfer H, Zhang

EA,

MV,

Lane

MD,

Potter WZ.

Chang

W-

1984.

De-

sipramine pharmacokinetics in Chinese and Caucasian volunteers. Br J. Clin. Pharmacol. 17:433-40 106. Yue QY, Svensson JO, Alm C, Sjéqvist F, Sawe J. 1989. Interindividual and in-

terethnic differences in the demethylation and glucuronidation of codeine. Br. J. Clin. Pharmacol. 28:629-37 107. Yue QY, Svensson JO, Sjéqvist F, Sawe J. 1991. A comparison of the pharmacoki-

Dahl

ML,

Johansson

I,

1996. Debrisoquine and S-mephenytoin hydroxylation phenotypes and genotypes in a Korean population. Pharmacogenetics 6:441-47

PZ,

Yue QY, Svensson JO, Sawe J, Bertils-

son L. 1995. Codeine metabolism in three Oriental populations: a pilot study in Chinese, Japanese and Koreans. Phar-

macogenetics 5:173-77 112a.

22:451-75

HK,

Ingelman-Sundberg M, Cha YN, et al.

Y, Kuang T-Y, et al. 1992. Pronounced

differences between native Chinese and Swedish populations in the polymorphic hydroxylations of debrisoquin and S-mephenytoin. Clin. Pharmacol. Ther. 51:388-97 104. Lou YC. 1990. Differences in drug metabolism polymorphism between Orientals and Caucasians. Drug Metab. Rev.

Roh

Nishida

Y, Fukuda

T, Yamamoto

I,

Azuma J. 2000. CYP2D6 genotypes in a Japanese population: low frequencies of CYP2D6 gene duplication but high frequency of CYP2D6* 10. Pharmacogenetics 10:567—70 ily. Wang SL, Huang JD, Lai MD, Liu BH, Lai ML. 1993. Molecular basis of genetic variation in debrisoquine hydroxylation in Chinese subjects: polymorphism in RFLP and DNA sequence of CYP2D6. Clin. Pharmacol. Ther. 53:410-18

840

XIE EVAL

114. Garcia-Barcel6 M, Chow LY, Kum Chiu HF, Wing YK, Shing Lee DT, et al. 2000. Genetic analysis of the CYP2D6 locus in a Hong Kong Chinese population. Clin. Chem. 46:18—23 Ile: Kubota T, Yamaura Y, Ohkawa N, Hara H, Chiba K. 2000. Frequencies of CYP2D6 mutant alleles in a normal Japanese population and metabolic activity of dextromethorphan O-demethylation in different CYP2D6 genotypes. Br J. Clin. Pharmacol. 50:31-34 116. Forbes NS, Bradford LD, Gotschall RR, Leeder JS, Gaedigk

122.

123.

A. 1999. CYP2D6

allele frequencies in African Americans: phenotype concordance with dextromethorphan. Clin. Pharmacol. Ther.

124.

65:170 (Abstr.)

Lele Leathart JBS, London SJ, Steward A, Adams JD, Idle JR. et al. 1998. CYP2D6 phenotype-genotype relationships in African-Americans and Caucasians in Los Angeles. Pharmacogenetics 8:529-41 118. London SJ, Daly AK, Leathart JBS, Na-

p25:

vidi WC, Carpenter CC, et al. 1997. Ge-

netic polymorphism of CYP2D6 and lung cancer risk in African-Americans and Caucasians in Los Angeles county. Carcinogenesis 18:1203-14

.

IGS). Martin DE, Tran JQ, Flockhart DA, Jorkasky DK. 1998. Analysis of CYP2D6 and CYP2C19 genotypes in large African-American and Caucasian populations. Clin. Pharmacol. Ther. 63:206 (Abstr. )

120. Evans WE, Relling MV, Rahman A, McLeod HL, Scott EP, et al. 1993. Genetic basis for a lower prevalence of deficient CY P2D6 oxidative drug metabolism phenotypes in black Americans. J. Clin. Invest. 91:2150-54 IP Aklillu E, Persson I, Bertilsson L, Johansson I, Rodrigues F, et al. 1996. Frequent distribution of ultrarapid metabolizers of debrisoquine in an Ethiopian population carrying duplicated and multi-

127;

128.

129.

plicated functional CYP2D6 alleles. J. Pharmacol. Exp. Ther. 278:441-46 Bathum L, Skjelbo E, Mutabingwa TK, Madsen H, Hgrder M, et al. 1999. Phenotypes and genotypes for CYP2D6 and CYP2C19 in a black Tanzanian population. Br. J. Clin. Pharmacol. 48:395-401 Masimirembwa C, Hasler J, Bertilsson L, Johansson I, Ekberg O, et al. 1996. Phenotype and genotype analysis of debrisoquine hydroxylase (CYP2D6) in a black Zimbabwean population. Reduced enzyme activity and evaluation of metabolic correlation of CYP2D6 probe drugs. Eur. J. Clin. Pharmacol. 51:117-—22 Masimirembwa C, Persson I, Bertilsson L, Hasler J, Ingelman-Sundberg M. 1996. A novel mutant variant of the CYP2D6 gene (CYP2D6*17) common in a black African population: association with diminished debrisoquine hydroxylase activity. Br. J. Clin. Pharmacol. 42:713-19 Chen S, Chou W-H, Blouin RA, Mao Z, Humphries LL, et al. 1997. The cytochrome P450 2D6 (CYP2D6) enzyme polymorphism: screening costs and influence on clinical outcomes in psychiatry. Clin. Pharmacol. Ther. 60:522-34 Marez D, Legrand M, Sabbagh N, Guidice JM, Spire C, et al. 1997. Polymorphism of the cytochrome P450 CYP2D6 gene in European population: characterization of 48 mutations and 53 alleles, their frequencies and evolution. Pharmacogenetics 7:193-202 Madsen H, Nielsen KK, Brgsen K. 1995, Imipramine metabolism in relation to the sparteine and mephenytoin oxidation polymorphisms-a population study. Br. J. Clin. Pharmacol. 39:433-39 Bathum L, Andersen-Ranberg K, Boldsen J, Brgsen K, Jeune B. 1998. Genotypes for the cytochrome P450 enzymes CYP2D6 and CYP2C19 in human longevity: role of CYP2D6 and CYP2C19 in longevity. Eur. J. Clin. Pharmacol. 54:427-—30 Marandi T, Dahl ML, Kiivet RA, Rago L,

ETHNIC VARIATION IN DRUG RESPONSE

Sjoqvist F.1996. Debrisoquine and Smephenytoin hydroxylation phenotypes and CYP2D6 genotypes in an Estonian population. Pharmacol. Toxicol. 78:303-7 130. Saarikoski ST, Sata F, HusgafvelPursiainen K, Rautalahti M, Haukka J, et al. 2000. CYP2D6 ultrarapid metabolizer genotype as a potential modifier of smoking behavior. Pharmacogenetics 10:5-10 131 Griese EU, Zanger UM, Brudermanns U, Gaedigk A, Mikus

G, et al. 1998.

Assessment of the predictive power of genotypes for the in vivo catalytic function of CYP2D6 in a Caucasian population. Pharmacogenetics 8:15—26

S27

Marandi

T, Dahl

M-L,

Rago

L, Ki-

ivet R, Sjéqvist F. 1997. Debrisoquine and S-mephenytoin hydroxylation polymorphisms in a Russian population living in Estonia. Eur. J. Clin. Pharmacol. 53:257-60

iS3: Yamada H, Dahl M-L, Lannfelt L, Viitanen M, Winblad B, et al. 1998.

CYP2D6 and CYP2C19 genotypes in an elderly Swedish population. Eur. J. Clin. Pharmacol. 54:479-81 134. Aynacioglu AS, Sachse C, Bozkurt A, 1r,4m°

no

-

141

RP 67,580

t,m> gp

151

WIN 51708

r>h

152

FK 888

h>r

FK 224

h~r

CGP 49823

fut

SR 140333

r~h~gp

GR 203040

li, (eabe5 g

156

h,g, gp>r

Si

h, gp>r

158

LY 303870

h, gp >r,m

159

L-733,060 BR

h, g, gp>r SS SS

SZT NKT 343

CRAOZ

aU

*h.= human

bgp = guinea pig eg = gerbil ¢r=rat °m = mouse

‘blank= not tested; ; ; ;

80, 153

yes

153 143, 154

yes

155)

see text

see text

p

® NEMEROFF

\

fe) ry

698-N ne

x

H

a ) Te

Y

H N >=0

N-NH

4

4 =

|

Nv LN H

S

090'€eZ-7

ok

Al OLBE0E

!ururyosnou ‘systuosejur

dOO

RorMayD sounjonys jo

€28'6b Ia

ans 7

4

=

/

4

Z0S LAN re

4

feeOrl

® OWENS

HS

STOUT

5 het

890

NK1I ANTAGONISTS

891

mild stress-induced anhedonia, in which rats subjected to various stressors develop an increased threshold for rewarding stimuli such as intra-accumbens electrical stimulation. Unlike tests of anxiety, behavior modification in these models is generally unidirectional; i.e. “depressogenic’” compounds are not commonly identified in this manner. Aside from equivocal results in the swim test (140), published

studies are lacking regarding activity of tachykinin receptor antagonists in these models. As noted previously, maternal separation of neonatal animals has been used as a model of anxiety or depression (112, 113). Relatively brief manipulation of these animals produces long-term endocrine and behavioral sequellae, which are responsive to antidepressant administration. Although the effects of NK, receptor antagonists on the long-term consequences of maternal deprivation have not been reported, Rupniak and colleagues did report that systemic pretreatment with L 733,060 or other NK, receptor antagonists reduced the vocalizations of guinea pig pups (2) or neonatal mice (115) subjected to maternal separation. Most importantly, antidepressant and anxiolytic agents reduced guinea pig vocalizations as well. Clearly, more preclinical studies are required in order to evaluate the antidepressant activity of NK, receptor antagonists. Evidence in laboratory animals is equivocal regarding the anxiolytic potential of NK, receptor antagonists. In the mouse black-and-white box test (similar in principle to the elevated plus-maze), systemic administration of CP 96,345 pro-

duced overall motor impairment but no anxiogenic or anxiolytic effect per se (141). Newer NK, receptor antagonists have not been reported to produce motor impairment; in general the more recent compounds have higher selectivity for the NK, receptor with lower affinity for other sites such as sodium and calcium channels (142). ICV injection of the NK, receptor antagonist, FK 888, produced anxiolytic effects in the mouse elevated plus-maze test (statistical significance depended largely on analysis of percent time in open arms of the maze rather than on analysis of percent of entries) without altering overall locomotor activity or motor coordination (80). The antagonist compound CGP 49823 reportedly produced anxiolysis in rats placed in an unfamiliar environment (143) and in the rat social interaction test (140), but according to the latter abstract was ineffective in the elevated plus-maze test (140).

The behavioral and physiologic effects of tachykinin receptor antagonists have been further described in other experiments not directly related to antidepressant or anxiolytic potential. [CV SP administration increased heart rate and blood pressure, effects that were reversed by CP 96,345 (144); neurokinin A and NK, receptor antagonists produced effects similar to SP and CP 96,345, respectively. Subse-

quently, many brain regions have been identified that elicit pressor responses to SP microinjections, including the locus coeruleus, periaqueductal grey, parabrachial nucleus, rostral ventrolateral medulla, central nucleus of the amygdala, paraven-

tricular nucleus, and other hypothalamic nuclei (27). In a related study, ICV NK,, but not NK, receptor blockade, reduced the cardiovascular response (increase in heart rate and mean arterial pressure) of rats exposed to acute stress induced by

892

STOUT

® OWENS

® NEMEROFF

s.c. injection of formalin (145). The doses required to produce this effect were reportedly less than those required to produce antinociception following intrathecal administration. However, the interpretation of the NK, receptor antagonist effect as a disruption of a general, physiologic stress response would have been clearer if a nonpainful stressor had been chosen in this study. Microinjection of NK, receptor antagonists, CP 96,345 or CP 99,994 into the caudal pontine reticular nucleus, a relay nucleus involved in processing acoustic stimuli, blocked footshock-induced sensitization of the acoustic startle response (146). One source of SP neurons projecting to this region is the laterodorsal tegmental nucleus; there are no known direct SP-ergic projections from the central nucleus of the amygdala (147), aregion necessary for acoustic startle response sensitization (148). Systemic CP 96,345 administration, or microinjection directly into the medial hypothalamus of cats results in blockade of the facilitation of hypothalamicstimulated defensive rage behavior by medial amygdaloid electrical stimulation (149). These results, along with the demonstration of a direct medial amygdaloidmedial hypothalamus pathway, suggest that when activated, neurons in the medial amygdaloid nucleus (a region of peak PPT-A expression within the brain) release SP from nerve terminals in the hypothalamus, resulting in potentiation of certain fear- or stress-induced behaviors. At least two studies have indicated a possible anxiolytic profile of NK, receptor antagonists, despite the reportedly sparse distribution of this receptor in the CNS. Acute peripheral injection of the NK, receptor antagonists SR 48968 and GR 159897 produced an increase in the amount of time mice or marmosets spent in exposed or potentially threatening locations, in light-dark box and human confrontation tests, respectively (150). ICV administration of SR 48968 was also

anxiolytic in the mouse elevated plus-maze test (80). The antidepressant activity of NK, receptor antagonists in animal models has not been studied.

CLINICAL STUDIES

Cerebrospinal Fluid (CSF) and Postmortem Measurements of Substance P In the absence of a direct method for obtaining neurochemical measurements of CNS activity in humans, neurotransmitters and metabolites are frequently measured in cerebrospinal fluid (CSF) samples obtained by lumbar puncture in psychiatric patients and controls. Although there are certainly limitations to this method, such as the lack of anatomic resolution and the preferential sampling of neurotransmitters released from brain regions in proximity to the ventricular system, the blood-brain barrier ensures that a change in neurotransmitter concentration in a particular patient population can at least be assumed to represent some sort of change in CNS activity. CSF findings in studies of depressed patients include

NK1I ANTAGONISTS

893

increased CRF (160), decreased somatostatin (161), and decreases in the sero-

tonin metabolite 5-HIAA, particularly within impulsive and/or suicidal subgroups of patients (162). The measurement of CSF SP concentrations in depressed patients is an understudied area. Rim6n et al (163) reported that depressed patients exhibited a fourfold mean elevation in SP-like immunoreactivity. Electrophoresis patterns also indicated an increase in SP degradation in the depressed group. A second study from the same investigators was not sufficiently powered to examine changes in depression, but it was noted that the three patients with depression had a substantially higher mean CSF SP concentration than did controls or patients with personality disorders, though the differences were not statistically significant (164). In neither study was SP concentration significantly altered in schizophrenic patients relative to controls. The findings of increased CSF SP in depression were not replicated by a second group (165), though the patient populations in the two studies differed widely. Furthermore, lower rather than higher CSF SP concentrations correlated with psychic anxiety and self-reported symptoms of sadness and inner tension in a group of nondepressed patients, most of whom suffered from chronic pain syndromes (166). Obviously, the relevance of the latter study to patients with mood disorders is unclear. In a study that unfortunately did not include healthy controls, successful treatment of depression with fluoxetine for six weeks did not significantly alter CSF concentration of (N-terminally extended) SP in these patients (167). Treatment effects on CSF SP concentrations have not been examined in any subsequent, published studies. To our knowledge, there have in fact been no studies of CSF SP in depression in the past decade. Without further data, it is impossible to draw conclusions from the extant, contradictory

literature. A second method commonly employed to examine the involvement of a neurotransmitter system in human disorders is postmortem examination of human tissue. In most of these studies, patterns of gene expression or membrane receptor density rather than neuropeptide concentration have been studied owing to a variety of technical reasons. In a study of NK, receptor autoradiography in cingulate cortex of controls, and patients with schizophrenia, unipolar and bipolar depression collected by the Stanley Foundation, total receptor binding was unchanged, but a decrease in superficial/deep cortical layer NK, receptor density ratio was noted in the unipolar depression group (168). Clearly, further work is required to characterize more fully the regional pattern of SP and NK, receptor expression in depression.

Clinical Trials The preclinical studies and CSF and postmortem data reviewed establish a clear link between SP neurotransmission and depression. antidepressant efficacy of the NK, receptor antagonist, M K-869 (2) a surprise to the field. MK-869 was selected from among various

above do not The reported was therefore NK, receptor

894

STOUT

® OWENS

® NEMEROFF

antagonists under development based on its affinity for human receptors, oral bioavailability, selectivity, brain penetrance, and favorable pharmacokinetics. Regarding selectivity, it is important to note that MK-869 has little or no affinity for serotonin, norepinephrine and dopamine transporters, or for monoamine oxidase A and B. Therefore, any efficacy of the drug may be assumed to result from its action on NK, receptors. In a double-blind study of 213 patients randomly assigned to once-daily placebo, the SSRI paroxetine (20 mg) or MK-869 (300 mg), clinical improvement as mea-

sured with the 21-item Hamilton rating scale for depression (HAM-D) was virtually identical over a 6-week period in the paroxetine and MK-869 groups (Figure 3). The therapeutic delay period was also similar in the two groups. In addition, the NK, receptor antagonist reduced symptoms of anxiety, as measured by the Hamilton anxiety scale (HAM-A) (Figure 3). Thus far, results of NK, receptor antagonists in patients with anxiety disorders have not been reported. Of note, MK-869 did not produce significantly higher rates of any side effect reported, when compared with placebo or paroxetine, and was associated with a substantially lower incidence of sexual dysfunction when compared to paroxetine (3% vs. 26%). This side effect is one of the principal liabilities of paroxetine and the other SSRIs; in fact, rates much higher than 26% have been reported (169, 170)

when patients are specifically asked about this side effect. In summary, the report indicated that NK, receptor antagonists may possibly offer similar efficacy in certain patient populations to established antidepressant agents with a more favorable side effect profile. Such a dramatic finding begs for confirmation and several clinical trials are currently underway to test the hypothesis that NK, receptor antagonists are effective antidepressants and/or anxiolytics. In a second trial, MK-869 apparently did not demonstrate higher efficacy than placebo (171). A more potent NK, receptor antagonist has subsequently been brought into clinical trials; no data have been published regarding this second Merck compound at the time of this writing.

CONCLUSIONS With the exception of the report from the Merck group detailing preclinical development of NK, receptor antagonists and the antidepressant efficacy of the compound, MK-869, a systematic literature review failed to provide strong evidence for the antidepressant efficacy of this class of compounds. In most cases, insufficient and/or contradictory evidence exists to establish particular arguments for or against the involvement of SP in depression. Alterations in SP synthesis and secretion in animal models of depression, and the effects of NK, receptor antagonists in these models are essentially untested. Measurement of SP in human tissue is a neglected field: To date it is equivocal whether CSF SP is altered in depression, and studies of PPT-A gene expression in postmortem brain tissue of depressed patients are lacking.

>

21 HAM-D in Change Mean

21 HAM-D in Change Mean

Week on mean Figure 3 Effect of treatment with MK-869 (300 mg/day) or paroxetine (20 mg/day) Anxiety Hamilton the and (A) (HAM-D21) Scale Depression Hamilton the on baseline from change (dark MK-869 of are s Scale (14 items), (B) in patients with major depressive disorder. Comparison 64). = squares,n (open placebo versus 68) = n circles, n = 66) or paroxetine (grey triangles, permission with s, modification minor with Reprinted intervals. Error bars show 95% confidence (2). Copyright 1998, American Association for the Advancement of Science.

896

STOUT

® OWENS

® NEMEROFF

The rich anatomic distribution of SP and NK, receptor-expressing neurons in limbic and monoaminergic regions of the brain is certainly compatible with the involvement of SP-ergic neurotransmission in depression and/or antidepressant efficacy. Studies to date have identified several regions and pathways of particular interest, including the following: (a) a medial amygdala projection to the hypothalamus involved in potentiation of defensive rage behavior, (b) the locus coeruleus,

in which NK, receptor activation may stimulate noradrenergic neuronal firing in response to stress, (c) the periaqueductal grey, which is heavily innervated by SP neurons and which has long been shown to be involved in behavioral and physiologic fear responses. These and other regions in the brain serve as plausible loci for mediating antidepressant responses to NK, receptor antagonist administration. Considerable, additional basic science efforts to determine the neurochemical and physiologic role of SP in these various brain regions may yield further insight into the psychopharmacologic potential of NK, receptor antagonists.

Based on the limited research in this field, the role of SP neuronal pathways in depression is unclear. However, the findings of the first clinical trial of an NK, receptor antagonist in depression are encouraging and should stimulate further preclinical and clinical evaluation of these compounds. The limitations of current antidepressant medications, including the delay for a full therapeutic response, a substantial rate of nonresponders, and bothersome side effect profiles, merit the full exploration of all plausible agents with novel antidepressant mechanisms of action. ACKNOWLEDGMENTS

The authors are supported by NIH MH-42088, MH-39415 and the Conte Center for the Neurobiology of Mental Disorders (MH-58922). Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED 1. Charney DS, Berman RM, Miller HL. 1998. Treatment of depression. See Ref. 172, pp. 705-731 . Kramer MS, Cutler N, Feighner J, Shrivastava R, Carman J, et al. 1998. Distinct mech-

anism for antidepressant activity by blockade of central substance P receptors. Science 281:1640-45 . Maggio JE. 1988. Tachykinins. Annu. Rev. Neurosci. 11:13—28

. Chang MM, Leeman SE. 1970. Isolation of a sialogogic peptide from bovine hypothalamic tissue and its characterization as substance P. J. Biol. Chem. 245:4784-90

. Krause JE, Chirgwin JM, Carter MS, Xu ZS, Hershey AD. 1987. Three rat preprotachykinin mRNAs encode the neuropeptides substance P and neurokinin A. Proc. Natl. Acad. Sci. USA 84:881—85 6. Kotani H, Hoshimaru M, Nawa H, Nakanishi S. 1986. Structure and gene organization of bovine neuromedin K precursor. Proc. Natl. Acad. Sci. USA 83:7074-78 7. Lee CM, Iversen LL, Hanley MR,

Sand-

berg BE. 1982. The possible existence of multiple receptors for substance P. NaunynSchmiedebergs Arch. Pharmacol. 318:28187

NK1 ANTAGONISTS

. Mantyh PW, Maggio JE, Hunt SP. 1984. The autoradiographic distribution of kassinin and substance K binding sites is different from the distribution of substance P binding sites in rat brain. Eur. J. Pharmacol. 102:361—64 . Danks

JA, Rothman

RB, Cascieri MA,

Chicchi GG, Liang T, et al. 1986. A comparative autoradiographic study of the distributions of substance P and eledoisin binding sites in rat brain. Brain Research 385:273-81

10. Buck SH, Burcher E, Shults CW, Lovenberg W, O’Donohue TL. 1984. Novel pharmacology of substance K—binding sites: a third type of tachykinin receptor. Science 226:987-89 . Laufer R, Gilon C, Chorev M, Selinger

Z. 1986. Characterization of a neurokinin B receptor site in rat brain using a highly selective radioligand. J. Biol. Chem. 261:10257-63 12: Maggi CA. 2000. The troubled story of tachykinins and neurokinins. Trends Pharmacol. Sci. 21:173-75 12a. Watling KJ. 1988. Tachykinin receptors. In The RBI Handbook of Receptor Classification and Signal Transduction, 34 ed. pp. 112-113. Natick, MA: RBI 3) Regoli D, Boudon A, Fauchere J-L. 1994. Receptors and antagonists for substance P and related peptides. Pharmacol. Rev. 46:551-99 14. Yokota Y, Sasai Y, Tanaka K, Fujiwara T, Tsuchida

K, et al. 1989.

Molecular

characterization of a functional cDNA for rat substance P receptor. J. Biol. Chem. 264:17649-52 15: Hershey AD, Krause JE. 1990. Molecular characterization of a functional cDNA encoding the rat substance P receptor. Sci-

ence 247:958-62 16. Nakanishi S. 1991. Mammalian tachykinin receptors. Annu. Rev. Neurosci. 14: 123-36 fe Shigemoto R, Yokota Y, Tsuchida K, Nakanishi S. 1990. Cloning and expres

897

sion of a rat neuromedin K receptor cDNA. J. Biol. Chem. 265:623-28

18. Takeda Y, Chou KB, Takeda J, Sachais BS, Krause JE. 1991. Molecular cloning,

structural characterization and functional expression of the human substance P receptor. Biochem. Biophys. Research Commun. 179:1232—40 10). Gerard NP, Eddy RL, Shows TB, Gerard C. 1990. The human Neurokinin A (Substance K) receptor, molecular cloning of the gene, chromosome localization and isolation of cDNA for tracheal and gastric tissues. J. Biol. Chem. 20455—62

20. Buell G, Schulz MF, Arkinstall SJ, Maury K, Missotten

M,

et al. 1992.

Molecu-

lar characterisation, expression and localisation of human neurokinin—3 receptor. FEBS Lett. 299:90-95 210 Arai H, Emson PC. 1986. Regional distribution of neuropeptide K and other tachykinins (neurokinin A, neurokinin B and substance P) in rat central nervous system. Brain Research 399:240—-49 2D, Shults CW, Quirion R, Chronwall B, Chase TN, O’Donohue TL. 1984. A comparison of the anatomical distribution of substance P and substance P receptors in the rat central nervous system. Peptides 5:1097— 1128 23% Warden MK, Young WS, III. 1988. Distribution of cells containing mRNAs encoding substance P and neurokinin B in the rat central nervous system. J. Comp. Neurol. 272:90-113 24. Marksteiner J, Sperk G, Krause JE. 1992. Distribution of neurons expressing neurokinin B in the rat brain: immunohistochemistry and in situ hybridization. J. Comp. Neurol. 317:341-56 2a) Mantyh PW, Hunt SP, Maggio JE. 1984. Substance P receptors: localization by light microscopic autoradiography in rat brain using [>H]SP as the radioligand. Brain Research 307:147-65 26. Jones RS, Olpe HR. 1982. A structure— activity profile of substance P and some

898

Dik

28.

Do}

30.

STOUT

® OWENS

® NEMEROFF

of its fragments on supraspinal neurones in the rat. Neurosci. Lett. 33:67—71 Ku Y-H, Tan L, Li L-S, Ding X. 1998. Role of corticotropin-releasing factor and substance P in pressor responses of nuclei controlling emotion and stress. Peptides 19:677-82 Ogata N. 1979. Effects of substance P on neurons of various brain regions in vitro. Brain Research 176:395—400 Aguiar MS, Brandao ML. 1996. Effects of microinjections of the neuropeptide substance P in the dorsal periaqueductal gray on the behaviour of rats in the plus-maze test. Physiol. Behavior 60:1183-86 de Araujo JE, Huston JP, Brandéo ML. 1998. Aversive effects of the C-fragment of Substance P in the dorsal periaqueductal gray matter. Exp. Brain Research 123:84— 89

Sul. Kelley AE, Iversen SD. 1979. Substance P infusion into substantia nigra of the rat: behavioural analysis and involvement of striatal dopamine. Eur. J. Pharmacol. 60:171— 79 32) Elliott PJ, Alpert JE, Bannon MJ, Iversen SD. 1986. Selective activation of mesolimbic and mesocortical dopamine metabolism in rat brain by infusion of a stable substance P analogue into the ventral tegmental area. Brain Research 363:14547 38! Dam TV, Martinelli B, Quirion R. 1990, Autoradiographic distribution of brain neurokinin—1/substance P receptors using a highly selective ligand [3H]-[Sar9, Met(O2)11]-substance P. Brain Research 531:333-37 . Maeno H, Kiyama H, Tohyama M. 1993. Distribution of the substance P receptor (NK-1 receptor) in the central nervous system. Brain Research. Mol. Brain Research 18:43-58 3). Yip J, Chahl LA. 1999. Distribution of Fos-like immunoreactivity in guinea-pig brain following administration of the neurokinin-1 receptor agonist, [SAR9,MET-

(O2)11]substance

P.

Neuroscience

94:

663-73 36. Dam T-V, Escher E, Quirion R. 1990. Visualization of neurokinin-3 receptor sites in rat brain using the highly selective ligand (3H]senktide.

Brain

Research

506:175-—

1g. Sie Ding YQ, Shigemoto R, Takada M, Ohishi H, Nakanishi

38.

Sus

40.

41.

2.

S, et al. 1996. Localization

of the neuromedin K receptor (NK3) in the central nervous system of the rat. J. Comp. Neurol. 364:290-310 Shughme PJ, Lane MV, Merchenthaler I. 1996. In situ hybridization analysis of the distribution of neurokinin-3 mRNA in the rat central nervous system. J. Comp. Neurol. 372:395-414 Dam TV, Escher E, Quirion R. 1988. Evidence for the existence of three classes of neurokinin receptors in brain. Differential ontogeny of neurokinin-1, neurokinin-2 and neurokinin-3 binding sites in rat cerebral cortex. Brain Research 453:372-76 Hagan RM, Beresford IJ, Stables J, Dupere J, Stubbs CM, et al. 1993. Characterisation, CNS distribution and function of NK2 receptors studied using potent NK2 receptor antagonists. Regul. Pept. 46:9-19 Mai JK, Stephens PH, Hopf A, Cuello AC. 1986. Substance P in the human brain. Neuroscience 17:709—39 Nomura H, Shiosaka S, Tohyama M. 1987. Distribution of substance P-like immunoreactive structures in the brainstem of the adult human brain: animmunocytochemical study. Brain Research 404:36570

. Tiberi M, Lemay J, Magnan J. 1987. Specific binding of [3H]-substance P to various regions of the human foetal brain. Peptides 8:947-49 44. Jordan D, Kermadi I, Rambaud C, Gilly R, Bouvier R, et al. 1995. Regional distribution of substance P binding sites in the brainstem of the human newborn. Brain Research 695:117—24 45. Arai H, Emson PC, Carrasco LH.

NK1 ANTAGONISTS

1987. Huntington’s disease: changes in tachykinin content in postmortem brains. Ann. Neurol. 22:587-94

46. Chawla MK, Gutierrez GM, Young W. Sr., McMullen NT, Rance NE. 1997. Localization of neurons expressing substance P and neurokinin B gene transcripts in the human hypothalamus and basal forebrain. J. Comp. Neurol. 384:429-42 47. Dietl MM, Palacios JM. 1991. Phylogeny of tachykinin receptor localization in the vertebrate central nervous system: apparent absence of neurokinin-2 and neurokinin-3 binding sites in the human brain. Brain Research 539:2i1-22 48. Mileusnic D, Lee JM, Magnuson DJ, Hejna MJ, Krause JE, et al. 1999. Neurokinin-

3 receptor distribution in rat and human brain: an immunohistochemical study. Neuroscience 89:1269-90 49. Owens MJ, Stout SC, Mulchahey JJ, Plotsky PM. 1996. Molecular and neurobiological mechanisms in the treatment of psychiatric disorders. In Textbook of Psychiatry, ed.

A Tasman, J Kay, JA Lieber-

man. Philadelphia: WB Saunders 50. Chan-Palay V, Jonsson G, Palay SL. 1978. Serotonin and substance P coexist in neurons of the rat’s central nervous system. Proc. Natl. Acad. Sci. USA 75:158286

Sil Bjorklund AJ, Emson PC, Gilbert RF, Skagerberg G. 1979. Further evidence for the possible coexistence of 5-hydroxytryptamine and substance P in medullary raphe neurones of rat brain. Brit. J. Pharmacol. 66:112P-—13P 52: Hokfelt T, Ljungdahl A, Steinbusch H, Verhofstad A, Nilsson G, et al. 1978. Immunohistochemical evidence of substance P-like immunoreactivity in some 5-hydroxytryptamine-containing neurons in the rat central nervous system. Neuroscience 3:517-38 533 Jones RS, Olpe HR. 1984. Monoaminergic modulation of the sensitivity of neurones in the cingulate cortex to iontophoret-

899

ically applied substance P. Brain Research 311:297-305

54. Barden N, Daigle M, Picard V, Di Paolo T. 1983. Perturbation of rat brain serotonergic systems results in an inverse relation between substance P and serotonin concentrations measured in discrete nuclei. J. Neurochem. 41:834—40 SD: Baker KG, Halliday GM, Hornung JP, Geffen

LB, Cotton RG, et al. 1991. Distribution, morphology and number of monoamine-synthesizing and substance P-containing neurons in the human dorsal raphe nucleus. Neuroscience 42:757—75 56. Charara A, Parent A. 1998. Chemoarchitecture of the primate dorsal raphe nucleus. J. Chem. Neuroanat. 15:111-27 Sule Sergeyev V, Hékfelt T, Hurd Y. 1999. Serotonin and substance P co-exist in dorsal raphe neurons of the human brain. NeuroReport 10:3967-70

58. Unger T, Becker H, Petty M, Demmert G, Schneider B, et al. 1985. Differential

effects of central angiotensin II and substance P on sympathetic nerve activity in conscious rats. Implications for cardiovascular adaptation to behavioral responses. Circ. Research 56:563-—75

2). Brent PJ, Johnston PA, Chahl LA. 1988. Increased plasma catecholamines and locomotor activity induced by centrally administered substance P in guinea-pigs. Neuropharmacology 27:743-48 60. Guan ZL, Ding YQ, Li JL, Lu BZ. 1998. Substance P receptor-expressing neurons in the medullary and spinal dorsal horns projecting to the nucleus of the solitary tract in the rat. Neurosci. Research 30: 213-18 61. Jansen AS, Loewy AD. 1997. Neurons lying in the white matter of the upper cervical spinal cord project to the intermediolateral cell column. Neuroscience 77:889-98 62. Guyenet PG, Aghajanian GK. 1977. Excitation of neurons in the nucleus locus coeruleus by substance P and related peptides. Brain Research 136:178-84

900

STOUT

® OWENS

® NEMEROFF

iors and increases serotonergic function. Proc. Natl. Acad. Sci. USA In press

63. McLean S, Ganong AH, Seeger TF, Bryce DK, Pratt KG, et al. 1991. Activity and

distribution of binding sites in brain of a nonpeptide substance P (NK1) receptor antagonist. Science 251:437—39

64. Shen KZ, North RA. 1992. Substance P opens cation channels and closes potassium channels in rat locus coeruleus neurons. Neuroscience 50:345-—S3 65. Hahn MK, Bannon MJ. 1999. Stressinduced C-fos expression in the rat locus coeruleus is dependent on neurokinin 1 receptor activation. Neuroscience 94: 11183-88 66. Hahn MK, Bannon MJ. 1998. Tachykinin

Ie Bannon MJ, Elliott PJ, Alpert JE, Goedert M, Iversen SD, et al. 1983. Role of

F2:

Se

NK, receptor antagonists enhance stressinduced c-fos in rat locus coeruleus. Eur.

J. Pharmacol. 348:155-60 66a.

Haddjeri

N,

Blier

P. 2000.

74. Effect

of

neurokinin-1 receptor antagonists on the function of S-HT and noradrenaline neurons. Neuroreport 11:1323-27.

67. Lorens SA, Paris JM, Brodin E. 1991. Neurokinin innervation of the rat median raphe nucleus does not originate in the brain stem. Ann. NY Acad. Sci. 632:43134

qs.

68. Paris JM, Lorens SA. 1987. Intra—median raphe infusions of muscimol and the substance P analogue DiMe-C7 produce hyperactivity: role of serotonin neurons. Behav. Brain Research 26:139-51 69, Mason GS, Elliott PJ. 1992. Behavioural consequences following infusion of selective neurokinin agonists into the median raphe nucleus of the rat. Neuropharmacology 31:757-60 70. Gradin K, Qadri F, Nomikos GG, Hille-

76.

76a.

77

gaart V, Svensson TH. 1992. Substance P

injection into the dorsal raphe increases blood pressure and serotonin release in hippocampus of conscious rats. Eur. J. Pharmacol. 218:363-67 70a. Santarelli L, Gobbi G, Debs PC, Sibille

78

EL, Blier P, et al. 2001. Genetic and phar-

macological disruption of NK1 receptor function decreases anxiety-related behay-

19

endogenous substance P in stress-induced activation of mesocortical dopamine neurones. Nature 306:791-—92 Bannon MJ, Deutch AY, Tam SY, Zamir N, Eskay RL, et al. 1986. Mild footshock stress dissociates substance P from substance K and dynorphin from Met- and Leu-enkephalin. Brain Research 381:1986 Walker PD, Riley LA, Hart RP, Jonakait GM. 1991. Serotonin regulation of tachykinin biosynthesis in the rat neostriatum. Brain Research 546:33-39 Chan-Palay V, Palay SL. 1977. Ultrastructural identification of substance P cells and their processes in rat sensory ganglia and their terminals in the spinal cord by immunocytochemistry. Proc. Natl. Acad. Sci. USA 74:4050-54 Randic M, Miletic V. 1977. Effect of substance P in cat dorsal horn neurons activated by noxious stimuli. Brain Research 128:164-69 Yaksh TL, Farb DH, Leeman SE, Jessell TM. 1979. Intrathecal capsaicin depletes substance P in the rat spinal cord and produces prolonged thermal analgesia. Science 206:481-83 Hill R. 2000. NK1 (substance P) receptor antagonists—why are they not analgesic inhumans? Trends Pharmacol. Sci. 21:244-6 Beattie DT, Connor HE, Hagan RM. 1995. Recent developments in tachykinin NK1 receptor antagonists: prospects for the treatment of migraine headache. Can. J. Physiol. Pharmacol. 73:871-77 Rupniak NM, Kramer MS. 1999. Discovery of the antidepressant and anti-emetic efficacy of substance P receptor (NK1) antagonists. Trends Pharmacol. Sci. 20: 485-90 Elliott PJ. 1988. Place aversion induced

NK1 ANTAGONISTS

by the substance P analogue, dimethyl-C7, is not state dependent: implication of substance P in aversion. Exp. Brain Research 73:354-56 80. Teixeira RM, Santos AR, Ribeiro SJ, Calixto JB, Rae GA, et al. 1996. Effects of central administration of tachykinin receptor agonists and antagonists on plus-maze behavior in mice. Eur. J. Pharmacol. 311:714 81. Aguiar MS, Brandao ML. 1994. Conditioned place aversion produced by microinjections of substance P into the periaqueductal gray of rats. Behav. Pharmacol. 5:369-73 82. Hasenohrl RU, Jentjens O, De Souza Silva MA,

83.

84.

85.

86.

87.

Tomaz

C, Huston

JP. 1998.

Anxiolytic-like action of neurokinin substance P administered systemically or into the nucleus basalis magnocellularis region. Eur. J. Pharmacol. 354:123-33 Hasendhri RU, Gerhardt P, Huston JP. 1992. Positively reinforcing effects of the neurokinin substance P in the basal forebrain: mediation by its C-terminal sequence. Exp. Neurol. 115:282-91 Nikolaus S, Huston JP, Hasendhrl RU. 1999. Reinforcing effects of neurokinin substance P in the ventral pallidum: mediation by the tachykinin NK1 receptor. Eur. J. Pharmacol. 370:93-99 Isaacson RL, Hannigan JHJ, Brakkee JH, Gispen WH. 1983. The time course of excessive grooming after neuropeptide administration. Brain Research Bulletin 11:289-293 Kulkosky PJ. 1988. Bombesin and ceruletide-induced grooming and inhibition of ingestion in the rat. Ann. NY Acad. Sci. 525:201-18 Drago F, Contarino A, Busa L. 1999. The expression of neuropeptide-induced excessive grooming behavior in dopamine D1 and D2 receptor-deficient mice. Eur. J.

Pharmacol. 365:125—31 88. Jones DN, Kortekaas R, Slade PD, Middlemiss DN, Hagan JJ. 1999. The behavioural

901

effects of corticotropin-releasing factorrelated peptides in rats. Psychopharmacology 138:124-32 89. Van Wimersma Greidanus TB, Maigret C. 1988. Grooming behavior induced by substance P. Eur. J. Pharmacol. 154:217— 20 90. Katz RJ. 1979. Central injection of substance P elicits grooming behavior and motor inhibition in mice. Neurosci. Lett. 12:133-36 Die Ravard S, Betschart J, Fardin V, Flamand O, Blanchard JC. 1994. Differential abi-

lity of tachykinin NK-1 and NK-2 agonists to produce scratching and grooming behaviours in mice. Brain Research 651:199— 208 oy, Stoess! AJ, Brackstone M, Rajakumar N, Gibson CJ. 1995. Pharmacological characterization of grooming induced by a selective NK-1 tachykinin receptor agonist. Brain Research 700:115—20 we. Hylden JL, Wilcox GL. 1981. Intrathecal substance P elicits a caudally-directed biting and scratching behavior in mice. Brain Research 217:212-15 94. Bristow LJ, Young L. 1994. Chromodacryorrhea and repetitive hind paw tapping: models of peripheral and central tachykinin NK1 receptor activation in gerbils. Eur. J. Pharmacol. 253:245-52 95. Piot O, Betschart J, Grall I, Ravard S, Garret C, et al. 1995. Comparative behavioural profile of centrally administered tachykinin NK1, NK2 and NK3 receptor agonists in the guinea-pig. Brit. J. Pharmacol. 116:2496—2502 96. Ribeiro SJ, De Lima TC. 1998. Naloxoneinduced changes in tachykinin NK3 receptor modulation of experimental anxiety in mice. Neurosci. Lett. 258:155—-58 OTe Stout SC, Nemeroff CB. 1994. Stress and psychiatric disorders. Seminars in the Neurosciences 6:271-80 98. Stout SC, Kilts CD, Nemeroff CB. 1995. Neuropeptides and stress: preclinical findings and implications for pathophysiology.

902

STOUT

® OWENS

® NEMEROFF

In Neurobiological and Clinical Consequences of Stress: From Normal Adaptation to PTSD, ed. A. Deutch. New York:

107.

Raven

99. Lisoprawski A, Blanc G, Glowinski J. 1981. Activation by stress of the habenulo-interpeduncular substance P neurons in the rat. Neurosci. Lett. 25:47-51 100. Siegel RA, Diiker E-M, Fuchs E, Pahnke

108.

U, Wuttke W. 1984. Responsiveness of mesolimbic, mesocortical, septal and hip-

pocampal cholecystokinin and substance P neuronal systems to stress, in the male

109.

rat. Neurochem. Int. 6:783—89

101. Siegel RA, Duker EM, Pahnke U, Wuttke W. 1987. Stress-induced changes in cholecystokinin and substance P concentrations in discrete regions of the rathypothalamus. Neuroendocrinology 46:75— 81 . Takayama H, Ota Z, Ogawa N. 1986. Effect of immobilization stress on neuropeptides and their receptors in rat central nervous system. Regul. Pept. 15:239-248 103. Rosen A, Brodin K, Eneroth P, Brodin E. 1992. Short-term restraint stress and s.c. saline injection alter the tissue levels of substance P and cholecystokinin in the peri-aqueductal grey and limbic regions of rat brain. Acta Physiol. Scand. 146:341-48 104. Brodin E, Rosen A, Schott E, Brodin K. 1994. Effects of sequential removal of rats from a group cage, and of individual housing of rats, on substance P, cholecystokinin and somatostatin levels in the periaqueductal grey and limbic regions. Neuropeptides 26:253—60 10S. Nakamura H, Moroji T, Nohara S, Nakamura H, Okada A. 1990. Effects of wholebody vibration stress on substance P- and neurotensin-like immunoreactivity in the rat brain. Environ. Research 52:155—63 106. Chowdrey HS, Larsen PJ, Harbuz MS, Lightman SL. 1995. Endogenous substance P inhibits the expression of corticotropin-releasing hormone during

110.

UE

a chronic inflammatory stress. Life Sct. 57:2021-29 Nussdorfer GG, Malendowicz LK. 1998. Role of tachykinins in the regulation of the hypothalamo-pituitary-adrenal axis. Peptides 19:949-68 Mantyh PW, Allen CJ, Ghilardi JR, Rogers SD, Mantyh CR, etal. 1995. Rapid endocytosis of a G protein-coupled receptor: substance P evoked internalization of its receptor in the rat striatum in vivo. Proc. Natl. Acad. Sci. USA 92:2622-26 Garland AM, Grady EF, Lovett M, Vigna SR, Frucht MM, et al. 1996. Mechanisms of desensitization and resensitization of G protein-coupled neurokinin1 and neurokinin2 receptors. Mol. Pharmacol. 49:438—46 Yu SS, Lefkowitz RJ, Hausdorff WP. 1993. Beta-adrenergic receptor sequestration: a potential mechanism of receptor sensitization. J. Biol. Chem. 268:33741 Smith DW, Hewson L, Fuller P, Williams AR, Wheeldon A, et al. 1999. The substance P antagonist L-760,735 inhibits stress-induced NK(1) receptor internali-

sation in the basolateral amygdala. Brain Research 848:90—95 1G ee Winslow JT, Insel TR. 1991. The infant rat separation paradigm: a novel test for novel anxiolytics. Trends Pharmacol. Sci. 12:402-04 ius, Ladd CO, Huot RL, Thrivikraman KV, Nemeroff CB, Meaney MJ, et al. 2000. Long-term behavioral and neuroendocrine adaptations to adverse early experience. Prog. Brain Research 122:81— 103 114. De Felipe C, Herrero JF, O’Brien JA, Palmer JA, Doyle CA, et al. 1998. Altered nociception, analgesia and aggression in mice lacking the receptor for substance P. Nature 392:394-97 U1). Rupniak NMJ, Carlson EC, Harrison T, Oates B, Seward E, et al. 2000. Pharmacological blockade or genetic deletion of

NK1 ANTAGONISTS

substance P (NKj) receptors attenuates neonatal vocalisation in guinea—pigs and mice. Neuropharmacology 39:1413-21

116. Cao YQ, Mantyh PW, Carlson EJ, Gillespie AM,

Epstein CJ, et al. 1998.

Pri-

mary afferent tachykinins are required to experience moderate to intense pain. Nature 392:390-94

117. Shirayama Y, Mitsushio H, Takashima M, Ichikawa H, Takahashi K. 1996. Reduc-

tion of substance P after chronic antidepressants treatment in the striatum, substantia nigra and amygdala of the ra. Brain Research 739:70-78 118. Brodin K, Ogren SO, Brodin E. 1994. Clomipramine and clonazepam increase cholecystokinin levels in rat ventral tegmental area and limbic regions. Eur. J. Pharmacol. 263:175-80 fee Oblin A, Zivkovic B, Bartholini G. 1984. Involvement of the D-2 dopamine receptor in the neuroleptic-induced decrease in nigral substance P. Eur. J. Pharmacol. 105:175-77 120. Hamon M, Gozlan H, Bourgoin S, Benoliel JJ, Mauborgne A, et al. 1987. Opioid receptors and neuropeptides in the CNS in rats treated chronically with amoxapine or amitriptyline. Neuropharmacology 26:531-39 12Ae Manier DH, Gillespie DD, SandersBush E, Sulser F. 1987. The serotonin/

noradrenaline-link in brain. I. The role of noradrenaline and serotonin in the regulation of density and function of beta adrenoceptors and its alteration by desipramine. Naunyn—Schmiedebergs Arch. Pharmacol. 335:109-14 122 Blier P, de Montigny C, Chaput Y. 1987. Modifications of the serotonin system by antidepressant treatments: implications for the therapeutic response in major depression. J. Clin. Psychopharmacol. 7:24S-35S 18: Baker GB, Greenshaw AJ. 1989. Effects of long-term administration of antidepressants and neuroleptics on receptors

903

in the central nervous system. Cell. Mol. Neurobiol. 9:1-44

124. Nestler EJ, McMahon

A, Sabban EL, Tallman JF, Duman RS. 1990. Chronic an-

tidepressant administration decreases the expression of tyrosine hydroxylase in the rat locus coeruleus. Proc. Natl. Acad. Sci. USA 87:7522-26

125. Brodin E, Ogren SO, TheodorssonNorheim E. 1987. Effects of subchronic treatment

with

imipramine,

zimelidine

and alaproclate on regional tissue levels of substance P- and neurokinin A/ neurokinin B—like immunoreactivity in the brain and spinal cord of the rat. Neuropharmacology 26:581-90 126. Rosén A, Franck J, Brodin E. 1995. Effects of acute systemic treatment with the 5 HT-uptake blocker alaproclate on tissue levels and release of substance P in rat periaqueductal grey. Neuropeptides 28: 317-24 WET Riley LA, Hart RP, Jonakait GM.

1991. Both zimelidine and clorgyline decrease preprotachykinin mRNA in adult medullary raphe nuclei. Molecular and Cellular Neurosciences 2:139-44 128. Walker PD, Riley LA, Hart RP, Jonakait GM. 1991. Serotonin regulation of neostriatal tachykinins following neonatal 6-hydroxydopamine lesions. Brain Research 557:31-36 120% Stout SC, Owens MJ, Nemeroff CB. 1999. Preprotachykinin-A and neurokinin 1 receptor mRNA expression are unchanged in rat forebrain after chronic antidepressant administration. Soc. Neurosci. Abstr. 25:1066 130. Zachrisson O, Mathé AA, Lindefors N. 1997. Decreased levels of preprotachykinin—A and tachykinin NK1 receptor mRNA in specific region of the rat striatum after electroconvulsive stimuli. Eur. J. Pharmacol. 319:191—95 iil. Stenfors C, Srinivasan GR, Theodorsson E, Mathé AA. 1992. Electroconvulsive stimuli and brain peptides: effect of

904

STOUT

® OWENS

® NEMEROFF

modification of seizure duration on neuropeptide Y, neurokinin A, substance P and neurotensin. Brain Research 596: 251-58

SZ

Brené S, Lindefors N, Herrera—Marschitz

M, Persson H. 1993. Differential regulation of preprotachykinin-A mRNA expression in striatum by excitation of hippocampal neurons. Eur J. Neurosci. 5:839-45

1333 Marcus MM, Nomikos GG, Malmerfelt A, Zachrisson O, Lindefors N, et al. 1997.

Effect of chronic antipsychotic drug treatment on preprosomatostatin and preprotachykinin A mRNA levels in the medial prefrontal cortex, the nucleus accumbens

and the caudate putamen of the rat. Mol. Brain Research 45:275—82

134. Figiel GS, McDonald WM, McCall WV, Zorumpski C. 1998. Electroconvulsive therapy. See Ref. 172, pp. 523-45 135), Snider RM, Constantine JW, Lowe JAd, Longo KP, Lebel WS, et al. 1991. A potent nonpeptide antagonist of the substance P (NK1) receptor. Science 251:435-37 136. Beresford IJ, Birch PJ, Hagan RM, Ireland SJ. 1991. Investigation into species variants in tachykinin NK1 receptors by use of the non-peptide antagonist, CP96,345. Brit. J. Pharmacol.

Sie

104:292-93

Gitter BD, Waters DC, Bruns RF, Mason NR, Nixon JA, et al. 1991. Species differences in affinities of non-peptide antagonists for substance P receptors. Eur. J. Pharmacol. 197:237-38 138. Barr AJ, Watson SP. 1993. Non-peptide antagonists, CP-96,345 and RP 67580, distinguish species variants in tachykinin NK1 receptors. Brit. J. Pharmacol. 108: 223-27 NSP), Porsolt RD, LePichon M, Jalfre M. 1977. Depression: a new animal model sensitive to antidepressant treatments. Nature 266:730-32 140. Vassout A, Schlaub M, Gentsch C, Ofner S, Schilling W, et al. 1994. CGP 49823, anovel NK1 receptor antagonist: behavi-

oural effects. Neuropeptides 26 suppl 1:38 141. Zemig G, Dietrich H, Maggi CA, Saria A. 1992. The substance P (NK1) receptor antagonist (+/—)-CP-96,345 causes sedation and motor impairment in Swiss albino mice in the black-and-white box behavioral paradigm. Neurosci. Lett. 143: 169-72 142. Guard S, Boyle SJ, Tang KW, Watling KJ, McKnight AT, et al. 1993. The interaction of the NK1 receptor antagonist CP-96,345 with L-type calcium channels and its functional consequences. Brit. J. Pharmacol. 110:385—91 143. File SE. 1997. Anxiolytic action of a neurokininl receptor antagonist in the social interaction test. Pharmacol. Biochem. Behay. 58:747-52 144. Tschdpe C, Picard P, Culmann J, Prat A, Itoi K, etal. 1992. Use of selective antagonists to dissociate the central cardiovascular and behavioural effects of tachykinins on NK1 and NK2 receptors in the rat. Brit. J. Pharmacol. 107:750—55 145. Culman J, Klee S, Ohlendorf C, Unger T. 1997. Effect of tachykinin receptor inhibition in the brain on cardiovascular and behavioral responses to stress. J. Pharmacol. Exp. Ther. 280:238-46 146. Krase W, Koch M, Schnitzler HU. 1994. Substance P is involved in the sensitization of the acoustic startle response by footshocks in rats. Behav. Brain Research 63:81-88 147. Kungel M, Ebert U, Herbert H, Ostwald J. 1994. Substance P and other putative transmitters modulate the activity of reti-_ cular pontine neurons: an electrophysiological and immunohistochemical study. Brain Research 643:29-39 148. Hitchcock JM, Sananes CB, Davis M. 1989. Sensitization of the startle reflex by footshock: blockade by lesions of the central nucleus of the amygdala or its efferent pathway to the brainstem. Behav, Neurosci. 103:509-18 149. Shaikh MB, Steinberg A, Siegel A. 1993.

NK1 ANTAGONISTS

Evidence that substance P is utilized in medial amygdaloid facilitation of defensive rage behavior in the cat. Brain Research 625:283—94

150. Walsh DM, Stratton SC, Harvey FJ, Beresford IJ, Hagan RM. 1995. The anxiolytic-like activity of GR159897, a non-peptide NK2 receptor antagonist, in rodent and primate models of anxiety. Psychopharmacology 121:186-91

ile

Garret C, Carruette A, Fardin V, Moussaoui S, Peyronel JF, et al. 1991. Pharma-

cological properties of a potent and selective nonpeptide substance P antagonist. Proc. Natl. Acad. Sci. USA 88:10208— 12 152. Sachais BS, Krause JE. 1994. Both extracellular and transmembrane residues contribute to the species selectivity of the neurokinin-1 receptor antagonist WIN 51708. Mol. Pharmacol. 46:122-28 153: Aramori I, Morikawa N, Zenkoh J, O’Donnell

N, Iwami

M,

et al. 1994.

Subtype— and species-selectivity of a tachykinin receptor antagonist, FK888, for cloned rat and human tachykinin receptors. Eur. J. Pharmacol. 269:277-81

154. Pozza MF, Olpe HR, Ofner S, Schoeffel C, Evans RH. 1998. The potency of the novel tachykinin receptor antagonist CGP49823 at rat and gerbil motoneurones in vitro. Eur, J. Pharmacol. 342:20308 X, Doutremepuich JD, 159) Emonds-Alt Heaulme M, Neliat G, Santucci V, et al. 1993. In vitro and in vivo biological activities of SR140333,

a novel

potent non-peptide tachykinin NK1 receptor antagonist. Eur. J. Pharmacol. 250:403-13 HE,

156. Beattie DT, Beresford IJ, Connor Marshall FH, Hawcock AB, et al. 1995.

The pharmacology of GR203040, a novel, potent and selective non-peptide tachykinin NK1 receptor antagonist. Brit. J. Pharmacol. 116:3149-57 ISI, Walpole CS, Brown MC, James IF, Camp-

905

bell EA, McIntyre P, et al. 1998. Compar-

ative, general pharmacology of SDZ

NKT

343, a novel, selective NK1 receptor an-

tagonist. Brit. J. Pharmacol. 124:83-92 158. McLean S, Ganong A, Seymour PA, Bryce

DK, Crawford RT, et al. 1996. Characterization of CP-122,721; a non-

peptide antagonist of the neurokinin NK1 receptor. J. Pharmacol. Exp. Ther. 277: 900-08 IK332), Iyengar S, Hipskind PA, Gehlert DR, Schober D, Lobb KL, et al. 1997. LY 303870, a centrally active neurokinin—1

antagonist with a long duration of action. J. Pharmacol. Exp. Ther. 280:774-85 160. Nemeroff CB, Widerlév E, Bissette G, Walleus H, Karlsson I, et al. 1984. Ele-

vated concentrations of corticotropin— releasing factor—like immunoreactivity in depressed patients. Science 226:1342—44

161. Widerlov E, Bissette G, Nemeroff CB. 1988.

Monoamine

metabolites,

corti-

cotropin releasing factor and somatostatin as CSF markers in depressed patients. J. Affect. Disord. 14:99-107

162. Risch SC, Nemeroff CB. 1992. Neurochemical alterations of serotonergic neuronal systems in depression. J. Clin. Psychiatry 53 Suppl:3-7 163. RimonR, Le Greves P, Nyberg F, Heikkila L, Salmela L, et al. 1984. Elevation of sub-

stance P-like peptides in the CSF of psychiatric patients. Biol. Psychiatry 19:509— 16 164. Heikkila L, Rim6n R, Terenius L. 1990. Dynorphin A and substance P in the cerebrospinal fluid of schizophrenic patients. Psychiatry Research 34:229-36

165. Berrettini WH, Rubinow DR, Nurmberger JIJ, Simmons-Alling

S, Post RM, et al.

1985. CSF substance P immunoreactivity in affective disorders. Biol. Psychiatry 20:965-70 166. Almay BG, Johansson F, Von Knorring L, Le Greves P, Terenius L. 1988. Substance P in CSF of patients with chronic pain syndromes. Pain 33:3-9

906

STOUT

® OWENS

® NEMEROFF

167. Martensson B, Nyberg S, Toresson G, Brodin E, Bertilsson L. 1989. Fluoxetine

treatment of depression. Clinical effects, drug concentrations and monoamine metabolites and N-terminally extended substance P in cerebrospinal fluid. Acta Psychiatr. Scand. 79:586—96

168. Burnet PW, Harrison PJ. 1999. Substance P (NK1) receptors in the cingulate cortex

in unipolar and bipolar mood disorder and schizophrenia. Biol. Psychiatry 47:80— 83

169. Rosen RC, Lane RM, Menza M. 1999. Effects of SSRIs on sexual function: a

critical review. J. Clin. Psychopharmacol. 19:67-85 170. Hirschfeld RM. 1999. Management of sexual side effects of antidepressant therapy. J. Clin. Psychiatry 60:27-30 LA. Kramer MS, Alexander R, Ball WA, Shedden A, Reines SA. 2000. The therapeutic potential of substance P (NK-1 receptor)

antagonists

(SPA).

Neuropsy-

chopharmacology 23:822 Liz: AF Schatzberg, CB Nemeroff, eds. 1998. Textbook of Psychopharmacology. Washington, D.C.: The American Psychiatric. 2nd ed.

SUBJECT INDEX A ACTH-secreting tumors V3R antagonists and treatment of, 190, 196 Activating transcription factor-1 (ATF-1),

487-88 Adenosine, 775-84 inflammation and, 475-84 ischemia and, 775-84 metabolism, 777—78 tissue protection and, 775,

780-81 Adenosine receptors, 775—84 characteristics, 776-77

mast cells and, 775, 781-82 pharmacology, 782-84 preconditioning and, 775 Adenylyl cyclases (ACs), 146-62 alcohol abuse and, 155-57 brain function and, 147—50

brown adipose tissue and, 153 calcium-inhibitable,

150-52 ceil differentiation and,

152-53 cyclic AMP synthesis and, 145 drug dependency and, 154-55 genetic diseases and, 157 genetic mice, 157-58 isoforms, 145-62

pharmacology of, 158-59 adenosine derivatives and, 159 forskolin and, 158 spermatozoa and, 153-54 tissue distribution, 146-48

ADP-ribosyl cyclase, 317-35 structure and mechanisms,

318-20 Ah receptor, 297-303 cytochrome P450 CYP1B1 expression and, 297-303

human, 300-1 rodent, 302-3 tumor, 297, 299

A-kinase anchoring proteins (AKAPs), 763-64, 766 Alcohol abuse cyclic AMP signaling pathway and, 155-57 Alzheimer’s disease drug treatment effects on, 625-26, 630-31, 637-39, 642-43 estrogen’s protective effects against, 570, 578, 582 glycogen synthase kinase-3 and, 802-3 pharmacogenetics and, 108 f-Aminobutyric acid See GABA y-Amyloid peptides formation and toxicity estrogen’s protective effects against, 582-83 Anesthetics (general)

gaseous, 27, 32, 42-43 ligand-gated ion channels and molecular modeling of interactions with, 34-40 recombinant, 25—27 minimum alveolar concentration, 25

molecular mechanisms of,

23-43 molecular sites of action,

32-34 Angiogenesis antiangiogenic activity of vitamin D, 421, 423, 433-35 Anticancer chemotherapy comparative genomic hybridization and, 110-11

See also Cancer chemoprevention Anticancer drugs cytochrome P450 CYP1B1 and, 299, 306, 309, 311 topoisomerase I-targeting, 53-64 camptochecin and,

54-61, 63-64 mechanisms of topisomerase I inhibition, 5S—56 topoisomerease II-targeting, 53, 64-68

daunorubicin, 64 doxorubicin, 64 mechanisms of inhibition, 64-66 mitoxantrone, 64

SUMO and, 53, 63 ubiquitin and, 53, 61-62

Antidepressant neurokinin; receptor antagonists as, 877-96 substance P and, 877-96 effects of antidepressant drugs on, 887 monoamine-substance P interactions, 881-83

stress and, 885-86

907

908

SUBJECT INDEX

Antidiuretic aquaretics and, 175, 186-90, 195 arginine vasopressin (AVP), 175-96 Antioxidants inhibition of peroxynitrite-mediated oxidation and nitration

SUMO and, 53, 63 ubiquitin and, 53, 61-62

Apoptosis camptothecin-induced, 57-59 DNA damage and, 367,

reactions, 223-34

386-89 p53 and, 388-89 receptor-mediated, 388 thioredoxin and, 273-75

thioredoxin, 269-70

topoisomerase II cleavage

Antipsychotics, 237-57 adrenergic receptor blocking properties, 24446 potential psychotogenic pathways and sites of

complexes and, 67—68 vitamin D’s proapoptotic

action, 246

psychotogenesis and hyperdopaminergic model of, 238, 242-44, 249-52 hypoglutaminergic model of, 238, 242-45,

249-52 monoamine-glutamateGABA interactions and, 237-57

Antisense oligonucleotides, 403-14 bel-2, 403, 411-14 phosphodiester, 403-14 phosphorothiate, 403-14

RNase H and, 403-4, 408-10 Antitumor drugs

topoisomerase I-targeting, 53, 55-64 camptothecin, 54-61, 63-64 topoisomerase II-targeting, 53, 64-68 daunorubicin, 64

doxorubicin, 64

mitoxantrone, 64 mechanisms of inhibition, 64-66

activities, 421, 423,

433-35 Apoptosis signal-regulating kinase | (ASK1),

273-75 Aquaretics, 175, 186-90, 195

nonpeptide V>-renal arginine vasopressin receptor antagonists and, 175, 186-90, 195 L-Arginine, 79-91 biochemistry of, 79-81 cardiovascular effects, 79, 83-88 endothelial function and, 79, 81-91 mechanisms of action, 79, 81, 85-88 nitric oxide synthase substrate provision and, 79, 81, 85-88 nitric oxide and, 79-83 as precursor for, 81-83 paradox, 91 pharmacokinetics, 79, 89-91 side effects, 88-89 vasodilation and, 79, 81, 83-88 Arginine vasopressin (AVP), 175-96 Arginine vasopressin/oxytocin (AVP/OT) receptors

molecular biology of, 175, 179-82, 191-92 Arginin vasopressin (AVP) receptor agonists nonpeptide, 175-76, 182-90, 192-96 strategies in designing, 182 therapeutic uses ACTH-secreting tumors, 190 aquaretics, 175, 186-90, 195 arterial hypertension, 175-76 cerebral edema, 188-89 congenital nephrogenic diabetes insipidus, 189 congestive heart failure, 186, 188, 195-96 dysmenorthea, 175-76, 195, hypernatremia, 175, 186, 195 liver cirrhosis, 175,

186-88 nephrotic syndrome, 175 syndrome of inappropriate secretion of antidiuretic hormone (SIADH), 175-76, 186 Arginine vasopressin (AVP) receptor subtypes V-vascular (V;R), 175-78, 182-85, 189-90 V2-renal (V2R), 175-76, 186-90 V3-pituitary (V3R),

175-76, 178-79, 190 Arnt-Schultz law, 16

Arterial hypertension nonpeptide vasopressin receptor agonists and treatment of, 175-76

Asthma adenosine receptors and, 775, 781-82

SUBJECT INDEX

Asymmetric dimethylarginine (ADMA), 86 Atherosclerosis endothelial function and L-arginine and, 79, 82-85

tyrosine nitration and, 212 Autocoids prostaglandins, 661-81

See also Prostaglandins

B Basal vascular tone endothelin and, 851,

856-58 Bel-2 antisense oligonucleotides, 403, 411-14 Bioethics pharmacogenetics and, 115-16 Bioinformatics, 103, 115 Biomarkers, 347-63

defined, 348 in drug development, 347-63 drug discovery and preclinical development, 347-63 early clinical development, 351-52 late clinical development, 352-53

evaluation of, 353-56 as indicators of drug exposure, 347 pharmacokineticpharmacodynamic studies, 351

surrogate endpoints, 347-58 validation as, 356-58 unanticipated adverse consequence of drug therapy and, 349-50 Bipolar disorder lithium treatment for, 789-91, 803

Bone remodeling stress-induced osteopontin and, 723, 733-36 Botulinum toxin pharmacological modulation of lower urinary tract function and, 709-10 Brain function adenylyl cyclases and, 147-50 Breast cancer osteopontin expression and, 737

Brown adipose tissue adenylyl cyclases and, 153

Cc CAAT-enhancer binding protein (c/EBP), 487-88 Calabrese, 16 Calcium/calmodulin-

dependent protein kinases

See Calcium signaling, Calmodulin kinases Calcium messengers, 317

multiple, 317 Calcium signaling in blood cells, 330-31 calmodulin (CaM) and, 471-93 calmodulin kinases and, 471-93 CaM KI, 477-93 CaM KII, 482-93 CaM KIV, 479-93 CaM KK, 480-82 in cardiac and smooth muscle cells, 330-31 homeostasis and, 471-72

inositol trisphosphate (IP3) and, 317-18, 334-35 intercellular, 333-34

in lymphatic cells, 330-31 nicotinic acid adenine

909

dinucleotide phosphate (NAADP) mediated, 318-35 pharmacology of,

320-22 nitric oxide and mediation of, 325-26 in oocytes, 327-28 in pancreatic acinar cells, 328-29 transcription regulation, 484-93 activating transcription factor-1 (ATF-1) and,

487-88 CAAT-enhancer binding protein (c/EBP) and,

487-88 cAMP-responsive binding element (CREB) and, 484-86, 492 2/histone deacetylase and, 489-91

myocyte enhancer factor-2 (MEF2) and,

489-91 serum response factor (SRF) and, 487-89 Calcium stores, 317, 334-35 Calmodulin (CaM), 471-93 Calmodulin kinases, 471-93

calcium signaling and, 471-93 CaM KI, 477-93 CaM KII, 482—93 CaM KIV, 479-93 CaM KK, 480-82 domain structure, 473-77 transcription regulation, 484-93 activating transcription factor-1 (ATF-1) and, 487-88 CAAT-enhancer binding protein (c/EBP) and, 487-88

910

SUBJECT INDEX

cAMP-responsive binding element (CREB) and, 484-86, 492 2/histone deacetylase and, 489-91 myocyte enhancer factor-2 (MEF2) and,

489-91 serum response factor (SRF) and, 487-89 cAMP See Cyclic AMP Camptothecin, 53-61, 63-64

apoptotic cell death and, 57-59 cytotoxicity replication collision model for, 58 topoisomerase I and,

53-61, 63-64 Cancer breast osteopontin expression and, 737 colon, 424-26

proapoptotic activities of vitmin D, 424-26 colorectal, 424-25 prostate, 425 skin, 426 Cancer chemoprevention vitamin D and synthetic analogs for, 421-35 antiangiogenic activity of, 421, 423, 433-35 antiproliferative activity of, 421, 423, 433-35 cheek pouch carcinogenesis and, 426 colon cancer and, 425-26 colorectal cancer and, 424-25 gastrointestinal carcinogenesis and, 426

proapoptotic activity of, 421, 423, 433-35 prodifferentiating activity of, 421, 423,

433-35 prostate cancer and, 425 skin carcinogenesis and, 425-26 sun exposure and cancer rates and, 423-25 See also Anticancer chemotherapy Cancer drugs resistance to thioredoxin and, 276 Cancer research cytochrome P450 CYP1B1 and, 299, 306, 309, 311 as early marker for tumorigenesis, 309, 311 Capsaicin phramacological modulation of lower urinary tract function and, 708-9 Carcinogenesis DNA cleavage and, 53-68 cheek pouch, 426 gastrointestinal, 426 skin, 426 See also DNA cleavage; Tumorigenesis Carcinogens cytochrome P450 CYP1B1 metabolism and, 297, 299, 3034 Cardiac hypertrophy endothelin and, 851, 858-60 Cardiovascular disease L-arginine’s beneficial effects chronic oral administration, (ESS intermittent infusion therapy, 79, 84-85

vasodilation and, 79,

81-88 nitric oxide and, 81-88 Catalase, 216 CD38, 317-20 structure and mechanisms, 318-20 Cell cycle checkpoints activation in response to DNA damage, 379-83,

389 checkpoint Rad pathway, 379 p53 and, 384-86 Cell survival osteopontin stimulation of, 723, 728 Cellular immunity osteopontin and, 723, 730-31 Central cardiorespiratory regulation endothelin and, 851, 858-62 Cerebral edema nonpeptide V>-renal arginine vasopressin receptor antagonists as treatment of, 188-89 Chicago Toxicology Laboratory, 2—5 Clinical endpoints defined, 348 intermediate endpoint, 348 surrogate endpoints, 347-50, 356-63 biomarkers and, 347-63 defined, 348

ultimate outcome and, 348 Clonidine, 245 Clozapine, 252, 256

Colon cancer proapoptotic activities of vitamin D, 424-26 Comparative genomic hybridization, 110-11 Computational biology, 3

SUBJECT INDEX

Congenital nephrogenic diabetes insipidus nonpeptide V>2-renal arginine vasopressin receptor antagonists and treatment of, 189 Congestive heart failure dual V; R/V2R antagonists

and treatment of, 186, 188, 195-96 Constitutive active receptor (CAR) genes regulated by, 136 nuclear activation of,

133-34 nuclear translocation of,

132-33 phenobarbital and, 123, 131-38 Cyclic ADP-ribose (CADPR),

317-35 agonistic and antagonistic analogs of, 321 calcium signaling and, 317-35 elevated levels after cell activation, 324-25 nitric oxide signaling pathways and, 325 ryanoldine receptor modulation and, 322-23 Cyclic AMP (cAMP) adenylyl cyclases and, 145—46, 152, 155-62 cell differentiation and,

152-53 compartmentation of cAMP action in heart, 751-67 B-adrenergic receptor subtypes, 756-58 caveolae, 759-63 functional compartments of signaling, 753-56 protein kinase A and A-kinase anchoring

proteins, 763-64, 766 protein kinase C and, 764-66 pathway alcohol abuse and, 155-57 drug dependency and, 154-55 Cyclic AMP-response binding element (CREB) transcription regulation and, 484-86, 492 Cytochrome P450, 123-38 enzymes, 535-59 drug interactions and, 535-59 genetic polymorphism and, 539-41, 547-59 induction of, 535-36,

542-43, 553-59 inhibition of, 535-36, 542-43, 547-53 variability in content of, 539-41 genes CYP1B1, 297-311 CYP2B, 124-28 CYP2H, 130 CYP3A, 128-30 CYPomI sil CYP102/CYP106, 130-31 gene transcription nuclear receptors and,

123-38 induction phenobarbital and, 123-38 xenochemical metabolism and, 124 Cytochrome P450 CYP1B1, 297-311 expression human, 300-1 murine cells and tissues, 310 rodent, 302-3, 310-11

911

tumor, 297, 299, 306, 309 in human fetal tissues, 308-9

metabolic probes, 305 murine metabolism of,

304-5 normal expression in liver,

306-8 rat metabolism of, 305 regulation, 300-3

Ah receptor and, 300-3 cAMP receptor and,

302-3 structure of human gene, 298

D Daunorubicin, 64 Defluorination, 458-60 Deltanoids

chemopreventive activity of, 428-35 antiangiogenic, 421,

423, 433-35 antiproliferative, 421, 423, 433-35 proapoptotic, 421, 423,

433-35 prodifferentiating, 421,

423, 433-35 hybrid, 421, 428-35 noncalcemic, 421, 428-35

Depression role of substance P and neurokinin; receptors,

888-96 Deslurane, 455

Dexamethasone/rifampicin response element,

128-29 Diabetic nephropathy drug treatment effects on, 647 Diffuse large B-cell lymphoma, 111 la,25-Dihydroxyvitamin D3

912

SUBJECT INDEX

analogs, 421, 429-35

Disease progression definition, 625—26

description methods ANOVA/ANCOVA, 632 change from baseline, 633 linear and nonlinear,

626, 633 survival analysis, 632 drug modifications of,

627-51 Alzheimer’s disease, 625-26, 630-31, 637-39, 642-43 diabetic nephropathy, 647 osteoporosis, 648-50

Parkinson’s disease, 625-26, 629-31, 633-37, 643-45 respiratory disease, 645-47 time course effects,

629-30 models asymptotic, 626-27 linear, 626, 633

Disease status measuring, 630-32 DNA nitric oxide modifications of, 219-22 DNA cleavage camptothecin and camptothecin cytotoxicity and, 57-59 mechanisms of topoisomerase I inhibition by, 55—57 topoisomerase I-mediated, 55-64 topoisomerase II-mediated, 64-68 tumor cell death and, 53-68 See also Carcinogenesis;

Tumorigenesis DNA damage, 367-89

apoptosis and, 367, 386-89 p53 and, 388-89 receptor-mediated, 388 cell cycle checkpoints and, 367, 379-83, 389 checkpoint Rad pathway, 379-84 p53 and, 384-86 DNA repair mechanisms,

367-79, 389 base excision repair (BER) pathway, 367-75 DNA double-strand break repair, 375—79 DNA glycolases and, 367-75 formation of covalent adducts with DNA, 367 hydrolytic deamination of cytosine and 5 methylcytosine bases, 367 hydrolytic depurination, 367 nonhomologous end-joining, 378-79 single-strand annealing, 378 Dopamine amphetamine-induced release of, 242 schizophrenia and, 237—57 dopamine hypothesis of, 237-39, 252, 254 Dopaminergic stabilizers schizophrenia and, 240-41 Doxorubicon, 64 Drug development, 347-63 biomarkers and, 347-63

confounding factors, 349 drug discovery and preclinical development, 351 early clinical development, 351-52 late clinical

development, 352-53

surrogate endpoints and, 347-63 regulatory review, 347, 350-51, 358-63 Drug interactions, 535—59 cytochrome P450 enzymes and inhibition and induction of, 535—36, 542-43, 547-59 intestinal first-pass metabolism and, 53841 liver and, 536-39 quantitative prediction of, 535 Drug metabolism cytochrome P450 CYP1B1 and, 297-311 anticancer drugs and, 299, 306, 309, 311 cytochrome P450 enzymes and, 535-59 inhibition of, 535 fluorine-containing drugs and, 443-64 drug disposition and, 447 drug distribution and, 447-48 drug metabolism rate, 448-52 prevention of metabolic bioactivation, 452—58

pharmacogenomic influences on, 101-16 polymorphic, 104~7 Drug metabolizing enzymes ethnic differences in, 815-26 CYP2C subfamily, 816-20 CYP206, 820-23 CYP3A4, 823-26 Drug receptors ethnic differences in drug

SUBJECT INDEX

disposition and response, 815, 827-30 a-adrenergic receptor, 827-28 B-adrenergic receptor, 828-30 Drug response fluorine substitution and,

443 interpatient variability in, 101, 103-4 molecular basis for ethnic differences in, 815-33 drug metabolizing enzymes and, 815-26 drug receptors and, 815,

827-30 drug transporters and,

815, 826-27 nitric oxide synthase and, 830-31 pertussis toxin—sensitive G; protein, 831-32

pharmacogenetics and, 101-16 polymorphic drug metabolism and, 104-7 polymorphic drug targets and, 107-9 Drug toxicity, 443 Drug transporters ethnic differences in drug disposition and response, 815, 826-27 p-glycoprotein, 826-27 Drug treatment effects,

625-51 Alzheimer’s disease, 625-26, 630-31, 637-39, 642-43 diabetic nephropathy, 647 osteoporosis, 648-50 Parkinson’s disease, 625-26, 629-31, 633-37, 643-45 protective, 625, 627-30 respiratory disease, 645-47

symptomatic, 625, 627-29 Dysmenorrhea nonpeptide V,-vascular arginine vasopressin receptor agonists and

913

aromatase and, 583

glutamate excitotoxicity and, 582 NMDA excitotoxicity

treatment of, 175-76,

and, 582 neurotrophic effects, 569,

185, 195 oxytocin and, 185

second messenger cascade,

E Embryogenesis, 54 Endothelin pathway, 852-53 Endothelin system, 851-67

components in brain, 851, 863-65 in development, 855-56 in heart, 851, 858-60 in kidney, 851, 862-63

in lung, 851, 861-62 in vessels, 851, 856-58 Enflurane, 455-56

Enzyme inhibition and induction drug interactions and,

535-36, 542-43, 547-59 Estradiol, 569-73, 575-84 antioxidant effects, 572, 575-76 neuroprotective effects, 569, 578-84 neurotrophic effects, 569, 576-78 17a, 569-73, 575-84 178, 569-73, 575-84 Estrogen antioxidant effects, 572, 575-76, 579-82 carcinogenicity oxidative metabolism and, 454

mechanism of action,

572-16 neuroprotective effects, 569, 576-78 antiapoptotic effects, 579 astroglial cells and

576-78 569, 572-75, 584 sites of estrogen action in brain, 570-72

Estrogen receptors, 569-85 crosstalk with neurotrophins, 574-75,

583-84 ERa, 569-85 ERB, 569-85 membrane-associated, 569, 574-75 role in estrogen’s neuroprotective effects, 583-84 selective estrogen receptor modulators (SERMs) and,

583-84 limitations of, 583-84

Estrogen replacement therapy, 578-84 benefits Alzheimer’s disease, 570, 578, 582 antiapoptotic effects, oy) antioxidant effects, 572,

575-76, 579-82 astroglial cells and aromatase and, 583 cardiovascular disease,

Se) mental performance and function, 578

neuroprotection against glutamate excitotoxicity, 582 neuroprotection against NMDA excitotoxicity,

582

914

SUBJECT INDEX

osteoporosis, 579 selective estrogen receptor modulators (SERMs),

583-84 limitations of, 583-84

Estrogen response element (ERE), 572-73, 576, 583-84 Ethnicity drug disposition and response and, 815—33

drug metabolizing enzymes and, 815-26 drug receptors and, 815,

827-30 drug transporters and, 815, 826-27 nitric oxide synthase and, 830-31 pertussis toxin—sensitive G; protein and, 831-32 N-Ethylamaleimide, 65

Etoposide, 64 F Ferritin, 217 Fluorine-containing drugs,

443-64 as antimetabolites,

460-63

drug disposition and, 447 drug metabolism rate and, 448-52 metabolic bioactivation prevention of, 452-58 metabolic defluorination, 458-60

physicochemical properties, 444 Fluorouracil, 460-62 Forskolin (FSK)

adenylyl cyclase and, 146-47, 149, 151, 154, 158 Free radicals nitric oxide, 203-24

G GABA monoamine-glutamateGABA interactions,

237-S7 schizophrenia and, 237,

241, 243, 245-47 GABAag receptor general anesthesia and,

23-43 Gastrointestinal carcinogenesis proapoptotic activities of vitamin D and, 426

Gene expression downregulation of antisense oligonucleotides and, 403-14 Gene regulation cytochrome P450, 123-38 constitutive active receptor (CAR) and,

123, 131-38 phenobarbital response elements and, 123-38 Genetic polymorphisms cytochrome P450 enzyme, 539-41, 547-S9 induction, 553-59 inhibition, 547-53 ethnic differences in drug disposition and response, 815-33 Genome instability See DNA damage Glutamate monoamine-glutamateGABA interactions, 237-57 glutamate monoamine interactions at presynaptic level, 244-45 glutamatergic control of monoamine release, 241-44

schizophrenia and, 237-38 Glycogen synthase kinase-3 (GSK-3B), 797-803 Alzheimer’s disease and, 802-3 lithium action and,

797-803 G protein-coupled receptor gene family lysophospholipid receptors, 507-11 G protein-coupled receptors arginine vasopressin receptor subtypes V,-vascular (V;R),

175-78, 182-85, 189-90 V>-renal (V2R), 175-76, 178, 186-96 V3-pituitary (V3R),

175-76, 178-79, 190 classification, 594-95 polymorphisms C-terminal domain, 606-8 extracellular loop domain, 605-6 genetic variations and, 593-615 intracellular loop domain, 603-5 noncoding-region, 608-12 single nucleotide polymorphisms (SNPs), 593-615 transmembrane domain, 599-603 prostaglandins and, 661-81 truncated receptors, 612-13 variations at loci, 595—96 G protein-coupled signaling pathways compartmentation in cardiac myocytes, 751-67 B-adrenergic receptor

SUBJECT INDEX

subtypes, 756-58 cAMP action, 751-56

caveolae, 759-63 protein kinase A and A-kinase anchoring proteins, 763-64, 766

protein kinase C and, 764-66 GRAS (generally recognized as safe)

origin of concept, 13

H Haloperidol, 249-52, 254-56,

448 Halothane, 455-57 hepatotoxicity, 456-57 Hippocrates, 1 2/Histone deacytylase,

489-91 History of toxicology, 1-20 Arnt-Schultz law, 16 Calabrese, 16 certification, 8-10

Chicago Toxicology Laboratory, 2-5

GRAS (generally recognized as safe), 13

Hippocrates, 1 hormesis, 16 journals, 5-6 molecular biology, 16 Orfila, 1 Paracelsus, 1, 16 Society of Toxicology, 7-8 textbooks, 6-7 threshold limit value (TLV), 13-15 transgenic toxicology, 16 US Army Chemical Warfare Service, 2 Hormesis, 16 Human Genome Project pharmacogenetics and, 102 Hypercholesterolemia endothelial function and

L-arginine and, 79,

82-85 Hyperdopaminergia model of schizophrenia, 238,

240-41, 249-52 Hypoglutamatergia model of schizophrenia, 238,

242-45, 249-52 Hyponatremia treatment of aquaretics and, 175, 195

nonpeptide vasopressin receptor antagonists and, 175, 186, 195

I Immunity cell-mediated osteopontin and, 723,

730-31 Induction cytochrome P450 3-methylcholanthrene and, 123-24 phenobarbital (PB), 123

phenobarbital response elements and, 123-38

Inflammation adensine receptors and, 775-84 osteopontin and, 728-31

cell-mediated immunity, 723, 730-31 macrophage phagocytosis, 729-30 monocyte/macrophage differentiation, 729

monocyte/macrophage migration, 728-29

915

In silico biology, 103 Intestinal first-pass metabolism drug interactions and cytochrome P450 enzymes and, 538-39 Iron regulatory protein (IRP), 216-17 Iron response element (IRE), 216-17 Ischemia-reperfusion injury adenosine receptors and,

7715-84 tyrosine nitration and, 212 Isoflurane, 455

K Kidney toxicant injury of osteopontin and, 732-33 L Leber’s congenital amaurosis, Sy Ligand-gated ion channels anesthetics’ molecular action on, 23-32 GABAag receptors and,

23-43 molecular modeling of, 34-40 nitrous oxide, 27, 32,

42-43 xenon, 27, 32, 42-43

Lipids nitric oxide and modifications of, 222—23 Lithium developmental effects, 791

Informatics, 103, 115

hematopoiesis, 791-92

Inositol monophosphatase lithium action and, 789, 794-96 Inositol trisphosphate (IP3), 317-18, 334-35 calcium signaling and, 317-18, 334-35

metabolic effects, 791

molecular targets glycogen synthase kinase-3, 797-803

inositol monophosphatase, 789, 794-96

916

SUBJECT INDEX

phosphomonoesterase, 794-97 Wnt signaling and, 789, 798-802 neuropsychiatric effects, 789-91, 803 bipolar disorder and,

789-91, 803 Liver cytochrome P450 CYP1B1 expression in, 306-8

cytochrome P450 enzymes in metabolic drug interactions and,

536-39 Liver cirrhosis nonpeptide vasopressin receptor agonists and treatment of, 186-88 Long-term depression (LTD) nitric oxide/cyclic ADP-ribose signaling pathway and, 325-26 Long-term memory (LTM) adenylyl cyclases and, 148-50 Long-term potentiation (LTP) adenylyl cyclases and, 148-50 Lower urinary tract, 691—710 anatomy and innervation,

691-92 pharmacological modulation of function afferent neuropeptides and, 702-4 botulinum toxin and, 710 capsaicin and, 708—9 cholinergic, 693-95 dopaminergic, 698-99 GABAergic inhibitory mechanisms, 702

glutamatergic excitatory mechanisms, 700-1

glycinergic inhibitory mechanisms, 702

neurotrophic factors and,

708 nitric oxide and, 704-5

opioid peptides and, 702 prostanoids and, 705—6 purinergic, 695-98

resiniferatoxin (RTX)

choline (SPC),

524-27 cardiovascular system functions, 524-26

production and function of extracellular, 526-27 tumor cells and, 526

and, 709-10 serotonergic, 698-99 voltage-gated ion channels and, 706-8 voiding dysfunction, 692-93 neural control of, 692 Lysophosphatidic acid (LPA) See Lysophospholipid

receptors; Lysophospholipids Lysophospholipid receptors, 507-27 G protein-coupled receptor gene family and, 507-11

lysophospholipidic acid receptors, 511-16 sphingosine 1-phosphate receptors, 521—24 sphingosylphosphorylcholine receptor, 524 Lysophospholipids, 507-27 lysophosphatidic acid (LPA), 507-24, 527 nervous system function, 517-19 production and degradation of extracellular, 520-21 tumor cells and, 519-20 sphingosine 1-phosphate (SIP), 507-11, 524-27 cardiovascular system functions, 524-26 production and function of extracellular, 526-27 tumor cells and, 526 sphingosylphosphoryl-

M m-AMSA, 64 Mast cells, 775, 781-82 Metabolism, drug See Drug metabolism Metastasis, 723 Methoxyflurane, 454-56 Minimum alveolar concentration, 25 Mitoxanthrone, 64

MK-801, 242-45, 249-56 Molecular modeling anesthetics’ molecular action on ligand-gated ion channels, 3440 Monoamines glutamate-monoamine interactions at postsynaptic level, 244-45 glutamatergic control of release of, 241-44 monoamine-glutamateGABA interactions, 237-57 schizophrenia and, 237 Movement disorders, 237 Myocardial inflammation tyrosine nitration and, 212 Myocyte enhancer factor-2 (MEF2), 489-91

N Nephrotic syndrome nonpeptide vasopressin receptor and treatment of, 175 N-ethylamaleimide, 65 Neurokinin; receptor

SUBJECT INDEX

antagonists, 877-96 MK-869, 877, 893-95 preclinical evaluation, 888-92 substance P and, 877—96

antidepressant activity, 877-96 monoamine-substance P interactions, 881—83

transgenic animals, 886-87 Neurotrophic facotors pharmacological modulation of lower urinary tract function and, 708 Nicotinic acid adenine dinucleotide phosphate (NAADP), 317-35 active anlogs of, 322 calcium signaling and, 317-35 in oocytes, 327-28 in pancreatic acinar cells, 328-29 Nitric oxide (NO) L-arginine and, 79-88 autooxidation of, 205-6 calcium signaling mediated by, 325-26 chemistry of, 204—6 cyclic ADP-ribose pathway and, 325 DNA modifications from, 219-22 effect on heme proteins, 214-16 inflammatory disease pathology and, 204 interactions with guanylate cyclase, 203 lipid modifications from, 222-23 nitration and, 203, 205, 209-14 nitrosation and, 205 S-nitrosylation and, 203, — 206-9

oxidation and, 205

p53 and, 218 pharmacological modulation of lower urinary tract function and, 704-5 prostaglandin H synthase and, 219 protein radicals and,

218-19 reactions with metal complexes, 204 ribonucleotide reductases and, 219

vascular physiology and pathophysiology and, 81 vasodilation and, 81, 83-88 zinc-finger-containing nuclear receptors and, 218 Nitric oxide synase (NOS) L-arginine and, 79-88 endothelial ethnic differences in drug disposition and response and, 830-31 isoforms, 203-4

Nitration manganese superoxide dismutase, 213 nitric oxide and, 203, 205,

209-14 tyrosine, 209-14 athersclerotic plaques and, 212 ischemia-reperfusion injury and, 212 myocardial inflammation and, 212

renal hypertension and, 212 Nitrosation, 205 Nitrosylation nitric oxide and, 203, 206-9 Nitrotyrosine, 209-14 Nitrous oxide effects on ligand-gated ion

917

channels, 27, 32, 42-43

NMDA, 569, 576, 579, 582 excitotoxicity estrogen’s neuroprotective effects against, 582 oxidative stress and, 579,

582 Nonpeptide arginine vasopressin (AVP) receptor antagonists, 175-76, 182-90, 192-96 See also Arginine vasopressin receptor antagonists Noradrenaline schizophrenia and, 237-38,

241, 256-57 O Opioid peptides pharmacological modulation of lower urinary tract function and,

702 Orfila, 1

Organophosphate insecticides, 3

Osteopontin (OPN), 723-41 bone remodeling and, 733-36 cellular signaling and, 723-28 inflammatory functions of, 728-31 cell-mediated immunity,

723, 730-31 macrophage phagocytosis, 729-30 monocyte/macrophage migration, 728-29

receptors, 724-28

CD44v, 725-26 integrins, 724—26 toxicant injury and, 723,

732-33 kidney and, 732-33

918

SUBJECT INDEX

tumorigenesis and, 736—40 vasculature and, 731—32 Osteoporosis drug treatment effects on, 625, 648-50 Oxidant injury thioredoxin and protection against, 269-71 Oxidative damage estrogen’s protective effect against, 569, 572,

575-76, 579-82 NMDA and, 569, 576, 579, 582 Oxytocin, 176, 185 Oxytocin receptor, 175-76, 179, 191-92 P p53 nitric oxide crosstalk with, 218 thioredoxin and binding of, 272-73 p54 DNA damage and repair and apoptosis and, 388—89 cell cycle checkpoints and, 384-86 Paracelsus, 1, 16 Parkinson’s disease drug treatment effects, 625-26, 629-31, 643-45 angiotensin-converting enzyme (ACE), 629 selegilene and tocopherol, 629 estrogen’s protective effect against, 570, 578, 582 Peroxisome proliferator activated receptor @, 137 Peroxynitrite, 206 Pertussis toxin—sensitive G;

protein, 831-32 Pharmacogenetics, 101-16 current technology, 102-3

drug development and, 113-14 drug metabolism and, 104-7 ethics and, 115-16 gene expression and, 111-13 high-density, 110-11 comparative genomic hybridization (CGH), 110-11 polymorphism drug metabolism, 104—7 drug targets, 107-9 single nucleotide

polymorphism (SNP), 108-10, 113-14 population, 815-33 thiopurine methyltransferase and (TPMT), 104-7, 113-14 Phenobarbital (PB) mechanism of induction, 131-35 constitutive active receptor (CAR) and, 123, 131-35 phenobarbitalresponsive enhancer molecule (PBREM)

and, 123, 126-27, 131 pregnane X receptor (PXR) and, 123, 129-32 xenobiotic responsive module and, 123, 128-29 Phenobarbital (PB) response elements, 123-38 cytochrome P450 and genes, 123-38 induction of, 123-38 xenobiotic responsive module (XREM), 123,

128-31 Phenobarbital-responsive enhancer molecule (PBREM), 123, 126, ISI SI1S7

Phosphomonophosphatase lithium action and, 789, 794-97 Phosphorothioate oligodeoxynucleotides, 403-14 Polymorphisms pharmacogenetics and polymorphic drug metabolism, 104-7 polymorphic drug targets, 107-9 single nucleotide polymorphism (SNP), 108-10, 113-14 Preconditioning adenosine receptors and, 775-81 Pregnane X receptor (PXR), 123, 129-32, 134-35 activation of, 134-35 phenobarbital induction and, 131-32, 134-35 Pregnane X receptor response element, 130 Prostaglandin H synthase interaction with nitric oxide, 219 Prostaglandins, 661-81 Prostanoid receptors, 661-81 DP receptors, 680-81 FP receptor, 670-71 IP receptors, 679-80 multiple E-prostanoid receptors, 671-79 TxA (TP) receptor, 668-70 Prostate cancer proapoptotic activities of vitamin D and, 425 Protein binding

thioredoxin and, 273-75 Protein folding thioredoxin and, 277

Protein kinase A compartmentation of G protein-coupled signaling

SUBJECT INDEX

pathways and, 751-67 Protein radicals nitric oxide reactions with,

218 Prozazine, 244

Psychogenesis hyperdopaminergia model of, 238, 240-41, 249-52 hypoglutamatergia model of, 238, 242-45, 249-52 monoamine-glutamaie-

Reperfusion injury protection against thioredoxin and, 276-77

Resiniferatoxin (RTX)

pharmacological modulation of lower urinary tract function and, 709-10 Respiratory disease drug treatment effects on,

645-47

GABA interactions and,

Retinoid X receptor (RXR),

237-S7

126-27 Ribonucleotide reductases nitric oxide interactions with, 219 RNase H, 403-4, 408-10 irrelevant cleavage and, 408-10 Ryanodine receptor (RyR), 320-35 cyclic ADP-ribose modulation, 322-23

Q Quinolone, 448

R Raloxifene, 583 Raltitrexed, 110 Raynaud’s phenomenon nitric oxide deficiency and, 85 Reactive nitrogen species (RNS), 203-4, 218 tumor suppressor p53 and, 218 Reactive oxygen species (ROS) DNA glycolase and,

370-73 Recombinant receptors anesthetic actions on,

23-27 Regulatory review biomarkers and, 347, 350-51, 358-63 investigational new drugs, 351 legal basis, 358 new drug applications, 350 Renal hypertension tyrosine nitration and, 212 Renal salt handling endothelin and, 851, 862-63

S Schizophrenia, 237-57

dopamine and, 237-57 dopamine hypothesis of, 237-39, 252, 254 hyperdopaminergic model of, 238, 240-41, 249-52 hypoglutaminergic model of, 238, 242-45, 249-52 monoamine-glutamateGABA interactions and, 237-57 noradrenaline and, 237—38, 241, 256-57 serotonin and, 237-38, 241-42, 246, 251, 254, 256-57

5‘UTR polymorphism, 609-10 Second messengers estrogen and G protein-linked, 569, 572-75, 584

919

Selective estrogen receptor modulators (SERMs),

583-84 limitations of, 583-84

Selective serotonin uptake inhibitors, 448 Selenium effects on thioredoxin peroxidase, 271

Serotonin schizophrenia and, 237-38,

241-42, 246, 251, 254, 256-57 Serum response factor (SRF),

487-89 Sevoflurane, 455

Signal transduction adenylyl cyclases and, 145-62 cAMP signaling pathway and, 147, 155-62 LTM and, 149 LTP and, 148-50 arginine vasopressin (AVP)

receptor subtypes and, 177-79 oxytocin receptor and, 179 Simvastatin, 352-53

Single nucleotide polymorphisms (SNPs), 108-10, 113-14, 593-615 G protein-coupled receptors and, 593-615 Skin cancer vitamin D and chemoprevention of,

423-26 Society of Toxicology, 7-8 Sphingosine 1-phosphate (SIP), 507-11, 524-27 cardiovascular system functions, 524-26

production and function of extracellular, 526-27

tumor cells and, 526

920

SUBJECT INDEX

Sphinosylphosphorylcholine (SPC), 524-27 cardiovascular system functions, 524-26

production and function of extracellular, 526—27

tumor cells and, 526 Stabilizers See Dopaminergic stabilizers Substance P antidepressant activity of, 877-96 monoamine-substance P interactions, 881—83 depression and, 892-96 substance P systems effects of antidepressant drugs on, 887

stress and, 885-86 SUMO, 53, 63 Superoxide radicals, 203

Surrogate endpoints, 347-63 biomarkers and, 347-63 as confirmatory evidence, 359 validation of biomarkers as, 356-58 Syndrome of inappropriate secretion of antidiurectic hormone (SIADH), 175-76, 186

T Tamoxifen, 583

Thioguanine, 104—5 Thiopurine methyltransferase (TPMT), 104-7, 113-14 deficiency in drug response and, 105-7 pharmacogenetics and, 104-7, 113-14 Thioredoxins (Trx), 261-82

Alzheimer’s disease and, 279 as antioxidant, 269-71

apoptosis inhibition and, 2, 20S atherosclerosis and, 278

cancer and, 279-81 cancer drug resistance and, 276 cloned forms, 262-63 as cofactor, 270

drugs that inhibit, 280-81 as growth factor, 268-69 HIV and, 279 immune function and, 278-79 interferon-gamma growth arrest and, 277 mutant forms, 262 peroxidase, 271

in plasma, 278 processed forms, 264 protein binding and, 273-75 protein folding and, 277 redox biochemistry, 262 reductase, 261-62

regulation of expression, 267 reperfusion injury protection and, 276-77 selenium’s effect on, 271 skin damage and, 278 structure, 264-67 subcellular localization, 268 transcription factor regulation and, 270-73 transgenic mice, 277 Threshold limit value origin of concept, 13-15 Thromboxane, 668-70 Topoisomerase, 53-68

anticancer drugs and, 54-68 topoisomerase I-targeting, 53, 64-68 topoisomerase

II-targeting, 63, 64-68 DNA damage and, 55-68

topoisomerase I-mediated, 55—64 topoisomerase II-mediated, 64-68 tumor cell death and, 63-68 Transcription calcium signaling cascades, 471-93 calmodulin kinases and, 471-93 regulation, 484-93 activating transcription factor-1 (ATF-1) and,

487-88 CAAT-enhancer binding protein (c/EBP) and, 487-88 cAMP-responsive binding element (CREB) and, 484-86, 492 2/histone deacetylase and, 489-91 myocyte enhancer factor-2 (MEF2) and, 489-9] serum response factor (SRF) and, 487-89 Transcription factor regulation thioredoxin and, 270-73 Tucaresol, 352 Tumor cell death, 53-68 anticancer drugs and, 54-68 topoisomerase I-targeting, 53, 6468 topoisomerase Il-targeting, 63, 64-68 topoisomerase-mediated DNA cleavage and, 53, 55-64 Tumor cells cytochrome P450 CYP1B1 expression in, 297, 299, 306, 309, 311

SUBJECT INDEX

as early-stage tumor marker, 309, 311

lysophosphatidic acid (LPA) and, 519-20 sphingosine 1-phosphate and, 526 sphingosylphosphory]choline and,

Vitamin D chemopreventive activity of, 421-35 antiangiogenic, 421, 423, 433-35

prostate cancer, 425

Ww Wat signaling lithium and, 789, 798-802

antiproliferative, 421,

423, 433-35 proapoptotic, 421, 423,

524 Tumorigenesis osteopontin and, 736-40 thioredoxin and, 270-71 See also Carcinogenesis; DNA damage

433-35 prodifferentiating, 421, 423, 433-35 dose-limiting toxicity, 427-28 altered calcium

U Ubiquitin topoisomerase I downregulation and, 61-62 Urea cycle L-arginine and, 80 US Army Chemical Warfare

preclinical evidence of chemopreventive efficiency, 425-28 cheek pouch carcinogenesis, 426 colon cancer, 425-26 gastrointestinal

metabolism, 427-28

Service, 2

921

carcinogenesis, 426 skin carcinogenesis,

x Xenobiotic-response elements CYPI1B1 gene expression and, 300

Xenobiotic responsive module (XREM), 123, 128-31, 137 CYP3A gene induction and, 128-29

Xenochemical metabolism cytochrome P450 and, 124 Xenochemical response elements, 123

Xenon effects on ligand-gated ion channels, 27, 32, 42-43

426

Vv Vasopressin receptors and antagonists See Arginine vasopressin receptors; Arginine vasopressin antagonists

xenograft models,

427-28 sun exposure and cancer rates, 423-25

colorectal cancer, 424-25

Z Zafirlukast, 351-52

Zinc-finger-containing nuclear receptors effect of nitric oxide on, 21

he

ii

CUMULATIVE INDEXES CONTRIBUTING AUTHORS, VOLUMES 37-41 Acosta D Jr, 38:63—96 Allen JW, 39:151-73

Catterall WA, 37:361—96

Chan PLS, 41:625-—59

Flexner C, 40:651—76 Fu H, 40:619-49

Amara SG, 39:431—-56 Ambudkar SV, 39:361—97 Anders MW, 38:501-37 Anderson SP, 40:491-518

Changeux J-P, 40:431-58

Fukushima N, 41:507—34

Chaudhuri G, 37:477-515 Choudhuri S, 39:267—94 Chun J, 41:507-34

Giachelli CM, 41:723-49

Aposhian HV, 37:397-419 Atkinson AJ Jr, 41:347—66

Clapham DE, 37:167—203 Coles P, 41:175—202 Collins MD, 39:399-430

Conn PJ, 37:205—37 Contos JJA, 41:507-34 Corringer P-J, 40:431-58 Corton JC, 40:491-518 Costa E, 38:321-S0

Goldstein A, 37:1—28

Bagdassarian CK, 41:661—90 Baker RC, 39:127-50 Bakhle YS, 38:97—120 Balboa MA, 39:175-89 Balsinde J, 39:175—-89

Costa LG, 38:21-43

Guyton KZ, 41:421-42

Dalton TP, 39:67-101

Halmes NC, 37:91-117 Hammond HK, 39:343-60 Hanoune J, 41:145—74

Aschner M, 39:151—73

Giacomini KM, 38:431-60 Gibbs JB, 37:143-66 Gillette JR, 40:19-41 Glass M, 38:179-—200 Gottesman MM, 39:361-97 Greenlee WF, 41:297-316

Gu Y-Z, 40:519-61 Guengerich FP, 39:1—17

Benovic JL, 38:289-319

Bertaccini E, 41:23-51 Blackburn TP, 40:319-34 Blaschke TF, 37:451—75

Blau HM, 40:295-317 Bode-Béger SM, 41:79-99 Boger RH, 41:79-99 Borges K, 39:221-41

Borjigin J, 39:53-65 Botting RM, 38:97-120

Bradfield CA, 40:519-61 Branchek TA, 40:319-34

Brett CM, 38:43 1-60 Breyer MD, 41:661—90 Breyer RM, 41:661—90 Briggs JM, 37:71-90 Broder S, 40:97-132 Brown JH, 40:459-89 Brunton LL, 41:751-73 Burgen ASV, 40:1—16 Burke MD, 41:297-316 Carlsson A, 41:237—-60 Carlsson ML, 41:237-60

Davila JC, 38:63—-96 Davis KL, 41:203-36 Debouck C, 40:193-—208 Defer N, 41:145-74 de Groat WC, 41:691-721 Dekant W, 38:501-—37

Denhardt DT, 41:723-49 Dennis EA, 39:175-89 De Vries L, 40:235-71

Harris RA, 41:23-51 Hefti F, 37:239-67 Heinrich M, 38:539-65 Hickson ID, 41:367-401 Hobbs AJ, 39:191—220 Hockerman GH, 37:361-96

Dey S, 39:361-97

Hoffman AR, 38:45-61 Hogenesch JB, 40:519-61

Dingledine R, 39:221-41 Doull J, 41:1-21

Holford NHG, 40:209-34; 41:625-59

Dunham EW, 37:53-69

Holm-Waters S, 41:237—-60

Elenko E, 40:235-71

Hook SS, 41:471-505 Hosokawa M, 38:257-88

Elliott JD, 40:177-91 Evans WE, 41:101-—21

Houghten RA, 40:273-82 Hrycyna CA, 39:361—97

Farquhar MG, 40:235-71 ‘ Felder CC, 38:179-200 Fischer T, 40:235-71 Fisher JW, 38:1—20

Insel PA, 39:175-89, 343-60; 41:593-624 Ishii I, 41:507-34 Ito K, 38:461—99

923

924

CUMULATIVE INDEXES

Iwatsubo T, 38:461—99

McEwen BS, 41:569-91 McLeod HL, 41:101-21

Joad JP, 37:29-52

Means AR, 41:471—505

Johnson BD, 37:361—96

Melchert RB, 38:63—96 Melvin WT, 41:297-316

Johnson DG, 39:295-—312

Metcalf B, 40:193—208

Kanamitsu S, 38:461—99 Kastrissios H, 37:451—75 Kedzierski RM, 41:851—76 Kensler TW, 41:421—42 Kim RB, 41:815—50

Meyer UA, 37:269-96 Miller RJ, 38:201—27 Moncada S, 39:191—220 Monks TJ, 38:229-55

Kimelberg HK, 39:151—73 Kimko HC, 40:209-34

Montfort WR, 41:261—95 Murad F, 41:203—36

Kitteringham NR, 41:443-70 Klaassen CD, 39:267—94

Murray GI, 41:297-316 Myers SA, 41:661—90

Klein PS, 41:789-813 Kobilka BK, 38:351—73

Myers SJ, 39:22141

Kramer RE, 39:127—50

Nagata K, 40:159-76 Nakajima Y, 38:461—99 Nathan L, 37:477-515 Neer EJ, 37:167—203 Negishi M, 41:123-43 Nemeroff CB, 41:877—-906 Nilsson M, 41:237-60 Norbury CJ, 41:367-401 North RA, 40:563—80

Krupnick JG, 38:289-319 Lau SS, 38:229-55 Law P-Y, 40:389-430 Lebedeva I, 41:403-19 Lee HC, 41:317-45 Lee SJ, 41:569-91

Lefer DJ, 40:283-94

Monteleone JPR, 40:209-34

Lemasters JJ, 37:327-38 Le Novére N, 40:431-58

Lesko LJ, 41:347—66

Li T-K, 41:53-77 Li X, 39:53-65 Eines 35—617 Linden J, 41:775—-87 Lipton SA, 38:159-77

Liu J, 39:267-94 Liu LF, 41:53-77 Loh HH, 40:389-430 LoPachin RM, 39:151—73

Lu AYH, 41:535-67 Maines MD, 37:517—54 Mao GE, 39:399-430 Marcus R, 38:45-61 Marrone TJ, 37:71—90

Martin E, 41:203-36 Masters SC, 40:619-49 McCammon JA, 37:71-90

Ohlstein EH, 40:177-91 Oliff A, 37:143—-66 O’ Neill PM, 41:443-70 Ortiz de Montellano BR,

38:539-65 Otterness DM, 39:19-52 Owens MJ, 41:877—906 Ozawa CR, 40:295-317

Park BK, 41:443-70 Pastan I, 39:361—97 Peck CC, 40:209-34 Peterson BZ, 37:361—96 Phiel CJ, 41:789-813 Pin J-P, 37:205=37 Pinkerton KE, 37:29-52 Plaa GL, 40:43-65

Pratt WB, 37:297-326 Puga A, 39:67-101 Pumford NR, 37:91—117 Ramachandra M, 39:361—97 Ramos KS, 39:243—65 Rana BK, 41:593-624

Rittling SR, 41:723-49 Robles M, 38:539-65 Rodan GA, 38:375-88 Rodriguez E, 38:539-65 Rodriguez RJ, 38:63—96 Rohrer DK, 38:351—73 Ruffolo RR Jr, 40:177-91

Safe SH, 38:121-58 Sagi SA, 40:459-89 Sah VP, 40:459-89 Satoh T, 38:257-88 Seal RP, 39:431-S6 Seasholtz TM, 40:459-89 Sheiner L, 40:67—96 Shertzer HG, 39:67-101 Shiina T, 41:593-624 Shoham M, 41:175—202 Sibley DR, 39:31341 Snyder SH, 39:53-65 Springer ML, 40:295-317 Starkov AA, 40:353-88 Stauber A, 40:491-518 Steimer J-L, 40:67—96 Stein CA, 41:403-19 Stein CM, 41:815—50 Steinberg SF, 41:751-73 Stout SC, 41:877-906 Strassburg CP, 40:581-618 Streit WJ, 39:151-73 Strosberg AD, 37:421-—50 Stuehr DJ, 37:339-59 Subramanian RR, 40:619-49 Sueyoshi T, 41:123-43 Sugiyama Y, 38:461—99 Surprenant A, 40:563-80 Szumlanski CL, 39:19-52

Posner GH, 41:421-42 Post SR, 39:343-60

Powis G, 41:261-95

Tedroff J, 41:237-60 Thibonnier A, 41:175—202

CUMULATIVE INDEXES

Thibonnier M, 41:175—202 Thummel KE, 38:389-430

Thurman RG, 37:327-38 Trudell JR, 41:23-51

Wallace KB, 40:353-88 Waring JF, 40:335-52 Waters N, 41:237—60

925

Wood AJJ, 41:815—-50

Xie H-G, 41:815-50

Tukey RH, 40:581-618

Wei L-N, 37:119-41 Weiner JA, 41:507-34

Turko IV, 41:203—36

Weinshilboum RM,

39:19-52

Yanagisawa M, 41:851-76

Ulrich RG, 40:335-52

West JE, 38:539-65

Yoshimura N, 41:691-—721

Yamakura T, 41:23-51 Yamazoe Y, 40:159-—76

White RE, 40:133-57 Vane JR, 38:97—120 Venter JC, 40:97-132 Walker CL, 39:295-312

Whitlock JP Jr, 39:103-—25

Wilkinson GR, 38:389-430

Witschi H, 37:29-52 Wong YH, 40:389-430

Zanger UM, 37:269-96 Zhang L, 38:43 1-60 Zheng B, 40:235-71 Zimmerman BG, 37:53-69

CHAPTER TITLES, VOLUMES 37-41

Prefatory Pharmacology A Rewarding Research Pathway A Quest for Erythropoietin Over Nine Decades Targets of Drug Action High-Throughput Screening in Drug Metabolism and Pharmacokinetic Support of Drug Discovery

A Burgen

37:1-28 38:1-20 40:1-16

RE White

40:133-57

JR Gillette

40:19-41

GL Plaa

40:43-65

JC Corton, SP Anderson, A Stauber J Doull

40:491-S18

A Goldstein JW Fisher

Toxicology Laboratory of Chemical Pharmacology, National Heart, Lung, and Blood Institute, NIH: A Short History Chlorinated Methanes and Liver Injury: Highlights of the Past 50 Years Central Role of Peroxisome Proliferator-Activated Receptors in the Actions of Peroxisome Proliferators

Toxicology Comes of Age

41:1-21

General Topics in Pharmacology and Toxicology Receptors Pharmacology and Functions of Metabotropic Glutamate Receptors The Role of the hsp90-Based Chaperone System in Signal Transduction by Nuclear Receptors and Receptors Signaling via Map Kinase Structure and Function of the 63-Adrenergic Receptor Cannabinoid Receptors and Their Endogenous Agonists Presynaptic Receptors From GABAA Receptor Diversity Emerges A Unified Vision of GABAergic Inhibition Insights from In Vivo Modification of Adrenergic Receptor Gene Expression

926

PJ Conn, J-P Pin

37:205—37

WB Pratt

37:297-326

AD Strosberg

37:421-S0

CC Felder, M Glass

RJ Miller

38:179-200 38:201-27

E Costa

38:321-S0

DK Rohrer, BK Kobilka

38:35 1-73

CHAPTER TITLES

Genetic Regulation of Glutamate Receptor Ion Channels New Insights into Dopaminergic Receptor Function Using Antisense and Genetically Altered Animals 5-HT¢ Receptors as Emerging Targets for Drug Discovery

Nicotinic Receptors at the Amino Acid Level

927

SJ Myers, R Dingledine, K Borges

39:221-41

DR Sibley

39:313-41

TA Branchek, TP Blackburn P-J Corringer, N Le Novére,

40:3 19-34 40:431-58

J-P Changeux Pharmacology of Clonded P2X Receptors Lysophospholipid Receptors

RA North, A Surprenant N Fukushima, I Ishii, JJ Contos, JA Weiner,

40:563-80 41:507-34

J Chun

Genetic Variations and Polymorphisms of G Protein-Coupled Receptors: Functional and Therapeutic Implications

BK Rana, T Shiina,

41:593-624

PA Insel

Prostanoid Receptors: Subtypes and Signaling

Role of Osteopontin in Cellular Signaling and Toxicant Injury

RM Breyer, CK Bagdassarian, SA Myers, MD Breyer

41:661-90

DT Denhardt,

41:723-49

CM Giachelli,

SR Rittling

Molecular Approach to Adenosine Receptors: Receptor-Mediated Mechanisms of Tissue Protection

J Linden

41:775-87

WC de Groat, N Yoshimura

41:691-721

HE Lee

41:317-45

GH Hockerman, BZ Peterson, BD Johnson, WA Catterall

37:361-96

LG Costa

38:21-43

Renal System Pharmacology of the Lower Urinary Tract

Signal Transduction Physiological Functions of Cyclic ADP-Ribose and NAADP as Calcium Messengers

Synaptic Functions Molecular Determinants of Drug Binding and Action on L-Type Calcium Channels

Signal Transduction in Environmental Neurotoxicity

928

CHAPTER TITLES

Inhibition of Nitric Oxide Synthase as a Potential Therapeutic Target Redox Regulation of c-Ha-ras and Osteopontin Signaling in Vascular Smooth Muscle Cells: Implications in Chemical Atherogenesis Cyclins and Cell Cycle Checkpoints The Regulator of G Protein Signaling Family

AJ Hobbs, A Higgs, S Moncada

39:191—220

KS Ramos DG Johnson, CL Walker

39:243-65 39:295-312 40:235-71

L De Vnies, B Zheng, T Fischer, E Elenko,

MG Farquhar Pharmacology of Selectin Inhibitors in Ischemia/Reperfusion States The Role of Rho in G Protein Coupled Receptor Signal Transduction

DJ Lefer

40:283-94

VP Sah, TM Seasholtz, SA Sagi, JH Brown

40:459-89

H Fu, RR Subramanian,

40:619-49

14-3-3 Proteins: Structure, Function, and

Regulations

SC Masters

Transporters Compartmentation of G Protein-Coupled Signaling Pathways in Cardiac Myocytes

SF Steinberg, LL Brunton LL

41:751-73

MD Maines

37:517-54

T Satoh,

38:257-88

Enzymes The Heme Oxygenase System: A Regulator of Second-Messenger Gases The Mammalian Carboxylesterases: From Molecules to Functions The Role of Receptor Kinases and Arrestins in G Protein-Coupled Receptor Regulation

M Hosokawa

JG Krupnick,

38:289-319

JL Benovic

Methylation Pharmacogenetics: Catechol O-Methyltransferase, Thiopurine Methyltransferase, and Histamine N-Methyltransferase

RM Weinshilboum, DM Otterness,

39:19-52

CL Szumlanski

Regulation and Inhibition of Phospholipase A>

J Balsinde, MA Balboa,

39:175-89

PA Insel, EA Dennis

Human UDP-Glucuronosyltransferases: Metabolism, Expression, and Disease

RH Tukey, CP Strassburg

40:581-618

CHAPTER TITLES

Tumor Cell Death Induced by Topoisomerase-Targeting Drugs Phenobarbital Response Elements of Cytochrome P450 Genes and Nuclear Receptors Regulation and Role of Adenylyl Cyclase Isoforms

929

T-K Li, LF Liu

41:53-77

T Sueyoshi,

41:123-43

M Negishi

J Hanoune, N Defer

41:145-74

NR Pumford, NC Halmes F Hefti

37:91-117 37:167—203 37:239-67

DJ Stuehr

37:339-59

HV Aposhian

37:397-419

TJ Monks, SS Lau

38:229-55

M Heinrich, M Robles,

38:539-65

Chemical Agents Protein Targets of Xenobiotic Reactive Intermediates G Protein Beta-Gamma Subunits Pharmacology of Neurotrophic Factors Structure-Function Aspects in the Nitric Oxide Synthases Enzymatic Methylation of Arsenic Species and Other New Approaches to Arsenic Toxicity The Pharmacology and Toxicology of Polyphenolic-Glutathione Conjugates Ethnopharmacology of Mexican Asteraceae (Compositae)

DE Clapham, EJ Neer

JE West,

BR Ortiz de Montellano,

E Rodriguez The Pineal Gland and Melatonin: Molecular

and Pharmacologic Regulation

Regulation of Gene Expression by Reactive Oxygen

Cytotoxicity of Short-Chain Alcohols Metallothionein: An Intracellular Protein to Protect Against Cadmium Toxicity Teratology of Retinoids The Clinical Pharmacology of L-Arginine The Basic and Clinical Pharmacology of Nonpeptide Vasopressin Receptor Antagonists

Novel Effects of Nitric Oxide

J Borjigin, X Li, SH Snyder

39:53-65

TP Dalton, HG Shertzer,

39:67-101

A Puga RC Baker, RE Kramer

39:127-5S0

CD Klaassen, J Liu, S Choudhuri MD Collins, GE Mao

RH Boger, SM Bode-Boger

M Thibonnier, P Coles, A Thibonnier, M Shoham KL Davis, E Martin, IV Turko, F Murad

39:267-94 39:399-430 41:79-99

41:175—202 41:203-36

930

CHAPTER TITLES

Biotransformation Molecular Mechanisms of Genetic Polymorphisms of Drug Metabolism In Vitro and In Vivo Drug Interactions Involving Human CYP3A Glutathione-Dependent Bioactivation of Haloalkenes Cytochrome P-450 3A4: Regulation and Role in Drug Metabolism Induction of Cytochrome P4501A1 Metabolism of Fluorine-Containing Drugs

Interindividual Variability in Inhibition and Induction of Cytochrome P450 Enzymes

UA Meyer, UM Zanger

37:269-96

KE Thummel, GR Wilkinson

38:389-430

MW Anders,

38:501-37

W Dekant

FP Guengerich JP Whitlock Jr BK Park, NR Kitteringham, PM O'Neill

39:1-17 39:103—25 41:443-70

JH Lin, AYH Lu

41:535-67

CJ Norbury,

41:367401

Nucleic Acids Cellular Responses to DNA Damage

ID Hickson

Ca*+/CaM-Dependent Kinases: From Activation to Function

SS Hook, AR Means

41:471-505

L Zhang, CM Brett,

38:43 1-60

Pharmacokinetics/Toxicokinetics Role of Organic Cation Transporters in Drug Absorption and Elimination

KM Giacomini

Biochemical, Cellular, and Pharmacological Aspects of the Multidrug Transporter

Mitochondrial Targets of Drug Toxicity

SV Ambudkar, S Dey, CA Hrycyna, M Ramachandra, I Pastan, MM Gottesman KB Wallace, AA Starkov

40:353-88

JB Gibbs, A Oliff

37:143-66

SH Safe

38:121-S8

G Powis, WR Montfort

41:261-95

39:361-97

Cancer and Carcinogenesis The Potential of Farnesyltransferase Inhibitors as Cancer Chemotherapeutics Interactions Between Hormones and Chemicals in Breast Cancer Properties and Biological Activities of Thioredoxins Cancer Chemoprevention Using Natural Vitamin D and Synthetic Analogs

KZ Guyton, TW Kensler GH Posner

r)

41:421-42

CHAPTER TITLES

931

Clinical Therapeutics Medication Compliance as a Feature in Drug Development Dual Protease Inhibitor Therapy in HIV-Infected Patients: Pharmacologic Rationale and Clinical Benefits Pharmacogenomics: Unlocking the Human Genome for Better Drug Therapy Antisense Oligonucleotides: Promise and Reality

H Kastrissios, TF Blaschke

37:451-75

C Flexner

40:65 1-76

HL McLeod, WE Evans

41:101-21

I Lebedeva, CA Stein

41:403-19

TJ Marrone, JM Briggs, JA McCammon

37:71-90

RA Houghten

40:273-82

CR Ozawa, ML Springer, HM Blau

40:295-317

L Lesko, AJ Atkinson Jr.

41:347-66

JR Vane, YS Bakhle, RM Botting

38:97-120

M Aschner, JW Allen,

39:151-73

Drug Development Science Structure-Based Drug Design: Computational Advances Parallel Array and Mixture-Based Synthetic Combinatorial Chemistry: Tools for the Next Millennium A Novel Means of Dmg Delivery: Myoblast-Mediated Gene Therapy and Regulatable Retroviral Vectors

Use of Biomarkers and Surrogate Endpoints in Drug Development and Regulatory Decision Making: Criteria, Validation,

Strategies

Systems Immune System/Inflammation Cyclooxygenases | and 2

Central Nervous System Glial Cells in Neurotoxicity Development

HK Kimelberg, RM LoPachin,

WJ Streit

Excitatory Amino Acid Transporters: A Family in Flux Molecular Mechanisms and Regulation of Opiod Receptor Signaling

RP Seal, SG Amara

39:43 1-56

P-Y Law, YH Wong,

40:389-430

HH Loh

932

CHAPTER TITLES

Anesthetics and Ion Channels:

Molecular Models and Sites of Action

T Yamakura,

41:23-51

E Bertaccini, JR Trudell,

RA Harris Interactions Between Monoamines,

Glutamate, and GABA in Schizophrenia: New Evidence

A Carlsson, N Waters,

41:237-60

S Holm-Waters, J Tedroff, M Nilsson, ML Carlsson

Drug Treatment Effects on Disease Progression Molecular Targets of Lithium Action Neurokinin! Receptor Antagonists as Potential Antidepressants

P Chan, N Holford

CJ Phiel, PS Klein

41:625-59 41:789-813

SC Stout, MJ Owens, CB Nemeroff

41:877-906

SR Post, HK Hammond, PA Insel

39:343-60

BK Rana, T Shiina,

41:593-624

Autonomic Nervous System B-Adrenergic Receptors and Receptor Signaling in Heart Failure Genetic Variations and Polymorphisms of G Protein-Coupled Receptors: Functional and Therapeutic Implications

PA Insel

Cardiovascular System Tissue Renin-Angiotensin Systems: A Site of Drug Action?

BG Zimmerman,

37:53-69

EW Dunham

Reperfusion Injury After Liver Preservation for Transplantation

JJ Lemasters,

37:327-38

RG Thurman

Endothelin System: The Double-Edged Sword in Health and Disease

RM Kedzierski, M Yanagisawa

41:851-76

L Nathan, G Chaudhuri

37:477-S15

R Marcus, AR Hoffman GA Rodan

38:45-61 38:375—88

SJ Lee, BS McEwen

41:569-91

Endocrine System Estrogens and Atherosclerosis Growth Hormone As Therapy for Older Men and Women Mechanism of Action of Biophosphates Neurotrophic and Neuroprotective Actions of Estrogens and Their Therapeutic Implications

CHAPTER TITLES

933

Pulmonary System The Toxicology of Environmental Tobacco Smoke

H Witschi, JP Joad,

37:29-52

KE Pinkerton

Microbial Systems Neuronal Injury Associated with HIV-1: Approaches to Treatment

SA Lipton

38:159-77

L-N Wei

37:119-41

JC Davila, RJ Rodriguez, RB Melchert, D Acosta Jr

38:63-96

K Ito, T Iwatsubo, S Kanamitsu, Y Nakajima,

38:461-99

Miscellaneous Techniques Transgenic Animals as New Approaches in Pharmacological Studies Predictive Value of In Vitro Model Systems in Toxicology

Quantitative Prediction of In Vivo Drug Clearance and Drug Interactions from In Vitro Data on Metabolism, and

Together with Binding and Transport

Y Sugiyama

The Impact of Genomics-Based Technologies on Drug Safety Evaluation

JF Waring, RG Ulrich

40:335-52

Y-Z Gu, JB Hogenesch,

40:519-61

Environmental Toxicity The PAS Superfamily: Sensors of Environmental and Developmental Signals

CA Bradfield

Pharmacology and Toxicology in the New Millennium Pharmacokinetic/Pharmacodynamic Modeling in Drug Development Sequencing the Entire Genomes of Free-Living Organisms: The Foundation of Pharmacology in the New Millenium High-Throughput Screening in Drug Metabolism and Pharmacokinetic Support of Drug Discovery Pharmacogenetics of Sulfotransferase

LB Sheiner, J-L Steimer

40:67—-96

S Broder,

40:97-132

JC Venter

RE White K Nagata, Y Yamazoe

40:133-57 40:159-76

934

CHAPTER TITLES

Drug Discovery in the Next Millennium

EH Ohlstein,

40:177-91

RR Ruffolo Jr,

JD Elliott

The Impact of Genomics on Drug Discovery Simulation of Clinical Trials

C Debouck, B Metcalf

NHG Holford, HC Kimko, JPR Monteleone, CC Peck

40:193-208 40:209-34

Soe